background background Proposed methods report for the Darwin catchments CSIRO logo A report from the CSIRO Northern Australia Water Resource Assessment to the Government of Australia CSIRO LAND AND WATER Cover image background ISBN: 978-1-4863-0715-9 Our research direction Provide the science to underpin Australia’s economic, social and environmental prosperity through stewardship of land and water resources ecosystems, and urban areas. Land and Water is delivering the knowledge and innovation needed to underpin the sustainable management of our land, water, and ecosystem biodiversity assets. Through an integrated systems research approach we provide the information and technologies required by government, industry and the Australian and international communities to protect, restore, and manage natural and built environments. Land and Water is a national and international partnership led by CSIRO and involving leading research providers from the national and global innovation systems. Our expertise addresses Australia’s national challenges and is increasingly supporting developed and developing nations response to complex economic, social, and environmental issues related to water, land, cities, and ecosystems. Citation CSIRO (2016) Proposed methods report for the Darwin catchments. A report from the CSIRO Northern Australia Water Resource Assessment to the Government of Australia. CSIRO, Australia. Copyright © Commonwealth Scientific and Industrial Research Organisation 2016. To the extent permitted by law, all rights are reserved and no part of this publication covered by copyright may be reproduced or copied in any form or by any means except with the written permission of CSIRO. Important disclaimer CSIRO advises that the information contained in this publication comprises general statements based on scientific research. The reader is advised and needs to be aware that such information may be incomplete or unable to be used in any specific situation. No reliance or actions must therefore be made on that information without seeking prior expert professional, scientific and technical advice. To the extent permitted by law, CSIRO (including its employees and consultants) excludes all liability to any person for any consequences, including but not limited to all losses, damages, costs, expenses and any other compensation, arising directly or indirectly from using this publication (in part or in whole) and any information or material contained in it. CSIRO is committed to providing web accessible content wherever possible. If you are having difficulties with accessing this document please contact csiroenquiries@csiro.au. CSIRO Northern Australia Water Resource Assessment acknowledgements This report was prepared for the Department of Agriculture and Water Resources. The Northern Australia Water Resource Assessment is part of the Australian Government’s Agricultural Competitiveness White Paper, the government’s plan for stronger farmers and a stronger economy. Important aspects of the Darwin Water Resource Assessment will be undertaken by the Northern Territory Government. This report was reviewed by Dr Glen Walker (Independent consultant), Mr Peter Jolly (Jolly Consulting) and Mr Clive Lyle (Independent consultant). Photo Snake bean, Humpty Doo, Northern Territory. Source: NT Farmers The Northern Australia Water Resource Assessment Team Project Director Peter Stone Project Leaders Chris Chilcott, Cuan Petheram, Ian Watson Project Support Caroline Bruce, Maryam Ahmad Communications Amy Edwards, Anne Lynch, Chris McKay Activities Agriculture and aquaculture viability Andrew Ash, Stuart Arnold, Mila Bristow1, Greg Coman, Rob Cossart2, Chris Ham2, Simon Irvin, Perry Poulton, Di Prestwidge, Clinton Revell2, Chris Stokes, Tony Webster, Bob Williams1, Stephen Yeates Climate Cuan Petheram, Daniel Aramini, Steve Charles Earth observation Neil Sims, Janet Anstee, Olga Barron, Elizabeth Botha, Eric Lehmann, Li Lingtao, Tim McVicar, Matt Paget, Luigi Renzullo, Catherine Ticehurst, Tom Van Niel, Garth Warren Ecology Carmel Pollino, Peter Bayliss, Rik Buckworth, Garry Cook, Aijun (Roy) Deng, Shane Griffiths, Rob Kenyon, Adam Liedloff, Linda Merrin, Christian Moeseneder, Eva Plaganyi-Lloyd, Barbara Robson, Danial Stratford Groundwater hydrology Andrew R. Taylor, Karen Barry, Kevin Cahill, Steven Clohessy2, Russell Crosbie, Phil Davies, Warrick Dawes, Rebecca Doble, Graham Herbert3, Tania Ibrahimi, Sandie McHugh2, Declan Page, Stan Smith, Steven Tickell1, Chris Turnadge, Joanne Vanderzalm, Des Yin Foo1, Ursula Zaar1 Indigenous aspirations and water values Marcus Barber, Carol Farbotko, Pethie Lyons, Emma Woodward Land suitability Ian Watson, Rebecca Bartley, Dan Brough3, Elisabeth Bui, Rob Dehayr3, Daniel Easey1, Neil Enderlin3, Luke Finn3, Mark Glover, John Grant2, Linda Gregory, Ben Harms3, Jason Hill1, Karen Holmes4, Jeremy Manders3, Angus McElnea3, David Morrison3, Seonaid Philip, Anthony Ringrose-Voase, Jon Schatz, Ross Searle, Henry Smolinski2, Mark Thomas, Dennis Van Gool2, Peter Wilson, Peter L. Wilson, Peter Zund3 Socio-economics Lisa McKellar, Carol Farbotko, David Fleming, Andrew Higgins, Nerida Horner, Neil MacLeod, Stephen McFallan, Sean Pascoe Surface water hydrology Justin Hughes, Dushmanta Dutta, Fazlul Karim, Steve Marvanek, Jin Teng, Jai Vaze, Bill Wang, Ang Yang Water storage Cuan Petheram, John Gallant, Geoff Hodgson, Klaus Joehnk, Arthur Read, Lee Rogers Note: Assessment team as at June 1, 2016. All contributors are affiliated with CSIRO unless indicated otherwise. Activity Leaders are underlined. 1Northern Territory Government, 2Western Australian Government, 3Queensland Government, 4CSIRO and Western Australian Government Shortened forms SHORT FORM FULL FORM AEM airborne electromagnetic APSIM Agricultural Production Systems Simulator AWRA-L Australian Water Resources Assessment – Landscape (model) AWRA-R Australian Water Resources Assessment – River (model) CMRSET CSIRO MODIS Reflectance-based Scaling EvapoTranspiration DEM digital elevation model DLRM Department of Land Resource Management (Northern Territory) ETa actual evapotranspiration EVI Enhanced Vegetation Index GCM global climate model GDE groundwater-dependent ecosystem GIS geographical information system GVMI Global Vegetation Moisture Index Landsat TM Landsat Thematic Mapper (multispectral sensor on Landsat 5 satellite) Landsat ETM+ Landsat Enhanced Thematic Mapper Plus (multispectral sensor on Landsat 7 satellite) Landsat OLI Landsat Operational Land Imager (multispectral sensor on Landsat 8 satellite) MAR managed aquifer recharge MGSH maximum gauged stage height MODIS MODerate-resolution Imagine Spectroradiometer MrVBF Multi-resolution Valley Bottom Flatness NDVI Normalised Difference Vegetation Index NDWI Normalised Difference Water Index NESP National Environmental Science Programme SRTM-H Shuttle Radar Terrain Missions SVAT Soil Vegetation Atmosphere Transfer TERN Terrestrial Ecosystem Research Network TraNSIT Transport Network Strategic Investment Tool Units UNIT DESCRIPTION GL gigalitre (1,000,000,000 litres) km kilometre (1000 metres) L litre m metre ML megalitre (1,000,000 litres) Preface Sustainable regional development is a priority for the Australian, Western Australian, Northern Territory and Queensland governments. In 2015, the Australian Government released the Northern Australia White Paper, which highlighted the opportunities for development of northern Australia’s water resources to enable regional development. In particular, many rural communities in northern Australia see irrigated agriculture as a means of reversing the long-term trend in population decline and as a critical element of broader regional development aspirations. This belief is supported by evidence from the southern Murray–Darling Basin, where studies have shown that irrigation production generates a level of economic and community activity that is three to five times higher than would be supported from rainfed (dryland) production. Development of northern Australia is not a new idea; there is a long history of initiatives to develop cultivated agriculture in the tropical north of Australia. Many of these attempts have not fully realised their goals, for a range of reasons. Most recently, it was highlighted that, although northern Australia’s environment poses challenges for irrigated agriculture, the primary reason that many of the schemes did not fully realise their goals is that they had insufficient capital to overcome the failed years that inevitably accompany every new irrigation scheme. The only large schemes still in operation in northern Australia had substantial government financial support at the construction phase, as well as ongoing support during establishment and learning phases. Northern Australia, however, is now seen to be located in the right place at the right time. Between 2000 and 2050, the world’s population is projected to grow from 6 to 9 billion people, and growth in food and fibre production is needed to meet an anticipated increase in demand. The majority of this growth is projected to occur in the tropics, particularly sub-Saharan Africa and South-East Asia. With two-thirds of the world’s food insecurity in Asia, sharp upward price movements in food have been identified as having the potential to result in political and social unrest. At the same time, it is projected that Asia will become home to the majority of the world’s middle class, which will result in an increasing demand for high-quality food produce from this region. The efficient use of Australia’s natural resources by food producers and processors is likely to increase the importance of understanding and sustainably managing Australia’s soil, water and energy resources. Finely tuned strategic planning will be required to ensure that investment and government expenditure on development are soundly targeted and designed. In terms of knowledge about, and development of, the natural resource base, northern Australia presents a relatively ‘blank slate’, with few ‘legacy issues’, particularly when compared with southern Australia. This presents a globally unique opportunity (a greenfield development opportunity in a first-world country) to strategically consider and plan the development of substantial areas of Australia. Most of northern Australia’s land and water resources have not been mapped in sufficient detail to support reliable resource allocation, mitigate investment or environmental risks, or provide policy settings that can support such decisions. Better data are required to enable private investment and government expenditure on development to be soundly targeted and designed, to account for intersections between existing and potential resource users, and to ensure that net development benefits are maximised. In 2013, the Australian Government commissioned CSIRO to undertake the Flinders and Gilbert Agricultural Resource Assessment in north Queensland. This assessment developed fundamental soil and water datasets, and provided a comprehensive and integrated evaluation of the feasibility, economic viability and sustainability of agricultural development in two catchments of northern Australia. It identified several opportunities for large-scale (>10,000 ha) irrigation development, based on the coincidence of suitable soils and new water storage capacity. The assessment provides a blueprint of the data and analysis required to identify and support actionable development opportunities in northern Australia. The outcome of the assessment was to reduce the uncertainty of investors and regulators, and to give the base information to allow development to occur in a sustainable manner. However, that work covered only 155,000 km2 (approximately 5%) of northern Australia, and acquiring a similar level of data and insight across northern Australia’s more than 3 million km2 would require more time and resources than are currently available. As a consequence, the Northern Australia White Paper prioritised about a dozen regions in northern Australia where more detailed water and agriculture resource assessments should be undertaken. It also provided $15 million to initiate the Northern Australia Water Resource Assessment in the Fitzroy catchment, Darwin catchments and Mitchell catchment (henceforth termed ‘regions’). The information from these Assessments will: • evaluate the soil and water resources • identify and evaluate water capture and storage options, and supply reliability • identify and test the commercial viability of agricultural opportunities, including irrigated agriculture, aquaculture and forestry • assess potential environmental, social and economic impacts and risks. The Fitzroy and Mitchell catchments were identified by the Northern Australia White Paper as being suitable candidates for a large-scale assessment of the economics and sustainability of irrigated agriculture because they appear to have large areas of soil suitable for irrigated agriculture, coincident with water. The river basins to the east of Darwin were chosen because they have the advantage of being close to a major population centre (Darwin) and have already seen much interest in development. Summary The Northern Australia Water Resource Assessment will provide a comprehensive and integrated evaluation of the feasibility, economic viability and sustainability of water resource development in three priority areas in northern Australia: the Fitzroy Catchment (Western Australia), the Darwin catchments (Northern Territory) and the Mitchell Catchment (Queensland). Each Assessment seeks to: • evaluate the climate, soil and water resources • identify and evaluate water capture and storage options • identify and test the commercial viability of irrigated agricultural, forestry and aquaculture opportunities • assess potential environmental, social and economic impacts and risks of water resource and irrigation development. In addition, each Assessment is designed to: • address explicitly the needs and aspirations for local development • meet the information needs of governments as they assess sustainable and equitable management of public resources, with due consideration of environmental and cultural issues • meet the due diligence requirements of private investors, by exploring questions of profitability and income reliability of agricultural and other developments. The objective of this report is to broadly outline the methods proposed for the Assessment in the Darwin catchments. The purpose is to openly communicate the scope of the Assessment and the proposed methods to a wide range of stakeholders, to allow them to provide feedback and engage with the Assessment team. The report also provides a mechanism for the Assessment team to acquire feedback on the proposed methods, to ensure that they are fit-for-purpose. The actual methods that the Assessment will use may differ as more information becomes available and local nuances are better understood. The final methods will be documented in technical detail in the final technical reports. The Assessment comprises following interrelated activities, which are discussed below. Climate opportunities and constraints The climate activity (Chapter 3) will seek to characterise the opportunities for water resource development, and agriculture and aquaculture production, in the context of current and future climates. The activity will generate several variants of 125 years of daily climate series representative of a 2.2 °C temperature rise relative to 1990, guided by global climate models (GCMs) from the Intergovernmental Panel on Climate Change Fifth Assessment Report (AR5) for an RCP8.5 emissions scenario. Pattern scaling will be used to transform the broadscale GCM outputs into catchment-scale variables that can be used by hydrological and agricultural simulation models. Availability of suitable soil The land suitability activity (Chapter 4) will develop digital land suitability maps of the entire Darwin catchments, showing areas that are more and less suitable under different combinations of land use and irrigation systems, and aquaculture. The activity will employ statistically based digital soil mapping methods to rapidly and objectively generate 30 m × 30 m grids of a wide range of soil attributes (e.g. depth of soil, texture, pH). The digital soil mapping will be informed by a limited soil sampling campaign. The existing set of rules for different combinations of land use and irrigation systems, and aquaculture will be refined. The rules will be applied to the digital soil mapping grids and climate grids to generate land suitability maps of the catchments. Availability of water The availability of surface water across the Darwin catchments will primarily be assessed using three types of hydrological models: (i) landscape model (AWRA-L), (ii) river system model, and (iii) hydrodynamic model (MIKEFLOOD) (see surface water hydrology activity, Chapter 5). The landscape model will be used to quantify water fluxes across the Darwin catchments. These fluxes will be used as input to the river system and hydrodynamic models. River system models are well suited to modelling regulated systems, and exploring how streamflow may be perturbed under future development, management and climate scenarios. The river system modelling provides an integrating framework for analysing the opportunities by which surface water development may enable regional development. Hydrodynamic models are physically based models that explicitly model the movement of water across the landscape. These models will be used to examine how large and small flood events, and the connectivity of offstream wetlands and the main river channel are perturbed by future development and climate scenarios. Interim digital soil attributes generated as part of the digital soil mapping process (Chapter 4) will be used to help parameterise all three hydrological models. Interim digital land suitability maps (Chapter 4), potential dam locations (Chapter 9) and key ecological assets (Chapter 12) will inform the structure of the river system model. The latter will inform the domain of the hydrodynamic model. The groundwater hydrology activity (Chapter 6) seeks to provide a comprehensive assessment of the most promising regional aquifers in the Darwin catchments, in the context of identifying opportunities for, and risks associated with, groundwater resource development to enable regional development. In the Darwin catchments, the field component of the groundwater activity will focus on the Wildman and Mary river basins. This will complement field hydrogeological investigations currently being conducted in the Wildman River Basin by the Northern Territory Government. Tasks being undertaken by the groundwater activity comprise: (i) mapping the spatial extent of the main aquifers, (ii) estimating aquifer hydraulic properties, (iii) mapping groundwater-dependent ecosystems (GDEs) and measuring their water use, (iv) estimating the scale of the flow system from water level mapping, (v) quantifying recharge using a wide variety of methods, (vi) measuring spring discharge, and (vii) applying simple analytical groundwater models to investigate the potential impacts of pumping on groundwater levels near GDEs. One of the challenges of working in northern Australia is the scarcity of data and the remoteness of the landscape. For these reasons, the Assessment will seek to use remotely sensed imagery (i.e. satellite imagery) where it can meaningfully inform the information needs of the Assessment. The Earth observation chapter (Chapter 7) details the methods that require detailed processing of remotely sensed data to generate products to help other activities achieve their objectives. This viii | Proposed project methods activity will primarily use data from MODIS Terra and AQUA satellites, and archival multi-temporal Landsat imagery from the Australian Geoscience Data Cube. The work will involve mapping flood inundation (to help constrain the hydrodynamic modelling), identifying persistent waterholes (key ecological refugia and critical for Chapter 12), measuring temporal changes in waterhole turbidity (to identify waterholes that are most susceptible to development, and waterholes that may in part be being replenished by groundwater), undertaking fine-scale spatial measurements of vegetation water use (to inform vegetation water use and watering requirements), mapping the extent of riparian vegetation, plant biomass and growth vigour, undertaking fine-scale (25 m), and mapping soil water content at a finer scale than is currently available. Agriculture and aquaculture viability The agriculture and aquaculture viability activity (Chapter 8) will work closely with the socio- economics activity (Chapter 10) to fully integrate biophysical agriculture and aquaculture production with an economic assessment. The activity will undertake crop and forage modelling and analysis using the Agricultural Production Systems Simulator (APSIM), the Grass Production Model (GRASP), and expert knowledge and experience. Some limited field studies will be undertaken as part of this activity to assist in validating crop and forage models, and estimates of crop and forage production. Although the agricultural activity will analyse individual crops and forages to help provide fundamental information on potential yields, water use, growing seasons and gross margins, for example, the aim is not to be prescriptive about cropping systems for particular locations; rather, the aim is to provide insights into the issues and opportunities associated with developing integrated cropping or crop–livestock systems, as opposed to individual crops. The aquaculture viability activity will assess the viability of irrigated and near-shore land-based aquaculture. Tasks will include an aquaculture pond suitability analysis, using information generated by digital soil mapping (Chapter 4) and hydrological modelling (Chapter 5). The activity will also include an economic analysis of various aquaculture enterprises, focusing on the opportunities by which aquaculture enterprises could complement irrigated cropping and grazing in the Darwin catchments. In conjunction with the socio-economics activity (Chapter 10), and the Indigenous aspirations and water values activity (Chapter 11), the opportunities for Indigenous aquaculture enterprises and aquaculture-related legislative constraints, respectively, will be explored. Water storage options The water storage activity (Chapter 9) will provide a comprehensive overview of the different water storage options in the Darwin catchments, to help decision makers take a long-term view of water resource development and to inform future allocation decisions. The construction of inappropriate storages and incremental releases of water can preclude the development of more appropriate water storages and water development options. The work will include a pre-feasibility assessment of large instream and offstream dams, including examining the opportunities for hydro-electric power generation. The activity will also include an assessment of large on-farm (e.g. 2 to 8 GL) hillside dams and ring tanks. The river system models (Chapter 5) will be used to explore how the reliability of harvesting water into ring tanks decreases with increasing catchment allocation and extraction, and other factors such as pumping capacity and the threshold above which water can be taken. Digital soil maps (Chapter 4) will be used to provide information on areas that are suitable for on-farm storage, such as ring tanks. One-dimensional hydrodynamic modelling will be used to assess the water level and flow at which natural floodways (which could be impounded) begin to flow and, using the river system models, the reliability with which they flow. This activity will also seek to quantify the volume of natural wetlands and waterholes in the Assessment area – there is increasing interest in exploring the opportunities by which water from wetlands could be used for consumptive purposes, while maintaining their ecological function. Finally, in conjunction with the groundwater hydrology activity, the opportunities for using managed aquifer recharge to store water below the ground surface will be assessed at a regional scale. Socio-economics of water resource development The socio-economics activity (Chapter 10) will examine the economic viability of development of water resources for food and fibre production, and the opportunities to use water for multiple purposes. Although the economics of agriculture will be examined as part of the agriculture and aquaculture viability activity (Chapter 8), the analysis will be limited to the farm scale. The socio- economics activity will extend the economic analysis to the scheme scale, using industry standard cost–benefit analysis methods. At the regional scale, econometric models will be used to examine regional socio-economic impacts of different types and scales of water resource development. As part of these analyses, the activity will apply the TraNSIT transport logistics model to quantify logistics costs, and identify logistics infrastructure bottlenecks, and best-bet infrastructure investment and regulatory strategies to reduce road access issues. Information on the possible locations and scale of water resource development will be provided by the land suitability (Chapter 4), surface water hydrology (Chapter 5), groundwater hydrology (Chapter 6) and water storage (Chapter 9) activities. Information on integrated cropping or crop–livestock and aquaculture systems will be provided by the agriculture and aquaculture activity (Chapter 8). The socio-economics activity will also examine how the values, perspectives and behaviours of individuals, groups and institutions may influence the nature and structure of any economic production. It will also seek to understand the policy and regulatory environment for the assessment of water resources and development of water-dependent industries. Indigenous aspirations and water values The Indigenous aspirations and water values activity will provide an overview of key Indigenous values, rights, interests and aspirations with respect to water and irrigated agricultural development in the Darwin catchments. This analysis is intended to assist, inform and underpin future discussions between developers and Indigenous people about particular developments, and their potential positive and negative effects on Indigenous populations. The activity will closely align with components of the agriculture and aquaculture viability activity (Chapter 8), the socio- economics activity (Chapter 10) and the ecology activity (Chapter 12). The fieldwork component of this activity will emphasise direct consultation with Indigenous Traditional Owners of, and residents in, the Assessment area. This will be undertaken through a variety of means, including telephone discussions, face-to-face interviews, group meetings and workshops. Other key components of the activity include a cultural heritage assessment, and a legal and policy analysis. Freshwater, riparian and marine ecology The ecology activity (Chapter 12) seeks to assess the potential for possible changes in flow regimes associated with new infrastructure across the Assessment area to affect freshwater and marine ecosystems. The Assessment focuses on water-related ecosystems (including terrestrial GDEs) because water developments, particularly irrigation, can result in substantial changes to streamflow, although typically water developments occupy only a small proportion of the landscape (<1% of a catchment). Impacts on terrestrial systems will be dealt with in an issues paper, which will identify future assessment priorities in the Assessment areas. Key tasks in the ecology activity will include identifying and prioritising assets in the Assessment area, for which conceptual models that capture flow–ecology relationships will be developed. A multiple-lines-of- evidence approach will be used to develop relationships between flow and ecology. These will be qualitative where information is poor and semi-quantitative or quantitative where information is rich. The activity will use hydrological outputs from the surface water hydrology modelling (Chapter 5) to which the flow–ecology relationships will be applied to identify likely ecological changes to freshwater and marine ecosystems as a result of different types and scales of water resource development. Similarly, hydrological data from the groundwater hydrology activity (Chapter 6) will be used to identify possible ecological change to key GDEs as a result of different levels of groundwater extraction. The trade-offs of water resource development will be examined in more detail in the case study experiments (Chapter 13). Case study experiments The Assessment will also undertake a small number of case study experiments in the Darwin catchments (Chapter 13). Their purpose is to show the reader how to ‘put everything together’ to answer their own questions about water resource development. As well, they aim to help readers understand the type and scale of opportunity for irrigated agriculture or aquaculture in selected geographic parts of the Assessment area, and explore some of the nuances associated with greenfield developments in the catchments, and northern Australia in general. Importantly, they are not designed to demonstrate, recommend or promote particular development opportunities being proposed by individual development proponents, nor are they recommendations on how development in the Darwin catchments should unfold. They are, however, designed to be realistic representations, and will explore a variety of potential water resource development options and scales of development. The case study experiments will draw on information, expertise and models from all activities in the Assessment. Contents The Northern Australia Water Resource Assessment Team ........................................................... i Shortened forms .............................................................................................................................iii Units ............................................................................................................................... iv Preface ............................................................................................................................... v Summary .............................................................................................................................. vii Part I 1 1 Introduction ........................................................................................................................ 2 1.1 Northern Australia Water Resource Assessment .................................................. 2 1.2 Reporting schedule ................................................................................................ 4 1.3 Review process ...................................................................................................... 5 1.4 Past studies and links to relevant current projects ............................................... 5 1.5 Objectives and contents of this report .................................................................. 7 1.6 References ............................................................................................................. 9 2 Key concepts ..................................................................................................................... 10 2.1 Water year, and wet and dry seasons ................................................................. 10 2.2 Scenario definitions ............................................................................................. 10 2.3 Case study experiments ...................................................................................... 11 2.4 References ........................................................................................................... 12 Part II 13 3 Climate ............................................................................................................................. 14 3.1 Climate opportunities for agriculture and aquaculture production ................... 14 3.2 Generation of historical and future climate time series ..................................... 15 3.3 References ........................................................................................................... 21 4 Land suitability .................................................................................................................. 23 4.1 Introduction ......................................................................................................... 23 4.2 Soil and land assessment ..................................................................................... 26 4.3 Data management and storage ........................................................................... 32 4.4 Assessment outputs ............................................................................................ 32 4.5 References ........................................................................................................... 33 5 Surface water hydrology ................................................................................................... 35 5.1 Introduction ......................................................................................................... 35 5.2 Model overview ................................................................................................... 35 5.3 Model calibration and modelling experiments ................................................... 38 5.4 Surface water quality ........................................................................................... 42 5.5 References ........................................................................................................... 43 6 Groundwater hydrology ................................................................................................... 45 6.1 Introduction ......................................................................................................... 45 6.2 Hydrogeology of the Darwin catchments............................................................ 47 6.3 Field hydrogeological investigations ................................................................... 48 6.4 Recharge and discharge estimation .................................................................... 54 6.5 Assessing the potential impacts of groundwater development ......................... 56 6.6 Risk of irrigation-induced salinity ........................................................................ 57 6.7 References ........................................................................................................... 58 7 Earth observation ............................................................................................................. 63 7.1 Flood inundation and waterhole persistence ..................................................... 64 7.2 Waterhole suspended sediment ......................................................................... 66 7.3 Riparian vegetation ............................................................................................. 67 7.4 High-resolution estimates of vegetation water use ............................................ 68 7.5 Soil water availability ........................................................................................... 68 7.6 References ........................................................................................................... 69 Part III 71 8 Agriculture and aquaculture viability ............................................................................... 72 8.1 Agriculture viability.............................................................................................. 72 8.2 Aquaculture viability ............................................................................................ 79 8.3 References ........................................................................................................... 84 9 Water storage ................................................................................................................... 87 9.1 Introduction ......................................................................................................... 87 9.2 Large instream and offstream storages .............................................................. 88 9.3 Farm-scale instream and offstream storages ...................................................... 92 9.4 Quantifying the volume of natural wetlands and waterholes ............................ 93 9.5 Managed aquifer recharge .................................................................................. 94 9.6 References ........................................................................................................... 99 10 Socio-economics ............................................................................................................. 102 10.1 Overview ............................................................................................................ 102 10.2 Multiscale economic assessment of water resources development for agriculture ....................................................................................................................... 103 10.3 Economic assessment of water resources development for aquaculture ........ 104 10.4 Application of TraNSIT to identify logistics infrastructure bottlenecks and opportunities .................................................................................................................. 104 10.5 Exploring stakeholder perspectives................................................................... 105 10.6 Understanding the policy and regulatory environment for the assessment of water resources, and development of water-dependent industries ............................. 107 10.7 Regional socio-economic impacts ..................................................................... 107 10.8 References ......................................................................................................... 108 11 Indigenous aspirations and water values ....................................................................... 109 11.1 Introduction ....................................................................................................... 109 11.2 Linkages to other Assessment activities ............................................................ 110 11.3 Linkages to other research projects in the Darwin catchments ....................... 110 11.4 Context and consultation .................................................................................. 110 11.5 Scope ................................................................................................................. 111 11.6 Research ethics .................................................................................................. 111 11.7 Methods ............................................................................................................ 112 11.8 Data analysis and preliminary dissemination .................................................... 114 11.9 References ......................................................................................................... 114 Part IV 117 12 Ecology ........................................................................................................................... 118 12.1 Region overview ................................................................................................ 119 12.2 Ecology activity breakdown ............................................................................... 119 12.3 References ......................................................................................................... 124 Part V 127 13 Case study experiments .................................................................................................. 128 13.1 Rationale ............................................................................................................ 128 13.2 Proposed case study framings ........................................................................... 128 14 Reports, products, protocols and standards .................................................................. 131 14.1 Reports, products and protocols ....................................................................... 131 14.2 Standards ........................................................................................................... 131 Figures Figure 1-1 Governance framework of the Northern Australia Water Resource Assessment ........ 4 Figure 1-2 The Darwin catchments ................................................................................................. 7 Figure 1-3 Schematic diagram illustrating the high-level linkages between the nine activities .... 9 Figure 3-1 Decadal analysis of the location and completeness of Bureau of Meteorology stations measuring daily rainfall used in the SILO database ...................................................................... 17 Figure 3-2 Decadal analysis of the location and completeness of Bureau of Meteorology stations measuring daily maximum air temperature used in the SILO database ...................................... 18 Figure 4-1 Darwin catchments, showing drainage, major and minor roads and legacy soil sampling sites. ............................................................................................................................... 24 Figure 5-1 Darwin catchment rivers (including South and East Alligator rivers to the east of the Darwin catchments), showing gauge location and gauge record information ............................ 38 Figure 5-2 Flood extent based on MODIS data (green), and hydrodynamic model domain (red polygons). HD is an abbreviation of “Hydro-dynamic” ................................................................. 43 Figure 6-1 Example of the spatial variability in the mean conductivity and resistivity of the subsurface at a depth of 40 to 60 m for existing airborne electromagnetic data in the Darwin catchments .................................................................................................................................... 51 Figure 6-2 Schematic representation of acquisition of airborne electromagnetic data (in this example, a time domain electromagnetic system) and interpretation ........................................ 52 Figure 8-1 Proposed market driven approach to agricultural viability and economic assessment .................................................................................................................................... 74 Figure 9-1 Schematic diagram illustrating various techniques that can be used for managed aquifer recharge, depending on the hydrogeological conditions ................................................ 96 Tables Table 1-1 Key deliverables .............................................................................................................. 5 Table 1-2 Other current projects with relevance to the Northern Australia Water Resource Assessment ..................................................................................................................................... 6 Table 4-1 Description of the covariates potentially available for use in the statistical analysis to support the sampling design, digital soil mapping attributes, and soil and land use suitability mapping. Each covariate layer has a cell dimension of 30 x 30 m ............................................... 27 Table 4-2 Soil attributes and methods of analysis ........................................................................ 30 Table 4-3 Selection of Darwin catchments land suitabilities, including engineering and a selection of prospective crops in broad categories ...................................................................... 32 Table 7-1 Spatial, temporal and spectral characteristics of selected satellite data sources that will be used in the Earth observation activity ............................................................................... 64 Table 8-1 Summary of the economic analytical approach proposed for different scales (paddock, farm, and irrigation scheme) ....................................................................................... 75 Table 8-2 Land use category and crop/forage modules currently available in APSIM that can be used in this Assessment ................................................................................................................ 77 Table 8-3 Pond types ..................................................................................................................... 81 Table 8-4 Limitations identified in preliminary aquaculture pond suitability analysis................. 81 Table 9-1 Major dams in the Darwin catchments (CSIRO 2009) .................................................. 88 Table 9-2 Proposed methods for assessing potential dam sites in the Darwin catchments ........ 90 Table 9-3 Types of offstream water storages (Lewis, 2002) ......................................................... 92 Table 9-4 Parameters likely to be included in a region-scale assessment of the suitability of infiltration-based managed aquifer recharge ............................................................................... 98 Part I Introduction 1 Introduction 1.1 Northern Australia Water Resource Assessment The Northern Australia Water Resource Assessment will provide a comprehensive and integrated evaluation of the feasibility, economic viability and sustainability of water resource development in three priority areas in northern Australia: the Fitzroy catchment (Western Australia), the Darwin catchments (Northern Territory) and the Mitchell catchment (Queensland). Each Assessment seeks to: • evaluate the climate, soil and water resources • identify and evaluate water capture and storage options • identify and test the commercial viability of irrigated agricultural, forestry and aquaculture opportunities • assess potential environmental, social and economic impacts and risks of water resource, aquaculture and irrigation development Techniques and approaches will be specially tailored for each of the Assessment areas. It is important to note that, although these four points appear in sequence, activities in one part of the Assessment will often inform (and hence influence) activities in an earlier part. For example, understanding ecological and cultural watering requirements (the third part of the Assessment, described in Part IV of this report) is particularly important in setting rules around water extraction and diversion (i.e. how much water can be taken and when it should be taken – the second part of the Assessment, described in Part III of this report). Thus, the procedure of assessing a site will inevitably include iterative steps, rather than a simple linear process. In covering the above points, each Assessment is designed to: • address explicitly the needs and aspirations of local development by providing objective assessment of resource availability, with consideration of the environmental and cultural issues • meet the information needs of governments as they assess sustainable and equitable management of public resources, with due consideration of environmental and cultural issues • meet the due diligence requirements of private investors, by exploring questions of profitability and income reliability of agricultural and other developments. Drawing on the resources of all three tiers of government, the Assessment will build on previous studies, draw on existing stores of local knowledge, and employ world-class scientific expertise, with the quality assured through peer-review processes. The Northern Australia Water Resource Assessment commenced on 16 December 2015 and will be completed by 30 June 2018. 1.1.1 SCOPE OF ASSESSMENTS In stating what the Assessments will do, it is equally instructive to state what they will not do. The Assessments will not advocate irrigation development. They will identify the resources that could be deployed in support of irrigation and aquaculture enterprises, and the scale of the opportunity that might exist. They are designed to evaluate the feasibility of development (at a catchment scale), not to enable particular developments. The Assessments will quantify the monetary and non-monetary values associated with existing use of resources, to enable a wide range of stakeholders to assess the likely costs and benefits of given courses of action. The Assessments are fundamentally resource assessments, the results of which can be used to inform planning decisions by citizens; councils; investors; and the state, territory and Australian governments. The Assessments do not seek to replace any planning processes, and they will not recommend changes to existing plans or planning processes. The Assessments will not invest or promote investment in infrastructure that may be required to support irrigation enterprises. They seek to lower barriers to investment in the Assessment area by exploring many of the questions that potential investors might have about production systems and methods, yield expectations and benchmarks, and potential profitability and reliability. This information will be established for each Assessment area, not for individual paddocks or farms. The Assessments do not assume that particular areas within the Assessment areas are in or out of scope. For example, the Assessments are ‘blind’ to issues such as land tenure that may exclude land parcels from development. The Assessments will identify those areas that are best suited for new agricultural, aquaculture developments and industries, and, by inference, those that are not well suited. The Assessments do not assume particular types or scales of water storage or water access. They will identify the types and scales of water storage and access arrangements that might be possible, and the likely consequences (both costs and benefits) of pursuing these possibilities. Having done that, they will not recommend preferred development possibilities, nor comment on the required regulatory requirements to make those water resources available. The Assessments will not assume a given regulatory environment. They will evaluate the availability and use of resources in accordance with existing regulations, but will also examine resource use unconstrained by regulations, to ensure that the results can be applied to the widest possible range of uses for the longest possible time frame. It is not the intention – and nor will it be possible – for the Assessments to address all topics related to irrigation and aquaculture development in northern Australia, due to time and/or resource constraints. Important topics that are not addressed by the Assessments (e.g. impacts of irrigation development on terrestrial ecology) are discussed with reference to, and in the context of, the existing literature. No attempt has been made to identify such topics in this report. Functionally, the Assessments will adopt an activities-based approach to the work (which is reflected in the content and structure of the outputs and products, as per Section 1.2), with the following activity groups: climate, land suitability, surface water hydrology, groundwater hydrology, agriculture and aquaculture viability, water storage, socio-economics, Indigenous aspirations and water values, and aquatic and marine ecology. These activities will be supported by an Earth observation activity, which will apply across the other activities. 1.1.2 PROGRAM GOVERNANCE FRAMEWORK The Program Governance Committee will provide high-level governance and leadership to the Northern Australia Water Resource Assessment (Figure 1-1). The Program Governance Committee will meet every 6 months and act as a conduit to government stakeholders in their respective jurisdictions. Each Assessment area will have a steering committee, which will guide the Northern Australia Water Resource Assessment Team. The committee will regularly report to key stakeholders on the program, and ensure that the expectations of stakeholders are considered and responded to appropriately. The Northern Australia Water Resource Assessment Team will plan, manage and deliver the Assessment. The team will consist of CSIRO staff, augmented with contracts to jurisdictions, universities and private contractors, where necessary. For more information on this figure please contact CSIRO on enquiries@csiro.au Figure 1-1 Governance framework of the Northern Australia Water Resource Assessment 1.2 Reporting schedule The contracted deliverables for each of the three Assessments are a suite of reports: • Technical reports present scientific work at a level of detail sufficient for technical and scientific experts to understand the work. Each of the activities of the Assessment has a corresponding technical report. • Three catchment reports – one for each Assessment area – synthesise key material from the technical reports, providing well-informed but non-scientific readers with the information required to make decisions about the opportunities, costs and benefits associated with irrigated agriculture. • Three overview reports – one for each Assessment area – are provided for a general public audience. • Three fact sheets – one for each Assessment area – provide key findings for all three Assessment areas for a general public audience. The dates for completion of key deliverables are listed in Table 1-1. Table 1-1 Key deliverables DELIVERABLE DATE FOR COMPLETION Methods report: Darwin catchments 30 June 2016 Catchment reports: Darwin catchments 30 May 2018 16-page summary report: Darwin catchments 30 May 2018 Final fact sheet: Darwin catchments 30 June 2018 1.3 Review process As part of CSIRO’s internal quality assurance process, all reports produced by the Assessment will be reviewed. CSIRO will manage the review process in accordance with CSIRO’s ePublish protocols. Technical reports will be reviewed by at least two reviewers. Additional comment will be sought from the Northern Territory Government, depending on the topic and content. A combination of external and internal reviewers will be used. After review and revisions in response to the review, each report will be sent to the Australian Government Department of Agriculture and Water Resources for final approval. 1.4 Past studies and links to relevant current projects A key component of the Assessment will be the collation and review of relevant literature, which will be a prerequisite for all activities. The methods described in this report will be modified to take into account the availability (and sometimes lack) of information on the Assessment area. The Assessment team will rely in part on the Program Steering Committee to ensure that all relevant literature is captured and to help identify local experts who should be consulted. Within CSIRO, the Assessment will link to relevant current projects, outlined in Table 1-2. Table 1-2 Other current projects with relevance to the Northern Australia Water Resource Assessment NAME OF PROJECT FUNDING OBJECTIVES TraNSIT: unlocking options for efficient logistics infrastructure in Australian agriculture Australian Government Agriculture Competitiveness White Paper Initiative To provide governments, industry, the farming community and other stakeholders with a baseline of freight transport costs between Australian agricultural value-chain enterprises, along with the capacity to identify and evaluate a range of scenarios to minimise transport costs and maximise long-term profitability Delivering capability into Northern Australia: exploring opportunities for multiple use of wetlands CSIRO Strategic Funding To examine the opportunities for using water from northern Australia’s wetlands for development purposes, while maintaining wetland values Cropping opportunities for northern Australia CSIRO Strategic Funding To better understand the research and development needs of corporate farming entities in northern Australia, and to do a capability audit of CSIRO’s ability to meet those needs National Environmental Science Programme (NESP) Northern Hub in the Fitzroy and Mitchell regions Australian Government To support sustainable development in northern Australia and inform practical solutions to the regions’ major environmental challenges 1.4.1 THE DARWIN WATER RESOURCE ASSESSMENT AREA The Assessment area is defined by the Finniss, Adelaide, Mary and Wildman Australian Water Resource Council river basins (Figure 1-2). Collectively, they encompasses an area of about 30,000 km2. The city of Darwin is located within the Finniss River Basin. The Adelaide and Finniss rivers are critical for Darwin’s domestic water supply. The Darwin catchments are tropical, and rainfall is highly seasonal. The wet season (November to April) accounts for 95% of annual rainfall and 96% of annual runoff (CSIRO, 2009). Annually, potential evaporation is greater than precipitation, and approximately 27% of precipitation is transformed to runoff. Large variability in annual runoff occurs, and the strong seasonality in rainfall results in large wet-season flows and small dry-season flows (CSIRO, 2009). The ecology of the catchments is adapted to the high seasonality and variability typical of tropical river systems. For more information on this figure please contact CSIRO on enquiries@csiro.au Figure 1-2 The Darwin catchments 1.5 Objectives and contents of this report The objective of this report is to broadly outline the methods that the Assessment intends to employ. The purpose is to openly communicate the scope of the Assessment and the proposed methods to a wide range of stakeholders, to allow them to provide feedback and engage with the Assessment team. The report also provides a mechanism for the Assessment team to acquire feedback on the proposed methods, to ensure that they are fit-for-purpose. The actual methods that the Assessment will use may vary as more information becomes available, and will be documented in detail in the technical reports. The Assessment is divided into 16 activities. Figure 1-3 illustrates the high-level linkages between the activities (in blue boxes) and the general flow of information in the Assessment. The figure does not seek to capture all linkages and dependencies between activities. This report is structured to align with the following three central questions (in italics below) that encompass the four deliverable points listed in Section 1.1, as well as the activities shown in Figure 1-3: •Part I – Introduction provides an overview of the Darwin Catchments and defines theAssessment Area and key concepts. –Chapter 1 – Introduction –Chapter 2 – Key concepts. •Part II – Resource assessment addresses the question ‘What soil and water resources areavailable to support regional development?’ by describing the information and methods neededto identify, map and quantify the available soil and water resources. The following Chapterspresent methods in Part II: –Chapter 3 – Climate –Chapter 4 – Land suitability –Chapter 5 – Surface water hydrology –Chapter 6 – Groundwater hydrology –Chapter 7 – Earth observation. •Part III – Economic viability addresses the question ‘What are the opportunities by which waterresource development may enable regional development?’ by evaluating the opportunities foragriculture and aquaculture, water storage, and supply of water for multiple uses, includingurban and hydro-electric power generation. It also evaluates the economic costs and benefits, and regional socio-economic impacts of these opportunities. The following Chapters presentmethods in Part III: –Chapter 8 – Agriculture and aquaculture viability –Chapter 9 – Water storage –Chapter 10 – Socio-economics –Chapter 11 – Indigenous aspirations and water values. •Part IV –Achieve a balance between competing priorities by addressing the question ‘How candevelopment opportunities be maximised by understanding, quantifying and managing impactsand trade-offs with existing industries and ecosystems?’ The following Chapters presentmethods in Part IV: –Chapter 12 – Aquatic, riparian and marine ecology. •Part V – Case study experiments, reports, key protocols and standards describes the rationalefor undertaking case studies, summarises the reports that will be delivered by the Assessmentand outlines key protocols for data management. –Chapter 13 – Case study experiments –Chapter 14 – Reports, products, protocols and standards. For more information on this figure please contact CSIRO on enquiries@csiro.au Figure 1-3 Schematic diagram illustrating the high-level linkages between the nine activities 1.6 References CSIRO (2009) Water in the Timor Sea Drainage Division. A report to the Australian Government from the CSIRO Northern Australia Sustainable Yields Project. CSIRO Water for a Healthy Country Flagship, Australia, 363–452. 2 Key concepts 2.1 Water year, and wet and dry seasons The Assessment area experiences a highly seasonal climate, with the majority of rain falling between December and March. Unless specified otherwise, the wet season is defined as the 6- month period from 1 November to 30 April and the dry season as the 6-month period from 1 May to 31 October. All results in the Assessment are reported over the ‘water year’, defined as the period 1 September to 31 August, which allows each wet season to be counted in a single 12- month period, rather than being split over two calendar years (i.e. counted as two separate seasons). This is the usual convention for reporting climate statistics in northern Australia, as well as from a hydrological and agricultural assessment viewpoint. 2.2 Scenario definitions The Assessment will consider four different scenarios of climate, surface water, groundwater and economic development, as used in the Northern Australia Sustainable Yields Project (CSIRO, 2009a, b, c): • Scenario A – historical climate and current development • Scenario B – historical climate and future development • Scenario C – future climate and current development • Scenario D – future climate and future development. 2.2.1 SCENARIO A Scenario A will include historical climate and ‘current’ development. The historical climate data will be for 125 years (water years from 1 September 1890 to 31 August 2015) of observed climate (rainfall, temperature and potential evaporation for water years). All results will be presented over this period, unless specified otherwise. Current development is the defined here as the level of surface water, groundwater and economic development as at 31 August 2015. The Assessment will assume that all current water entitlements are being fully used. Historical tidal data will be used to specify downstream boundary conditions for flood modelling undertaken in the Assessment. Scenario A will be used as the baseline against which assessments of relative change are made. 2.2.2 SCENARIO B Scenario B will include historical climate and future water resource developments. Scenario B will use the same historical climate data as Scenario A. Future water resource development will be described by each case study storyline, and river inflow and water extractions will be modified accordingly. Case study storylines will be developed in consultation with key stakeholders (see Section 0). The impacts of changes in flow regime due to future development will be assessed and compared to other scenarios including: • impacts on instream, riparian and near-shore ecology • impacts on Indigenous water values • economic costs and benefits • opportunity costs of expanding irrigation • institutional, economic and social considerations that may impede or enable adoption of irrigated agriculture. 2.2.3 SCENARIO C Scenario C will include future climate and current development. It will be based on the 125-year climate data sequence, scaled for conditions in about 2060. These climate data will be derived from a range of global climate model (GCM) projections for a 2.2 °C global temperature rise scenario, which encompasses different GCMs from the Intergovernmental Panel on Climate Change Fifth Assessment Report (AR5) under an RCP8.5 emissions scenario (RCP8.5 is one of four greenhouse gas concentration (not emissions) trajectories adopted by the IPCC for its fifth Assessment Report (AR5) in 2014); the projections will then be used to modify the observed historical daily climate sequences (see Chapter 3). Like Scenario A, current development is the level of surface water, groundwater and economic development as at 31 August 2015. Tidal level data will be manipulated to reflect a sea-level rise for about 2060 (i.e. the median date at which the GCMs reach a 2.2 °C global temperature rise). 2.2.4 SCENARIO D Scenario D is future climate and future development. It will use the same future climate series as Scenario C. River inflow, groundwater recharge and flow, and water extraction will be modified to reflect proposed future development, in the same way as in Scenario B. 2.3 Case study experiments The case study experiments in the Assessment will be used to provide examples of how information produced by the Assessment can be assembled to help the reader ‘answer their own questions’. They will also help readers understand the type and scale of opportunity for irrigated agriculture or aquaculture in selected geographic parts of the Assessment area, and explore some of the nuances associated with greenfield developments in the Darwin catchments. Importantly, the case study experiments are illustrative only. They are not designed to demonstrate, recommend or promote particular development opportunities being proposed by individual development proponents, nor are they CSIRO’s recommendations on how development in the Darwin catchments should unfold. However, they are designed to be realistic representations. That is, the case study experiments will be ’located‘ in specific parts of the Assessment area, and use specific water and land resources, and realistic intensification options. For more information on the case study experiments, see Chapter 13. 2.4 References CSIRO (2009b) Water in the Gulf of Carpentaria Drainage Division. A report to the Australian Government from the CSIRO Northern Australia Sustainable Yields Project. CSIRO Water for a Healthy Country Flagship, Australia. CSIRO (2009a) Water in the Timor Sea Drainage Division. A report to the Australian Government from the CSIRO Northern Australia Sustainable Yields Project. CSIRO Water for a Healthy Country Flagship, Australia. CSIRO (2009c) Water in the Northern North-East Coast Drainage Division. A report to the Australian Government from the CSIRO Northern Australia Sustainable Yields Project. CSIRO Water for a Healthy Country Flagship, Australia. Part II Resource assessment: What soil and water resources are available to support regional development? 3 Climate Climate variables are generally considered, together with soil data, to be the most important environmental factors in determining the suitability of particular locations for agriculture. And because climate is so very closely linked to hydrology and water availability, understanding of climate and its variability is especially important in assessments of semi-arid and subtropical sites in northern Australia for irrigated land use. This activity will characterise the current and future climate of the Darwin catchments in the context of the opportunities for water resource development, and agriculture and aquaculture productivity (Section 3.1). Section 3.2 then describes methods by which future climate time series are generated for use by the land suitability activity (Chapter 4), surface water hydrology activity (Chapter 5), groundwater hydrology activity (Chapter 6) and agricultural productivity activity (Chapter 8). Historical and future climate and development scenarios are defined in Chapter 2. The key questions that this activity seeks to address in the Darwin catchments include: • What are the large- and small-scale processes controlling climate? • How do climate variables vary spatially and temporally? • What is an appropriate timescale over which to assess the water resources? • Are there decadal wetter and drier patterns in the observed record, and do they correlate with well-known modes of climate variability? • What does the paleoclimatic record tell us about rainfall variability in northern Australia’s geological past? • What is the accuracy of short-term and seasonal climate and streamflow forecasting? • What is the drought, cyclone and flood risk to agriculture, aquaculture and infrastructure? • Do sea surface temperatures, salinity concentrations and storm surges in the Darwin catchments pose a risk to aquaculture production? • How climatically suitable is the Darwin catchments for food, fibre and tree production? • Do the recent (AR5) future climate projections indicate that the Darwin catchments will be ‘wetter’ in about 2060 or ‘drier’? 3.1 Climate opportunities for agriculture and aquaculture production The climate activity will investigate the climate-related opportunities, constraints and risks to agriculture and aquaculture in the Darwin catchments under current and projected future climates. This analysis will include examination of the: • large- and small-scale processes controlling climate in the Darwin catchments • spatial and temporal variability and trends of key climate parameters across the Darwin catchments • accuracy of short-term and seasonal forecasting of rainfall and streamflow in the Darwin catchments • paleoclimates of northern Australia • sea surface temperatures, salinity and storm surges in the Darwin catchments. These analyses will be undertaken by examining historical climate data and modelled ocean data and will draw upon other programs of work being undertaken by CSIRO and the Bureau of Meteorology and expert knowledge within these two organisations. A large source of uncertainty in assessing agriculture productivity in northern Australia is the absence of information on local-scale variability in key climate variables, such as temperature, humidity and solar radiation, which are sparsely measured. Local-scale variations of these parameters can be important to plant growth. Based on an examination of available climate data, the Assessment may install temporary climate stations to try to detect the presence and relative importance of local-scale nuances in key climate parameters. 3.2 Generation of historical and future climate time series This section describes how the climate data will be generated for the four scenarios listed in Chapter 2. 3.2.1 SCENARIO A The Scenario A gridded daily rainfall, temperature, radiation and humidity (i.e. relative humidity, vapour pressure deficit) series for 1890 to 2015 (125 water years) will be used as input for all numerical modelling, unless specified otherwise. This time period was chosen to maximise the length of climate data, to best capture the range of hydroclimate variability (wet and dry sequences) over different time scales (annual, multi-year and multi-decadal).. Before this time period is confirmed, an analysis will be undertaken to confirm that there is no evidence of a significant change in rainfall that could be attributed to a changing climate – with respect to hydrological modelling, other climate parameters are of secondary importance to the magnitude and timing of rainfall. However, it is likely that stand-alone irrigated agricultural production modelling, which is less reliant on having a long time series of rainfall than hydrological modelling, will begin simulation modelling from 1965, when the reliability of other climate parameters (e.g. temperature, radiation) improved in the observed record. A cursory evaluation of the spatial and temporal extent of the rainfall information across northern Australia (Li et al., 2009) and in the Darwin catchments ( Figure 3-1) reveals that: • over the past 100 years, the Gulf of Carpentaria region has the best coverage of rainfall stations of all regions across northern Australia (see Li et al., 2009), although the Darwin catchments are not as well represented as the more southerly parts of the Gulf region • other climate variables are less well represented than rainfall across northern Australia (although typically exhibit less temporal variation and intuitively, spatial correlations are higher) (Figure 3-2) • before 1930, fewer than ten stations across northern Australia measured daily maximum temperature. Preliminary Agricultural Production Systems Simulator (APSIM) modelling (Keating et al., 2003) in the Darwin catchments will be undertaken to determine whether the model results are affected by having temporally varying sparse temperature and radiation data. This will help confirm the most appropriate timescale for assessing the viability of irrigated agriculture and dryland agriculture. From these observations, a decision was made to begin historical data assessments from 1890. A two-step method will be used: 1. Compile the 1890 to 2015 daily rainfall, maximum and minimum temperature, radiation and humidity (i.e. vapour pressure deficit) time series for northern Australia on a 0.05 × 0.05 degree grid from the SILO data drill (http://www.longpaddock.qld.gov.au/silo/; Jeffrey et al., 2001). 2. Calculate the 1890 to 2015 daily areal potential evaporation (PE) series for northern Australia, again on a 0.05 × 0.05 degree grid, using the SILO data drill temperature and vapour pressure data as input to Morton’s wet environment evaporation algorithms (Morton, 1983; Chiew and Leahy, 2003). Before further analysis and generation of future climate sequences, the robustness of the SILO climate data in the Darwin catchments will be investigated to ensure that the appropriate spatial and temporal patterns in the Darwin catchments were preserved during the SILO database interpolation processes. This will be done by: • spatially plotting trends in the variability of key climate parameters • identifying potential step changes in the mean and variability of key climate parameters • comparing the SILO data drill data with observed data at high-quality rainfall stations within, and in the vicinity of, the Darwin catchments. Previous studies by Petheram et al. (2009) identified that, in some small headwater catchments in the upper Darwin catchments, the volume of rain falling on the catchment provided by the SILO data drill is less than the volume of streamflow discharged. In these regions, nearby rain gauges will be used to modify the SILO data drill rainfall datasets to attempt to better reflect the actual rainfall. 3.2.2 SCENARIO B Climate inputs of Scenario B will be identical to those of Scenario A. For more information on this figure please contact CSIRO on enquiries@csiro.au Figure 3-1 Decadal analysis of the location and completeness of Bureau of Meteorology stations measuring daily rainfall used in the SILO database The decade labelled ‘1910’ is defined from 1 January 1910 to 31 December 1919, and so on. At a station, a decade is 100% complete if there are observations for every day in that decade. The Murrumbidgee catchment in south-eastern Australia is shown for comparison. For more information on this figure please contact CSIRO on enquiries@csiro.au Figure 3-2 Decadal analysis of the location and completeness of Bureau of Meteorology stations measuring daily maximum air temperature used in the SILO database The decade labelled ‘1910’ is defined from 1 January 1910 to 31 December 1919, and so on. At a station, a decade is 100% complete if there are observations for every day in that decade. The Murrumbidgee catchment in south-eastern Australia is shown for comparison. 3.2.3 SCENARIO C Global climate models (GCMs) are an important tool for simulating global and regional climate. The climate activity will provide several variants of 125 years of daily climate series, representative of a 2.2 °C temperature rise relative to 1990, guided by GCMs from the Intergovernmental Panel on Climate Change Fifth Assessment Report (AR5) for an RCP8.5 emissions scenario. However, GCMs provide information at a resolution that is too coarse to be used directly in catchment-scale hydrological modelling. For example, rainfall occurs too often and at too low intensity (Stephens et al., 2010). This is particularly important in tropical regions, where even high-resolution coupled climate models simulate tropical cyclones as weaker and larger in spatial extent than those observed (IPCC, 2007). Hence, an intermediate step is generally performed: the broadscale GCM outputs are transformed to catchment-scale variables. Two fundamental approaches exist for downscaling broadscale GCM output to a finer spatial resolution: dynamic downscaling and statistical downscaling (Fowler et al., 2007, and references therein). Dynamic downscaling is computationally very intensive, while sophisticated statistical downscaling methods (e.g. regression models, weather typing schemes, weather generators) are laborious. Numerous studies, however, also suggest that no single downscaling method is superior across a range of hydrological metrics (Mitchell, 2003; Whetton et al., 2005; Prudhomme and Davies, 2009; Chiew et al., 2010; Segui et al., 2010). In a comparative assessment of scaling methods in an area dominated by large-scale storm events, Salathé (2003) found simple scaling methods to be effective in simulating hydrological systems. For these reasons, and because climate change assessment is not the primary focus of the Assessment, a simple scaling technique – the pattern scaling method (Whetton et al., 2000) – will be used. The GCM selection process is outlined in the next section and is followed by a brief summary of the pattern scaling method. Selection of global climate models A commonly held premise of hydrological prediction is that models that are better able to simulate the past are more likely to accurately simulate the future. In an Australia-wide assessment of rainfall simulations using 23 GCMs, Chiew et al. (2009) found that there was no clear difference in future rainfall projections across northern Australia between the better and poorer performing GCMs, and that the use of weights to favour the better GCMs gave similar rainfall results to modelling using all 23 GCMs. The use of paleo-observations to determine which, if any, GCMs have the proper sensitivity in tropical regions is also of little value because tropical oceanic and terrestrial paleoclimate proxies are conflicting (Rind, 2008). To assess the uncertainty and simulate the range of future runoff predictions, future climate projections from a large range of archived GCM simulations will be downloaded from the Coupled Model Intercomparison Project (CMIP5) website (http://cmip-pcmdi.llnl.gov/cmip5/). There are 42 GCMs available for the RCP8.5 emissions scenario with rainfall and temperature data; however, a small subset (10) of these GCMs do not have solar radiation and/or humidity data, which are required for the AWRA-L (Chapter 5), WAVES (Chapter 6), APSIM and GRASP (Chapter 7) modelling. Only those 32 GCMs that have data available for rainfall, temperature, solar radiation and humidity will be examined. The exact number of GCMs with all of these climate parameters available over the time slices of interest is still being determined. Ultimately, three GCMs – a ‘wet’, a ‘mid’ and a ‘dry’ – will be selected for further analysis and reporting. The GCMs will be selected on the basis of ranked runoff, with the wet, mid and dry GCMs corresponding to those GCMs with the 10th, 50th and 90th percentile, respectively, mean annual runoff averaged across the Darwin catchments. GCMs will be selected on the basis of ranked runoff because this is the primary factor controlling water resource development. Scaling method The seasonal pattern scaling method to be employed in the Assessment will use the output from the 32 GCMs to scale the 125 years of historical daily rainfall, temperature, radiation and humidity sequences (i.e. SILO climate data), to construct the 32 by 125-year sequences of future daily rainfall, temperature, radiation and humidity. The method comprises two broad steps. The first step involves estimating the seasonal scaling factors for four 3-month blocks (December to February, March to May, June to August and September to November) for two time slices centred around 1990 (1975 to 2005) and 2060 (2046 to 2075). This is representative of a 2.2 °C temperature rise under an RCP8.5 emissions scenario. For each season and over each time slice, the total rainfall is calculated. Seasonal scaling factors are then calculated as the ratio of the total season’s rainfall over the 2060 time slice divided by the total rainfall over the 1990 time slice. The historical climate sequence will then be scaled using seasonal scaling factors. The second step involves rescaling the entire series so that it matches the annual scaling factors. This process will be repeated for each GCM, for each season and for each GCM grid cell. The method is then repeated for each climate parameter. The method of using a pattern scaling method to transform broadscale GCM outputs to catchment-scale variables is denoted as ‘GCM- PS’. Selection of GCMs for further analysis The 32 future climate GCM-PS time series will be used as input to the AWRA-L model (Chapter 5) to generate 32 gridded time series of future runoff across the Darwin catchments; a mean annual catchment-average runoff will be calculated for each. The GCM-PS catchment-average runoff will then be ranked from highest to lowest, and depending upon the distribution of the results the GCM-PS corresponding to the 10th (wet scenario, or Cwet), 50th (mid scenario, or Cmid) and 90th (dry scenario, or Cdry) percentile values will be selected for further analysis. Alternatively if the results are heavily skewed by several outliers, GCM-PS corresponding to the 20th (wet scenario, or Cwet), 50th (mid scenario, or Cmid) and 80th (dry scenario, or Cdry) percentile values will be selected. GCM-PS models will be selected on the basis of ranked mean annual runoff because this is the hydroclimate parameter that relates most to the potential scale of water resource development. The Cwet, Cmid and Cdry GCM scaling factors will be plotted across the Darwin catchments to check whether anomalies have been introduced as a result of GCM grid cells straddling land and water. If anomalies are identified, seasonal and annual scaling factors from adjacent land-based GCM grid cells will be adopted. 3.2.4 SCENARIO D Climate inputs for Scenario D will be identical to those of Scenario C. 3.3 References Chiew FHS and Leahy C (2003) Comparison of evapotranspiration variables in evapotranspiration maps of Australia with commonly used evapotranspiration variables. Australian Journal of Water Resources 7, 1–11. Chiew FHS, Teng J, Vaze J and Kirono DGC (2009) Influence of global climate model selection on runoff impact assessment. Journal of Hydrology 379(1–2), 172–180. Chiew FHS, Kirono DGC, Kent DM, Frost AJ, Charles SP, Timbal B, Nguyen KC and Fu G (2010) Comparison of runoff modelled using rainfall from different downscaling methods for historical and future climates. Journal of Hydrology 387(1–2), 10–23. Fowler HJ, Ekstr.m M, Blenkinsop S and Smith AP (2007) Estimating change in extreme European precipitation using a multimodel ensemble. Journal of Geophysical Research: Atmospheres 112(D18). DOI: 10.1029/2007JD008619. IPCC (2007) Climate change 2007: the physical science basis. Contributions of Working Group 1 to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, United Kingdom. Viewed 2 May 2011, http://www.ipcc.ch. Jeffrey SJ, Carter JO, Moodie KM and Beswick AR (2001) Using spatial interpolation to construct a comprehensive archive of Australian climate data. Environmental Modelling and Software 16(4), 309–330. Keating BA, Carberry PS, Hammer GL, Probert ME, Robertson MJ, Holzworth D, Huth NI, Hargreaves JNG, Meike H, Hochman Z, McLean G, Verburg K, Snow V, Dimes JP, Silburn M, Wang E, Brown S, Bristow KL, Asseng S, Chapman S, McCown RL, Freebairn DM and Smith CJ (2003) An overview of APSIM, a model designed for farming systems simulation. European Journal of Agronomy 18, 267–288. Li LT, Donohue RJ, McVicar TR, Van Niel TG, Ten J, Potter NJ, Smith IN, Kirono DGC, Bathols JM, Cai W, Marvanek SP, Gallant SN, Chiew FHS and Frost AJ (2009) Climate data and their characterisation for hydrological scenario modelling across northern Australia. A report to the Australian Government from the CSIRO Northern Australia Sustainable Yields Project. CSIRO Water for a Healthy Country Flagship, Australia. Mitchell TD (2003) Pattern scaling – an examination of the accuracy of the technique for describing future climates. Climatic Change 60(3), 217–242. Morton FI (1983) Operational estimates of lake evaporation. Journal of Hydrology 66(1–4), 77– 100. Petheram C, Rustomji P and Vleeshouwer J (2009) Rainfall-runoff modelling across northern Australia. A report to the Australian Government from the CSIRO Northern Australian Sustainable Yields Project. CSIRO Water for a Healthy Country Flagship, CSIRO, Australia. Prudhomme C and Davies H (2009) Assessing uncertainties in climate change impact analyses on the river flow regimes in the UK. Part 2: Future climate. Climatic Change 93(1–2), 197–222. Rind D (2008) The consequences of not knowing low- and high-latitude climate sensitivity. Bulletin of the American Meteorological Society 89(6), 855–864. Salathé EP (2003) Comparison of various precipitation downscaling methods for the simulation of streamflow in a rainshadow river basin. International Journal of Climatology 23(8), 887–901. Segui PQ, Ribes A, Martin E, Habets F and Boe J (2010) Comparison of three downscaling methods in simulating the impact of climate change on the hydrology of Mediterranean basins. Journal of Hydrology 383(1–2), 111–124. Stephens GL, LE T, Forbes R, Gettlemen A, Golaz J, Bodas-Salcedo A, Suzuki K, Gabriel P and Haynes J (2010) Dreary state of precipitation in global models. Journal of Geophysical Research 115. DOI: 10.1029/2010JD014532. Whetton PH, Hennessy KJ, Katzfey JJ, McGregor JL, Jones RN and Nguyen K (2000) Climate averages and variability based on a transient CO2 simulation, Annual Report 1997–98. Department of Natural Resources and Environment, Victoria. Whetton PH, McInnes KL, Jones RN, Hennessy KJ, Suppiah R, Page CM, Bathols J and Durack PJ (2005) Australian climate change projections for impact assessment and policy application: a review. Viewed 5 June 2012, http://www.cmar.csiro.au/e- print/open/whettonph_2005a.pdf. 4 Land suitability This chapter describes the methods that will be followed for the land suitability activity of the Assessment in the Darwin catchments. The land suitability focusses on areas suitable for irrigated and other agriculture, forestry and aquaculture. This chapter is in four parts. The first provides an overview of the land suitability activity and previous soil assessments undertaken in the Darwin catchments. The second part describes the methods to be used for the digital soil mapping and land suitability analysis. The third and fourth parts briefly outline data management and products produced by the land suitability activity. In many respects the methods described here build on techniques that were developed to assess land suitability for the Flinders and Gilbert Agricultural Resource Assessment, i.e. the digital soil mapping (Thomas et al., 2015) and crop suitability modelling (Harms et al., 2015). 4.1 Introduction Land development in northern Australia is challenging because of the highly variable wet and dry seasons. Also, many of the soils are poor quality because of their age and highly weathered status (Reimann et al., 2012), and so are low in fertility and affected by salinity and/or sodicity (Webb et al., 1974). These limitations combine to make the soils susceptible to water erosion in areas of even reasonably low slope (Pillans, 1997; Brooks et al., 2009), and much of the arable land is subject to lengthy periods of flooding. There are, however, deposits of younger soils (e.g. Quaternary alluvium) that may be suitable for agricultural development. To identify these areas, it is important to first understand the spatial location and characteristics of the soils, and then assess their suitability with respect to a range of land uses, principally agriculture and aquaculture. Given the size of the Darwin catchments ( Figure 4-1), the variability of the soils, and the budget and time frames of this Assessment, it is impractical to use traditional methods to map the soils. In any case, soil mapping has evolved in recent years, through a combination of computing advances and statistics, into a new approach called digital soil mapping. The key questions that this activity seeks to address in the Darwin catchments include: • Where are the soils with characteristics suitable for agricultural intensification? • Are these soils located near suitable water sources, whether groundwater or surface storage? • Which crop and land use options are suitable from a biophysical perspective? • What are the soil limitations for irrigated agriculture, and where are they located? • What are the limitations for aquaculture? • Are there potential on-site or off-site impacts that will be created through intensification, such as secondary salinisation, wind or water erosion, or acid sulfate conditions? Digital soil mapping allows soil properties (variables), such as pH, sampled at specific locations, to be related to an expanding Australian database of national covariates. Covariates, which are GIS- format datasets, are selected because they directly correlate to landscape and soil properties. Examples of covariates are slope, correlating to soil depth, and rainfall deficit correlating to leaching intensity and pH. Digital soil mapping discovers relationships at the geographic intersection of the sampled variable (e.g. pH) and multiple ‘stacked’ covariate datasets, builds statistical models from these relationships, and then applies the models to predict (map) the variable values at all other unsampled locations in the study area from the covariates (McBratney et al., 2003). Unlike traditional soil mapping used to map soil types, digital soil mapping produces maps of individual soil properties (e.g. pH or permeability). As a result, the approach is especially suited to land suitability assessment. A particular strength of digital soil mapping methods over the traditional mapping methods is that digital soil mapping produces spatial statistical measures of the quality of the mapped parameter (i.e. uncertainty estimation maps). For more information on this figure please contact CSIRO on enquiries@csiro.au Figure 4-1 Darwin catchments, showing drainage, major and minor roads and legacy soil sampling sites. Digital soil mapping has a growing successful track record in similar assessments in Australia (e.g. Kidd et al., 2012; Thomas et al., 2015) and further afield (e.g. Behrens and Scholten, 2006; Sanchez et al., 2009). 4.1.1 PREVIOUS ASSESSMENTS Soil mapping of the Darwin catchments was first published by CSIRO in 1953 (Christian and Stewart, 1953). This mapping employed, for the first time (Christian et al., 1960), the concept of land systems – mapping units that feature distinctive aggregations of repeating patterns of soils, geomorphology and vegetation (McKenzie et al., 2008), rather than soil units per se. In keeping with the current Northern Territory policy to promote economic growth, numerous contemporary soil surveys to assess land potential have been completed or are under way for the Darwin catchments area to explore opportunities for expanding farming. The area is covered by broad scale mapping (e.g. 1:100,000), while some of the area is also covered by finer scale mapping (e.g. 1:5,000 to 1:50,000). Broad scale mapping, for example, is suitable for an overview of regional opportunities for development, whereas finer scale mapping allows local targeting within the broadscale mapping for further assessments. Actual on-ground planning requires even finer scale targeted assessments of prospective areas for operational planning. There are more than 6000 legacy survey sites within the catchments. Finally, a desktop land evaluation of northern Australia (Wilson et al., 2009), which used existing soils, landscape and climate data, showed that the ability to evaluate the potential for agricultural development in most of northern Australia, including the Darwin catchments area, was severely restricted by a lack of suitable soil mapping themes (e.g. soil pH) at appropriate mapping scales. 4.1.2 ACTIVITY OBJECTIVES The three main objectives of the land suitability activity in the Darwin catchments are to: • devise a statistically robust field sampling design that most effectively captures the range of the covariates • produce maps of soil properties • produce an agricultural (or other) land suitability framework and mapping at scales that enable broad appreciation of the scale of opportunity for intensification. 4.1.3 LINKAGES TO OTHER ACTIVITIES Soil information collected as part of this activity will assist crop modelling that will be used by the agriculture viability activity (Chapter 8) to estimate parameters such as crop yield, plant water usage and deep drainage as well as providing general information on the potential of soils for agriculture and other uses. Outputs from Chapter 8 (e.g. crop yield, plant water usage, deep soil drainage per hectare) will be combined to generate water demand for the river system models (Chapter 5), assess the potential for rise in the watertable (Chapter 6) from surface water irrigation and inform the costs and benefits for the socio-economics activity (Chapter 10). The digital soil mapping data will also provide the basis for situating and designing hypothetical irrigation schemes for the case study experiments. 4.2 Soil and land assessment 4.2.1 FIELD SAMPLING DESIGN This section describes the approach to designing a statistically robust field sampling design. Maximum mapping reliability from digital soil mapping is achieved using a sampling design that has no statistical biases – that is, sampling that is representative of the whole spectrum of soils in the mapping area. Statistical approaches can select sampling sites from the data distributions in covariates datasets; success (i.e. a lack of bias) is achieved when the data distribution of covariates at the sampling sites closely matches that of the covariates over the whole area. While it is highly desirable that the resulting survey design should have no biases for the reasons above, the most important criterion is that the survey must be practically achievable to complete in the time available. Very real tensions often arise between the need for a lack of sampling bias and finite project resources, which lead to insufficient time to access all sampling sites or properly capture the full variability in the covariates, especially if some of the variability occurs in areas that are remote and difficult to access. The Assessment will apply a sampling approach to minimise these tensions. First, conditioned Latin hypercube sampling (Minasny et al., 2007; Roudier et al., 2012) of a selection of covariates (Table 4-1) will ensure that all sampling sites capture the range of soils in the Assessment area. Second, if sampling sites prove inaccessible in the field, alternative sites within a radius of 200 m will be available as a fall-back (Brungard and Johanson, 2015). To increase field efficiency, sampling sites will be restricted to within a 500 m buffer each side of roads and tracks. It is acknowledged, with sampling now restricted only to soils contained in the buffered road strips, that the ‘cost’ incurred from this action is sampling bias. The survey design will comprise 100 sampling sites. The site selection methodology will be communicated with field teams to ensure that all are aware of the design criteria and selection rationale, necessary, for example, when crews may need to select alternative sites or when dealing with other conflicts that might arise. The Assessment will use suitable legacy soil data from sites in and around the Darwin catchments. The suitability of such data will be assessed based on age and geolocation precision. During formulation, the sampling design will incorporate the legacy data to avoid sampling redundancy. Table 4-1 Description of the covariates potentially available for use in the statistical analysis to support the sampling design, digital soil mapping attributes, and soil and land use suitability mapping. Each covariate layer has a cell dimension of 30 x 30 m THEME COVARIATE CONCEPTUAL PEDOGENIC RELATIONSHIP Relief Slope percentage Hillslope accumulation and erosion patterns Profile curvature Toposequence patterns Plan curvature Hillslope water and matter accumulation Multi-resolution valley bottom flatness Broad landscape deposition zones Multi-resolution ridge top flatness Broad landscape erosion zones Slope focal median Dominating components of slope shape and slope accumulation zones Slope length 300 m Hillslope erosion and deposition intensity (fine scale) Slope length 1000 m Hillslope erosion and deposition intensity (coarse scale) Digital elevation model Elevation, relief, temperature, local rainfall Aspect Local heating intensity, local moisture deficits Contributing area Upslope contributing area for water and matter, broad landscape accumulation patterns Terrain Wetness Index (TWI –SAGA) Localised areas of water shedding and water accumulation Relief Local landform patterns, water and matter shedding and accumulation patterns Terrain roughness Erosion potential, soil depth Erosional surfaces Erosion potential, soil depth Parent material Weathering Index of Australia 2011 Index of weathering intensity of parent material 1:1 million surficial geology mapping Regional geology, geochemistry Magnetics Bedrock/parent material composition Gravity Bedrock/parent material composition Kaolinite Clay mineralogy, weathering intensity, geochemistry Smectite Clay mineralogy, geochemistry, hydraulic patterns Illite Clay mineralogy Radiometric Th High weathering intensity, geology Radiometric U Groundwater-affected areas, geology Radiometric K Fresh bedrock, geology Radiometric total count Weathering or freshness dominance Electrical conductivity Landscape/hillslope groundwater and slope process patterns, geology Total Na Landscape/hillslope slope process patterns, geology Total K Landscape/hillslope slope process patterns, geology Vegetation National Dynamic Land Cover dataset Vegetation class, nutrient cycling Normalised Difference Vegetation Index (Landsat) Photosynthetic activity Remnant Vegetation 100k Qld Vegetation class, nutrient cycling Climate Prescott Index climate surface Leaching intensity THEME COVARIATE CONCEPTUAL PEDOGENIC RELATIONSHIP Annual rainfall (mm) Soil recharge potential Minimum temperature Soil moisture and thermal dynamics Maximum temperature Soil moisture and thermal dynamics Potential evapotranspiration Leaching intensity Solar radiation Thermal dynamics Other Landsat TM Various 4.2.2 ESTIMATION OF SOIL ATTRIBUTES Land suitability assessment requires maps of the soil attributes that are important for the land uses being assessed. For the digital soil mapping approach being used (see below) these attributes should be estimated at each sampling site. Many attributes will be estimated from observations of field morphology using standard descriptions found in the Field Handbook (National Committee on Soil and Terrain, 2009). Other attributes will be estimated using mid-infrared spectroscopy (MIR) of small soil specimens taken from specified depths at all sampling sites. MIR has an advantage over conventional laboratory analysis because the MIR spectrum of a specimen can be acquired in only a few minutes and can be used to estimate a range of attributes, making it cost-effective to obtain spectra from all soil layers at all sampling sites. MIR allows a quantitative estimation of attributes such as clay silt and sand contents, pH, cation exchange capacity, exchangeable cations, organic carbon, and moisture content at 15 bar potential. For those soil attributes that can be estimated using MIR, the quantitative estimates from MIR are preferable to the categorical estimates provided from field descriptions because the resulting digital soil maps can show how the attribute changes continuously over the landscape. In addition, the estimates are less subjective, especially since statistical methods can be used to remove any poor quality estimates. Estimation of soil attributes requires development of a calibration for each attribute that uses conventional laboratory analysis (Rayment and Lyons, 2011) from a small subset of specimens. It is important that the specimens selected for calibration cover the full range of MIR spectra for the catchments. However, because of the extra field time required to take a large, bag specimen for conventional analysis, it is necessary to select in advance those sampling sites where calibration specimens will be taken. A two-stage approach will be used. The first stage involves using the same conditioned Latin hypercube method described above and selecting a random subset of 20 sites (20 %) of the 100 sampling sites that cover as wide a range of landscape positions as possible. Specimens for conventional analysis will be taken from the standard depths at these sites. In the second stage, MIR spectra will be acquired for all the specimens from the first stage and a statistical procedure used to select a subset of about 40% that best represent the range of MIR spectra. This subset will then be sent for conventional analysis, the results of which will be used to develop the calibrations. 4.2.3 DIGITAL SOIL MAPPING This section describes the approach to the production of maps of soil properties. Thomas et al. (2015) described the application of digital soil mapping to large northern Australian catchments, and this forms the basis of the method used in this Assessment. The digital soil mapping will rely on a suite of approaches combining environmental correlation and data mining. These statistical techniques include piecewise linear regression decision trees, using the Cubist toolset (Holmes et al., 1999) to predict continuous variables (e.g. pH, organic carbon), and using the C5 toolset (RuleQuest, 2006) for categorical variables (e.g. rock/non-rocky classes). Both of these toolsets are packages operated using the R statistical software program (Hornik, 2016). The attraction of these approaches, in addition to their proven predictive power, is that they generate model outputs that allow scientists to interact and query the models from a pedogenic perspective. Other machine learning approaches may be tested, including DSMART (Odgers et al., 2014) or Random Forests (Hothorn et al., 2006). All covariates in Table 4-1 will be considered in the digital soil mapping because toolsets such as Cubist exploit covariates in proportion to the level of their predictive support so that there is no ‘dilution’ of overall predictive power caused by weaker covariates. The digital soil mapping will employ the soil attribute data described in the previous section as well as the wealth of legacy data from the Darwin catchments contained in the Northern Territory Government’s soil and landscape information database. The legacy data will be assessed for suitability based on attributes including age, analytical methods and locational precision. Reliability maps will be generated to accompany the prediction maps, allowing users to objectively assess: (i) the overall reliability of the map (e.g. concordance coefficient), and (ii) the spatial distribution of prediction errors. Approximately 20% of sample sites will be withheld from model creation to allow independent accuracy testing (e.g. concordance coefficient). 4.2.4 FIELD METHODS The soil survey will involve collection of new field observations and samples. Sampling will involve using a vehicle-mounted hydraulic coring rig to sample to a maximum depth of 1.5 m (or refusal). Samples will be taken at depth intervals of 0 to 0.1, 0.2 to 0.3, 0.5 to 0.6, 0.8 to 0.9, 1.1 to 1.2 and 1.4 to 1.5 m. To secure sufficient soil bulk for chemical analyses (Rayment and Lyons, 2011), three cores will collected within a 1 m diameter, and the depth–interval samples will be combined and bagged. Rapid field testing will be done in-situ, including dispersion (sodicity) and electrical conductivity (salinity) measurements. Site and soil information will be entered into field sheets, applying Australian descriptions (National Committee on Soil and Terrain, 2009) and soil classification standards (Isbell and CSIRO, 2016). Two field trips will be made to the Darwin catchments during the 2016 dry season to collect soil and landscape data. A final field trip, for validation, is scheduled for the 2017 dry season. Table 4-2 shows the soil attributes that will be determined through quantitative and qualitative methods in the field or laboratory. The digital soil mapping approaches extrapolates these variables over the larger area. These mapped covariates will, in turn, be used in the land suitability mapping. Table 4-2 Soil attributes and methods of analysis ATTRIBUTE METHOD PURPOSE Quantitative soil attribute Soil type Field assessment Defined soil type from the Australian Soil Classification, allows communication and correlation Soil depth restriction (before 1.5 m) Measured in field at depths of 0.5, 1 and 1.5 m Defining characteristic of the soil; driver for a range of other attributes, including rooting depth Organic carbon Measured in laboratory Affects plant available water capacity, erosion and fertility Percentage of clay Measured in laboratory Affects plant available water capacity, erosion, workability and adhesiveness Clay type Measured in laboratory Affects plant available water capacity, workability and adhesiveness Percentage of fine sand Measured in laboratory Affects plant available water capacity, erosion, workability and adhesiveness Percentage of coarse sand Measured in laboratory Affects plant available water capacity, erosion, workability and adhesiveness Percentage of silt Measured in laboratory Affects plant available water capacity, erosion, workability and adhesiveness pH Measured in field and laboratory Affects plant available water capacity and nutrient toxicities Electrical conductivity Measured in field and laboratory Affects plant available water capacity and salinity Cations Measured in laboratory Affects nutrient supply Soil moisture characteristics Measured in laboratory Affect plant available water capacity Bulk density and porosity Measured in field and laboratory, and estimated by pedotransfer function Affect plant available water capacity, impermeable layers, workability and erosion risk Qualitative soil attribute Permeability Field observation Describes ability of soil to transmit water internally, influences soil wetness and rooting depth Dispersion class Measured in field Affects erosion Wetness/drainage Field observation Describes ability of soil to transmit water Field texture Field observation Influences soil physical properties Rockiness Field observation Influences crop types and management factors Surface structure Field observation Affects infiltration, erosion and workability Rooting depth Recorded in field as depth to impermeable layers or bedrock Determines plant available water capacity Plant available water capacity Pedotransfer function Measure of the amount of water available to plants 4.2.5 LAND SUITABILITY MODELLING This section describes the approach to the agricultural (or other) land suitability framework. Following the collection of new soil data, the suitability assessments for each land use will be determined using the framework for land evaluation of the Food and Agriculture Organization of the United Nations (FAO, 1976). This framework was applied in earlier Flinders and Gilbert Agricultural Resource Assessment land suitability modelling (Harms et al., 2015). Land suitability is the fitness of a given area for a land use type. It is commonly expressed as a set of discrete classes, from Class 1 (completely suited) to Class 5 (completely unsuited). The first stage in land suitability assessment is constructing rules that reflect the effect of various soil and land properties on a given land use. The effect of each attribute is ranked as follows: 1. Attribute poses no significant limitation to sustained application of the specified use. 2. Attribute poses a minor limitation to the sustained application of the specified use that will cause a minor reduction of productivity or benefits and will not raise inputs above an acceptable level. 3. Attribute poses a major limitation to the sustained application of the specified use that reduces productivity or benefits and increases required inputs to the extent that the overall advantage to be gained from the use, although still attractive, will be significantly less than from Class 1 or 2 land. 4. Attribute poses a severe limitation to the sustained application of the specified use that so reduces productivity and benefits, or increases required inputs, that this expenditure will be only marginally justified. 5. Attribute poses such a severe limitation that it precludes the sustained application of the specified use. Next, the soil and landscape attributes at each location (in this case, in each mapping pixel) are scored for suitability for a given land use, using the rules above. Finally, the overall suitability is determined, usually on the basis of the most limiting attribute using the following classes: 1. Highly suitable land with no significant limitations to sustained application of the specified use. 2. Suitable land with minor limitations to the sustained application of the specified use that will cause a minor reduction of productivity or benefits and will not raise inputs above an acceptable level. 3. Moderately suitable land with major limitations to the sustained application of the specified use that reduce productivity or benefits and increase required inputs to the extent that the overall advantage to be gained from the use, although still attractive, will be significantly less than from Class 1 or 2 land. 4. Marginally suitable land with severe limitations to the sustained application of the specified use that so reduce productivity and benefits, or increase required inputs, that this expenditure will be only marginally justified. 5. Unsuitable land with such severe limitations that they preclude the sustained application of the specified use. Land suitability assessments use the underlying assumption that the most limiting factor determines the overall suitability rating. The land suitability assessment in the Assessment will apply a set of rules to the spatial attribute data produced by the digital soil mapping (Table 4-2). For example, it may be that land is unsuitable for a particular crop if the pH of the surface soil is greater than 8.5. Land suitability will be determined for a selection of promising crops (Table 4-3) under dryland or furrow, spray, trickle, flood and rainfed irrigation systems, which appear to be promising for the Darwin catchments. The suitability of land will also be determined for engineering land uses, such as aquaculture and the building of offstream water storages (e.g. ring tanks). The land suitability activity will work closely with the agriculture and aquaculture viability activity to determine the limitations for particular crops, and the land suitability rules that will be used. Table 4-3 Selection of Darwin catchments land suitabilities, including engineering and a selection of prospective crops in broad categories CATEGORY LAND USE / CROP Engineering Offstream water storage, ring tanks Aquaculture Agriculture Tree crops (nuts, fruit) Silviculture, plantation (sandalwood) Cereals Forage grazing, hay silage, legumes Vegetables Food legumes Industrial crops: cotton, sugarcane Root crops 4.3 Data management and storage All soil and covariate data will be stored on the CSIRO Bowen Cloud. Analyses and modelling will be conducted on CSIRO virtual computers with sufficient computational power and storage. Archiving and data standards will follow Assessment and CSIRO guidelines. Once quality is checked, soil landscape data will be uploaded into the Australian Soil Resource Information System (ASRIS, 2013). Selected samples and accompanying analyses will be submitted to the National Soil Archive. New data generated by the Assessment will be shared with the Northern Territory Government and more widely. 4.4 Assessment outputs The primary outputs of this activity are sets of: (i) digital layers of individual soil and landscape properties, and (ii) digital land suitability maps for various crops under different irrigation systems, as well as support for engineering (e.g. ring tanks and aquaculture). The Darwin catchments will be mapped, and the maps will be supplied with uncertainty estimates that allow broadscale appreciation of the potential for land use intensification through agriculture (typically irrigated) or aquaculture. Users will use the uncertainty maps, for example, to indicate how reliable the mapping is in areas of their special interest so indicating the level of confidence in decisions they wish to take. Areas that show mapping with high uncertainty indicate where further soil sampling may be required to bolster modelling to increase the mapping reliability, and hence, confidence in future decisions. 4.5 References ASRIS (2013) Australian Soil Resource Information System. Viewed 6 August 2013, http://www.asris.csiro.au. Behrens T and Scholten T (2006) Digital soil mapping in Germany – a review. Journal of Plant Nutrition and Soil Science 169(3), 434–443. DOI: 10.1002/jpln.200521962. Brooks AP, Shellberg JG, Knight J and Spencer J (2009) Alluvial gully erosion: an example from the Mitchell fluvial megafan, Queensland, Australia. Earth Surface Processes and Landforms 34(14), 1951–1969. Brungard CW and Johanson J (2015) The gate’s locked! I can’t get to the exact sampling spot ... can I sample nearby? Pedometron 37(July 2015), 8–10. Christian CS and Stewart GA (1953) General report on survey of Katherine–Darwin region, 1946. CSIRO, Melbourne. Christian CS, Stewart GA and Perry RA (1960) Land research in northern Australia. Australian Geographer 7(6), 217–231. DOI: 10.1080/00049186008702349. FAO (1976) A framework for land evaluation. Food and Agriculture Organization of the United Nations, Rome. Harms B, Brough D, Philip S, Bartley R, Clifford D, Thomas M, Willis R and Gregory L (2015) Digital soil assessment for regional agricultural land evaluation. Global Food Security 5(0), 25-36. Doi: http://dx.doi.org/10.1016/j.gfs.2015.04.001. Holmes G, Hall M and Prank E (1999) Generating rule sets from model trees. Springer, Berlin Heidelberg. Hornick, K (2016) “The R FAQ”, https://CRAN.R-project.org/doc/FAQ/R-FAQ.html. Hothorn T, Hornik K and Zeileis A (2006) Unbiased recursive partitioning: a conditional inference framework. Journal of Computational and Graphical Statistics 15(3), 651–674. DOI: 10.1198/106186006x133933. Isbell RF and CSIRO (2016) The Australian soil classification. CSIRO Publishing, Melbourne. Kidd D, Webb MA, Grose CJ, Moreton RM, Malone BP, McBratney AB, Minasny B, Viscarra Rossel RA, Cotching WE, Sparrow LA and Smith R (2012) Digital soil assessment: guiding irrigation expansion in Tasmania, Australia. In: Minasny B, Malone BP and McBratney AB (eds) Digital soil assessments and beyond. Taylor and Francis Group, London, 3–8. McBratney AB, Mendonca ML and Minasny B (2003) On digital soil mapping. Geoderma 117, 3–52. McKenzie NJ, Grundy MJ, Webster R and Ringrose-Voase AJ (2008) Guidelines for surveying soil and land resources. CSIRO Publishing, Collingwood, Victoria. Minasny B, McBratney AB and Walvoort DJJ (2007) The variance quadtree algorithm: use for spatial sampling design. Computers & Geosciences 33(3), 383–392. DOI: 10.1016/j.cageo.2006.08.009. National Committee on Soil and Terrain (2009) Australian soil and land survey field handbook. CSIRO Publishing, Collingwood, Victoria. Odgers NP, Sun W, McBratney AB, Minasny B and Clifford D (2014) Disaggregating and harmonising soil map units through resampled classification trees. Geoderma 214–215, 91– 100. Pillans B (1997) Soil development at snail’s pace: evidence from a 6 Ma soil chronosequence on basalt in north Queensland, Australia. Geoderma 80(1), 117–128. Rayment GE and Lyons DJ (2011) Soil chemical methods: Australasia. CSIRO Publishing, Collingwood, Victoria. Reimann C, De Caritat P, Team GP and Team NP (2012) New soil composition data for Europe and Australia: demonstrating comparability, identifying continental-scale processes and learning lessons for global geochemical mapping. Science of the Total Environment 416, 239–252. Roudier P, Beaudette DE and Hewitt AE (2012) A conditioned Latin hypercube sampling algorithm incorporating operational constraints. In: Minasny B, Malone BP and McBratney AB (eds) Digital soil assessments and beyond. Taylor and Francis Group, London, 227–231. RuleQuest (2006) See5/C5.0. Morgan Kaufmann Publishers Inc., San Francisco. Sanchez PA, Ahamed S, Carre F, Hartemink AE, Hempel J, Huising J, Lagacherie P, McBratney AB, McKenzie NJ, Mendonca-Santos MD, Minasny B, Montanarella L, Okoth P, Palm CA, Sachs JD, Shepherd KD, Vagen TG, Vanlauwe B, Walsh MG, Winowiecki LA and Zhang GL (2009) Digital soil map of the world. Science 325(5941), 680–681. DOI: 10.1126/science.1175084. Thomas M, Clifford D, Bartley R, Philip S, Brough D, Gregory L, Willis R and Glover M (2015) Putting regional digital soil mapping into practice in tropical northern Australia. Geoderma 241–242, 145–157. Webb AA, Beeston GR and Hall TJ (1974) The soils and vegetation of part of the Mayvale land system in the Gulf of Carpentaria. Agric. Chem. Lab. Br. Tech. Rep (5). Wilson P, Ringrose-Voase A, Jacquier D, Gregory L, Webb M, Wong M, Powell B, Brough D, Hill J and Lynch B (2009) Land and soil resources in northern Australia. Northern Australia Land and Water Science Review. Northern Australia Land and Water Taskforce, Canberra. 5 Surface water hydrology The surface water hydrology activity uses a modelling framework to obtain water storage and flux estimates over various spatial and temporal scales across the Darwin catchments. The key questions that this activity seeks to address in the Darwin catchments include: • How much water has discharged from the catchments each day, month and year since 1890? • What types of water sharing arrangements may be of benefit, given surface water development in the area? • What are the opportunities to use surface water for multiple uses? • Where is most runoff generated? • How does the persistence of waterholes relate to streamflow in different river reaches? • With what degree of reliability can increasing volumes of water be extracted in different parts of the Darwin catchments, and how will streamflow be perturbed downstream? • What is the maximum flood extent, and how do flood extent and duration vary with different size events? • How do flood extent and duration change under different levels of water harvesting and large dam development? • How will projected climate change scenarios affect streamflow and water resource development in the Darwin catchments? This chapter provides an overview of the key surface water modelling frameworks to be used in the Assessment. This is followed by a brief description of the available data, and an overview of the model calibration and model experiment process. Examples of use of the model output are then provided, and surface water quality is discussed briefly. 5.1 Introduction Streamflow in the Darwin catchments is highly seasonal, reflecting contrasting wet and dry seasons. The catchments are relatively flat and feature extensive floodplains of low relief. The lower portions of the rivers are tidally affected. The Adelaide River shows a tidal influence more than 100 km (of river length) from the estuary mouth. Current irrigation development in the Assessment area is largely based on groundwater. 5.2 Model overview Three types of interdependent models will mainly be used: (i) landscape, (ii) river system, and (iii) hydrodynamic. Broadly speaking, the landscape model simulates fluxes that will be used as input to the river system model and the hydrodynamic model. Output from the river system model will be used as an upstream boundary condition for the hydrodynamic model. Landscape models Landscape models are used to estimate the hydrological response of landscapes (at the scale of interest). The most widely used and recognisable landscape model is a rainfall-runoff model, which features calibrated parameters, and typically estimates runoff at a point or grid cell from daily precipitation and potential evaporation inputs. More complex landscape models, such as the Australian Water Resources Assessment – Landscape (AWRA-L), (Viney et al., 2015) model, also have calibrated parameters, but have spatially variable parameters for soil and vegetation, for example. These models are better suited to providing estimates of other water storages and fluxes, such as soil water and deep drainage. The AWRA-L model will be used in the Assessment to simulate a wide range of landscape water fluxes, but the AWRA-L parameter of primary interest for the purposes of the Assessment is runoff. River system models The river system model aggregates runoff estimates obtained from the AWRA-L model (or conceptual rainfall-runoff model) and routs the water along a stream network. Streamflow is usually estimated at various points along the river system. These points are typically referred to as nodes, with connecting stream lines referred to as ‘links’ or reaches. Each link features various sub-models to estimate in-reach processes such as routing, irrigation diversion, losses to groundwater, losses to floodplains, anabranch flow and reservoirs. Each reach or link uses inflows from reaches upstream, climate data, configuration information and calibrated parameters to estimate states related to configured processes and estimate flow at the end of the reach. The model could be used with or without a loss function to improve goodness of fit. The river system modelling activity will use two modelling frameworks in the Fitzroy Water Resource Assessment: i) AWRA – River (AWRA-R), (Dutta et al., 2015a), and ii) Source (Welsh et al., 2012). AWRA-R can be used with the AWRA-L model or a conceptual rainfall-runoff model such as Sacramento (Burnash et al., 1973). Initially, the river system model used in the Darwin Water Resource Assessment will be AWRA – River (AWRA-R), (Dutta et al., 2015a). The AWRA-R model is very flexible, enabling it to be quickly modified, as a result of its simple reach-by-reach operation where each reach is simulated in full before the simulation of the next reach. The model is also designed to enable fast run times and can be used in conjunction with a variety of auto-calibration routines (Dutta et al., 2015b). This will enable modelling experiments to be rapidly undertaken so as to ascertain the most appropriate conceptualisation and calibration strategies. The model structure and parameters may be transferred to the Source modelling framework, depending upon the legacy model preferences of the Northern Territory Government. If model parameters are transferred to the Source modelling framework, comprehensive testing will be undertaken to ensure that the AWRA-R and Source models perform similarly. Where processes and computational differences cannot be resolved between the two model platforms, a separate calibration process for Source will ensure that final delivered parameters are appropriate for Source simulation. Hydrodynamic models Hydrodynamic models are physically based models that explicitly simulate the movement of floodwaters through waterway reaches, storage elements and hydraulic structures. A combination of one-dimensional and two-dimensional hydrodynamic modelling will be used to simulate flow along sections of the stream network that have tidal influence, and floodplain areas that are subject to widespread flooding. The hydrodynamic modelling will use a one-dimensional river flow model (MIKE11) and a two-dimensional floodplain inundation model (MIKE 21) under the modelling platform of MIKEFLOOD (DHI, 2007, 2009). 5.2.1 DATA AVAILABILITY The surface water activity will build on work previously undertaken in the Darwin catchments, namely the Northern Australia Sustainable Yields (NASY) project (CSIRO, 2009). As part of the NASY Project, runoff was generated using an ensemble of conceptual rainfall-runoff models (Petheram et al., 2009). In the Darwin Water Resource Assessment, a more complex suite of hydrological models will be used than was used in NASY because more detailed modelling is required. Furthermore, a greater length of streamflow data are now available since the NASY Project was completed in 2008. The Darwin Water Resource Assessment will have greater data requirements than the NASY Project because more detailed analysis and modelling are required to address the objectives of the Assessment, and more physically based models will be used. The primary dataset used for all surface water model calibration is stream gauge data. For the river system modelling, all available gauges in the Darwin catchments will be assessed for use; however, the AWRA-L modelling may also include nearby gauges. In the Darwin catchments, 66 separate gauge records of variable quality and duration have been identified. These will be assessed for their inclusion in the river system node–link network. Gauge location, length of record and data quality for the Darwin catchments are shown in Figure 5-1. For more information on this figure please contact CSIRO on enquiries@csiro.au Figure 5-1 Darwin catchment rivers (including South and East Alligator rivers to the east of the Darwin catchments), showing gauge location and gauge record information MGSH = maximum gauged stage height Source: Petheram et al. (2009) 5.3 Model calibration and modelling experiments 5.3.1 INPUT DATA AND DATA COLLECTION Climate data, as described in Chapter 3, will be sourced from the SILO database, subject to data quality checks. Soil information is required by all three modelling frameworks (AWRA-L, AWRA-R and MIKEFLOOD). The AWRA-L model will use interim grids of soil texture generated as part of the digital soil mapping process that is undertaken in the land suitability activity (Chapter 4). Using an appropriate pedotransfer function (e.g. Dane and Puckett, 1994), saturated hydraulic conductivity will be estimated. The AWRA-R model will use interim digital soil mapping products to help constrain groundwater losses within the model and to determine appropriate soil parameters for the irrigation module within the model. The MIKEFLOOD model will use the interim digital soil mapping products to constrain modelled infiltration. The Assessment will use the hydrologically corrected Shuttle Radar Terrain Missions (SRTM-H) digital elevation model (DEM) as the baseline elevation dataset. However, river channels are poorly resolved by the SRTM-H DEM because it was flown in February 2000, when many rivers in northern Australia had high flows. For this reason, the SRTM-H DEM will be supplemented using high-resolution elevation data. These data will be acquired from laser altimetry, which was flown at the same time as the airborne electromagnetics survey of the Darwin catchments. It may be supplemented with lidar data, which are being acquired for the Adelaide River floodplain by the Northern Territory Government. If required, additional data may be acquired using laser altimetry flown from a helicopter along key transects. These data will be spliced back into the SRTM-H DEM. These high-resolution elevation data are particularly useful in helping to parameterise channel features in the MIKEFLOOD model. In key perennial reaches of the river, it may be necessary to undertake a bathymetric survey (water storage activity – Chapter 9). The results will be then spliced back into the SRTM-H DEM before use. Roughness information is required to parameterise the MIKEFLOOD model. This will be derived from vegetation mapping data and, potentially, satellite radar data. It is likely that groundwater will influence streamflow along the major floodplains of the Darwin catchments. To account for these effects, available groundwater monitoring data will be used (if close to the streamline), from which an annual pattern of groundwater heads will be generated. The result can be used to constrain groundwater losses from the river system model along appropriate reaches. These data will be used in model calibration. The inclusion of this process can enable the river system model to better simulate the effects of groundwater-based irrigation scenarios on streamflow. A limited number of pressure sensors may be deployed in selected persistent waterholes in the Darwin catchments. Waterholes will be selected in consultation with the ecology activity (Chapter 12) and the Northern Territory Government, and by analysis of the Water Observations from Space (WOfS) dataset. The WOfS on-ground sensors will be used to try to establish ‘commence to fill’ discharge and the flow required to fill selected waterholes after each dry season. This information can be used to make the output from the river system models (typically daily time series of water fluxes) more ecologically meaningful. In consultation with the Northern Territory Government, field data may be collected to help establish the physical (minimum) limits to water extraction (i.e. minimum depth and discharge at which water could be pumped) in key reaches of the Darwin catchments. 5.3.2 MODEL CALIBRATION AWRA-L The following modelling experiments to support the calibration of the AWRA-L model will be undertaken: • The Assessment will investigate various strategies for making best use of the available streamflow data. These will include modelling experiments to determine an appropriate data quality and length threshold for use in the calibration process. • A variety of objective functions will be explored using the data from the Darwin catchments, to best simulate both low and high flows. • A single set of parameters will be determined for the AWRA-L model for each of two approaches: – parameters calibrated to gauges across all of northern Australia – parameters calibrated to gauges in the vicinity of the Darwin catchments. The parameter set that proves to have the best predictive capacity will be used to estimate runoff at all locations across the Darwin catchments at a 5 km grid. The model parameters will be evaluated on an independent subset of catchments, using various goodness-of-fit measures and compared to the result of a conceptual rainfall-runoff models. For more information on the calibration routine, including objective functions and optimisers, see Viney et al. (2015). AWRA-R Calibration of the AWRA-R model will proceed as follows: • A baseline node–link network will be established for the Darwin catchments. This is simply the physical connection of river reaches with each other and is required to enforce the calculation order for each reach (reach models are run upstream to downstream in a workflow). This step will be influenced by the availability and suitability of gauge data, and physical aspects of the river system. For example, the location and extent of floodplains may influence how many reaches are represented. • If experimental results warrant, gauge data will be filtered to remove any data with unacceptable quality codes. • Sub-models that enable various processes (e.g. overbank flow, groundwater loss) can be switched on or off. • AWRA-L runoff estimates will be aggregated to provide estimates of ungauged flow (i.e. residual inflow) for each river reach. • The observed streamflow record in the headwater catchments will be ‘patched’ with AWRA-L aggregated runoff estimates, i.e. simulated runoff will be used where there are gaps in the headwater observed time series.Calibration against observed flows will be undertaken using two approaches: – reach-by-reach calibration, which involves calibrating each reach to its respective gauge without consideration of other gauges. For calibration of any Source models, this approach will be used. – system calibration, which involves using an algorithm that calibrates all reaches simultaneously, and is typically used without a loss/gain function. Recent research has shown that system calibration has the potential to produce more realistic system states without compromising goodness of fit for simulated flows – that is, over-fitting can be avoided by the use of system calibration (Hughes et al., 2015, 2016). In both calibration methods, the differential evolution algorithm (Mullen et al., 2011) will be used for calibration. The objective function will be a combination of Nash–Sutcliffe efficiency on root transformed values and mean annual absolute bias: ....=..1+S..................,........-................,..........2........ =1S..................,........-....................................2........ =1..·..1+S..S..................,....-................,....................................365....=.............. =1·1.... .. (5.1) where a range of values of . will be tested (as per Petheram et al. (2012)), n is the number of daily observations for the gauge, and m is the number of years of observation. Total bias as a part of the objective function is not favoured because of the tendency for optimisation to select parameters that correct for subperiods of poor goodness of fit, by having another subperiod of equally poor fit of the opposite sign for the residual – that is, one poor period corrects the bias of another poorly fitting period. Such behaviours are hidden by the use of total bias. • The parameter set that has the best predictive performance assessed against subperiods of stream gauge data will be used in the final calibration model. This will be referred to as the baseline calibration model. MIKEFLOOD MIKEFLOOD will be calibrated and audited against observed water levels (there are seven stream gauges within the hydrodynamic model domain), and flood maps derived from Landsat and MODIS imagery for a number of historical flood events. Landholder survey information about flood events and heights will also be recorded, and the data will be incorporated into the calibration process, where appropriate. The calibration of the MIKEFLOOD model is an iterative process in which the model parameters are manually adjusted and the simulation results are compared with observations until a satisfactory calibration has been achieved. The calibrated model will be used to assess changes in flood dynamics under future climate and development scenarios. Figure 5-2 shows a composite map of MODIS imagery, indicating the maximum flood extent in the Darwin catchments. Also shown is the proposed domain of the MIKEFLOOD model. 5.3.3 USE OF MODEL SIMULATION RESULTS AWRA-L As mentioned above, output from the AWRA-L model will be used as input to both the AWRA-R and MIKEFLOOD models. Runoff data from the AWRA-L model will also be used by the water storage activity (Chapter 9) to estimate the preliminary yield from large and small dams. AWRA-R Depending upon the results of the modelling experiments and the Northern Territory Government preference regarding legacy model framework, either AWRA-R or Source will be used to assess the reliability with which increasing quantities of water can be extracted in different reaches of each catchment, and under different operation and management rules. The model will also be used to undertake more detailed yield assessments of large dams in the Darwin catchments, and to simulate downstream perturbations to flow, which will be used by the ecology activity (Chapter 12) to identify likely ecological change. Key simulated streamflow metrics will be related to the ‘commence to fill’ waterhole observations and key metrics of waterhole persistence (Chapter 7), so that the persistence of waterholes can be assessed under future climate and development scenarios. The river model model will be the integrating framework in case study experiments that are dominated by surface water development (Chapter 13). MIKEFLOOD The simulated flood dynamics across the floodplains will be analysed to provide a comprehensive understanding of inundation characteristics (extents, residence time, volume, depth, frequency) at the modelled resolution. The modelled inundation information can then be used to assess hydrological connectivity between the main river channels and offstream wetlands for floods of different magnitudes, with a range of recurrence intervals. It is anticipated that this information will be summarised in the form of a multidimensional lookup table, linking streamflow and tidal level to the duration, depth and timing of inundation. This method enables changes to the hydrological connectivity between the main river channels and floodplain wetlands to be assessed under future climate and development scenarios (Dutta et al., 2013) using the AWRA-R model, which has a much faster runtime than the MIKEFLOOD model (and the Source model). The results will be presented in graphical format to illustrate the changes to the hydrological connectivity between the river channels and ecologically significant offstream wetlands under different scenarios. 5.4 Surface water quality Existing data on surface water quality collected by the Australian and Northern Territory governments, universities and other organisations will be analysed to determine whether the current quality of water in the Darwin catchments poses a limitation to different forms of water resource development. If warranted and appropriate, a limited field campaign may be undertaken to sample water in areas of concern. Surface water turbidity of waterholes in the Darwin catchments will be inferred using remotely sensed data (Chapter 7). For more information on this figure please contact CSIRO on enquiries@csiro.au Figure 5-2 Flood extent based on MODIS data (green), and hydrodynamic model domain (red polygons). HD is an abbreviation of “Hydro-dynamic” 5.5 References CSIRO (2009) Water in the Gulf of Carpentaria Drainage Division. A report to the Australian Government from the CSIRO Northern Australia Sustainable Yields Project. CSIRO Water for a Healthy Country Flagship, Australia. Dane JH, Puckett W (1994) Field soil hydraulic properties based on physical and mineralogical information. In ‘Proceedings of the International Workshop on Indirect Methods for Estimating the Hydraulic Properties of Unsaturated Soils’. (Eds MTh van Genuchten et al.) pp. 389–403. (University of California: Riverside, CA). DHI (2007) MIKEFLOOD: Modelling of River Flooding: A Step-by-Step training Guide, DHI: Denmark; 12p. DHI (2009) MIKE 21 Flow Model: Scientific Documentation, DHI: Denmark; 60p. Dutta D, Karim F, Ticehurst C, Marvanek S and Petheram C (2013) Floodplain inundation mapping and modelling in the Flinders and Gilbert catchments. A technical report to the Australian Government from the CSIRO Flinders and Gilbert Agricultural Resource Assessment, part of the North Queensland Irrigated Agriculture Strategy. CSIRO Water for a Healthy Country and Sustainable Agriculture flagships, Australia. Dutta D, Kim S, Hughes J, Vaze J and Yang A (2015a) AWRA-R technical report. CSIRO Land and Water, Australia. Dutta D, Kim S, Hughes J, Yang A and Vaze J (2015b) AWRA-R version 5.0 calibration tools. CSIRO Land and Water Flagship, Australia. Guerschman JP, Sims N, Warren G, Arthur T and Colloff M (2009) Remote sensing of inundation. In: Overton IC, Colloff MJ, Doody TM, Henderson B, Cuddy SM (eds) Ecological outcomes of flow regimes in the Murray–Darling Basin. A report prepared for the National Water Commission by CSIRO Water for a Healthy Country Flagship. CSIRO, Canberra, 422. Hughes JD, Kim SSH, Yang A, Dutta D and Vaze J (2015) Whole of system calibration of river models: weighting functions and their effect on individual gauge and system performance. MODSIM2015 proceedings: 21st International Congress on Modelling and Simulation. Modelling and Simulation Society of Australia and New Zealand, Australia, 2054–2060. Hughes JD, Kim SSH, Dutta D and Vaze J (2016) Optimisation of a multiple gauge, regulated river- system model. A system approach. Hydrological Processes. DOI: 10.1002/hyp.10752. Mullen KM, Ardia D, Gil DL, Windover D and Cline J (2011) DEoptim: an R package for global optimisation by differential evolution. Journal of Statistical Software 40(6): 1–26. Petheram C, Rustomji P and Vleeshouwer J (2009). Rainfall-runoff modelling across northern Australia. A report to the Australian Government from the CSIRO Northern Australian Sustainable Yields Project. CSIRO Water for a Healthy Country Flagship, Australia. Petheram C, Rustomji, Chiew and Vleeshouwer J (2012) Rainfall-runoff modelling in northern Australia: a guide to modelling strategies in the wet–dry tropics. Journal of Hydrology Special Edition on Tropical Hydrology 462–463, 28–41. DOI: 10.1016/j.jhydrol.2011.12.046. Viney N, Vaze J, Crosbie R, Wang B, Dawes W and Frost A (2015) AWRA-L v5.0: technical description of model algorithms and inputs. CSIRO, Australia. Welsh WD, Vaze J, Dutta D, Rassam D, Rahman JM, Jolly ID, Walbrink P, Podger GM, Bethune M, Hardy JM, Teng J and Lerat J (2012) An integrated modelling framework for regulated river systems. Environmental Modelling and Software 39, 81-102 Xu H (2006) Modification of normalised difference water index (NDWI) to enhance open water features in remotely sensed imagery. International Journal of Remote Sensing 27, 3025– 3033. 6 Groundwater hydrology The purpose of the groundwater hydrology activity is to provide a comprehensive assessment of the most promising regional aquifers in the Darwin catchments (Finniss, Adelaide, Mary and Wildman river basins) in the context of identifying opportunities for, and risks associated with, groundwater resource development. In this chapter, methods are described by which groundwater resources will be assessed in the Darwin catchments. Much of the contextual hydrogeological information has already been collated by the Northern Australia Sustainable Yields Project (CSIRO, 2009). The key questions that this activity seeks to address in the Darwin catchments include: • What types of aquifers exist? • What is the scale of the available groundwater resource? • Is the groundwater quality suitable for stock, domestic and irrigation use? • Are the aquifers suitable for managed aquifer recharge? • What is the risk of irrigation-induced salinity? • What are the potential impacts on groundwater-dependent ecosystems associated with groundwater resource development? 6.1 Introduction Groundwater resources in northern Australia occur in aquifers hosted in a variety of hydrogeological settings. The term ‘aquifer’ refers to one or a combination of geological units that transmit water relatively quickly. Conversely, ‘aquitard’ refers to one or a combination of geological units that transmit water relatively slowly. The flow of water through a combination of these two units in a defined area is known as a groundwater flow system. In general, aquifer types include alluvial aquifers of unconsolidated sediments (sand, silt and gravel) associated with rivers and their floodplains, more consolidated but permeable sedimentary aquifers such as sandstone and limestone formations, and fractured and weathered basement rock aquifers of consolidated volcanic and crystalline rocks. Many of these aquifers occur across vast geographical areas where water resources are non-prescribed and groundwater infrastructure is sparse. More often than not, the installation of infrastructure occurs sporadically, in association with individual mining or agricultural developments and occasional environmental studies. This makes understanding groundwater processes, understanding the potential for groundwater resource development and identifying risks associated with development particularly challenging. 6.1.1 ACTIVITY OBJECTIVES Activity objectives will vary between river catchments areas, depending upon the existing availability of data and the current degree of conceptual understanding. In the Adelaide and Finniss river catchments, in which groundwater resources are relatively well developed, a desktop summary will be undertaken. This will summarise the current conceptualisation of groundwater flow systems, including (i) geometry and extent, (ii) properties of aquifer and aquitard units, (iii) recharge and discharge processes (including temporal dynamics). Other aspects to be covered in these reviews are the spatial variability of groundwater quality and the extent of current resource development. In the Mary River catchment, characterisation of the groundwater flow system will involve: (i) defining the geometry and extent of aquifers (using analyses of airborne geophysical survey data and exploratory drilling); (ii) estimation of aquifer hydraulic properties (using pumping tests, analyses of downhole geophysical data, and analyses of water level time series data); (iii) estimation of the spatial dynamics of groundwater flow (from analyses of water levels and composition); and (iv) identifying locations and magnitudes of recharge and discharge fluxes, including temporal dynamics (using analyses of field-based and remotely sensed data). These activities will result in the production of a conceptual model of the groundwater flow system in the Mary River Basin. In addition, potentiometric surfaces (indicating groundwater levels and flow directions) will be produced for both wet and dry season conditions. Importantly, for an area featuring few groundwater data such as the Mary catchment, the groundwater hydrology activity will identify which data types will assist in further constraining flow system conceptualisation (and therefore the regional-scale water balance), which will assist in future predictions of potential cumulative impacts. The activity will also identify which data types will assist in the monitoring of potential impacts of groundwater extraction on groundwater-dependent ecosystems and groundwater-surface water interactions. A similar range of objectives and activities will be undertaken in the Wildman River catchment, which will represent a relatively larger component of field investigation activities. Groundwater flow system characterisation will involve the use of exploratory drilling to define the extent and properties of aquifers. A range of approaches (i.e. pumping tests, downhole geophysical logs, and time series analysis of water levels) will be used to estimate aquifer hydraulic properties. The spatial dynamics of groundwater flow will be estimated from analyses of groundwater levels and composition. Locations and magnitudes of recharge and discharge fluxes, including temporal dynamics will be identified using analyses of field-based and remotely sensed data. These activities will result in the production of a conceptual model of the groundwater flow system in the Wildman catchment. If possible, a water balance will also be estimated. Potentiometric surfaces will be produced for both wet and dry season conditions, indicating groundwater levels and flow directions. Using a combination of analytical and numerical models, the extent of buffer zones around groundwater-dependent assets will be estimated. These will be used to identify locations suitable for sustainable groundwater extraction. The groundwater hydrology activity will also identify which data types will assist in (i) further constraining flow system conceptualisation and (ii) the monitoring of potential impacts of groundwater extraction on groundwater-dependent assets. In addition to basin-specific activities, other analyses will be undertaken across all four catchments. Rates and locations of groundwater recharge and discharge will be estimated using a catchment water balance model, a soil-vegetation-atmosphere model, and analyses of remotely sensing data. If possible, recharge may also be estimated using analyses of groundwater levels and composition. Where possible, point-scale estimates derived from field investigations in the Mary and Wildman catchments will be used to constrain the results of these analyses. These regional- scale approaches will be used to produce spatial distributions of recharge and discharge across all four catchments. In summary, a range of data types and tools will be used to develop a sound conceptual understanding of the hydrogeology in the Darwin catchments. Tools include both traditional tools (drilling, aquifer pump tests and water level mapping) and novel tools (geophysics, remote sensing and environmental tracers). The work plan for the Darwin catchments will be individually tailored based on the level of existing data and knowledge, and in consultation with Northern Territory Government groundwater hydrologists, universities and private consultants. Overall, the groundwater hydrology activity will comprise six overarching components: • Develop and/or refine the conceptual understanding of the hydrogeology of the Assessment area. • Quantify the scale of the available resource across the Assessment area. • Estimate groundwater recharge and discharge across the Assessment area. • Examine opportunities for managed aquifer recharge (MAR). • Assess the potential risk of irrigation-induced salinity that could result from surface water based irrigation. • Identify potential risks arising from groundwater resource development. The opportunities for managed aquifer recharge are discussed in the water storage activity chapter (Chapter 9). The work for that activity will depend on the conceptual understanding of the hydrogeology in the Darwin catchments developed by the groundwater hydrology activity, which is essential for evaluating the feasibility of MAR and the techniques that are most suited to the environment. See Chapter 9 for more information. 6.1.1 LINKAGES TO OTHER ACTIVITIES A number of datasets will be sourced from other activities. Spatially distributed precipitation data will be sourced from the climate activity (Chapter 3). Spatial distributions of soil parameters will be sourced from the land suitability activity (Chapter 4). These will be used in conjunction with pedotransfer functions to parameterise SVAT models to estimate groundwater recharge. Catchment water balance model-based estimates of deep drainage will be sourced from the surface water activity (Chapter 5). These data will provide an independent line of evidence to compare to recharge estimates derived from SVAT models. Spatially distributed data such as MODIS reflectance, EVI, NDVI and GVMI datasets will be sourced from the Earth observation activity. These will be used to identify the locations and extents of groundwater-dependent ecosystems (i.e. monsoon vine forests) and as inputs to the CMRSET algorithm for estimation of ET. Deep drainage outputs produced by Agricultural Production Systems Simulator (APSIM) models will be sourced from the agriculture and aquaculture viability activity (Chapter 8). These will be used as an input for estimating risks of irrigation-induced groundwater salinisation. 6.2 Hydrogeology of the Darwin catchments The Darwin catchments comprise five main aquifer types: (i) unconsolidated lateritic sediments, (ii) unconsolidated Cretaceous sediments, (iii) a basal aquifer consisting of gravel, sand and clay forming the boundary between Mesozoic and Proterozoic units, (iv) the Koolpinyah Dolostone aquifer, and (v) Proterozoic fractured rock basement aquifers, consisting of fractured and jointed consolidated rocks (CSIRO, 2009; Fell-Smith and Sumner, 2011). Lateritic sedimentary aquifers occur from the ground surface and vary in thickness; they have high permeability but low storage. Net recharge is low, with recharge waters infiltrating to underlying Cretaceous aquifers. Cretaceous sediment aquifers consisting of sandstone and claystone occur sporadically across the catchments and are considered to be the most suitable targets for future groundwater development. Bores screened in Cretaceous sediments typically feature yields of 2 to 5 L/s (Fell- Smith and Sumner, 2011). Basal-layer aquifers that underlie Cretaceous sediments are composed of gravel and sand, sometimes also featuring a clay matrix. Bores sited in basal-layer aquifers have produced high yields. Cretaceous sediments are often underlain by the Koolpinyah Dolostone for which transmissivity has been estimated at 80 to 9000 m2/d (from pumping tests), and airlift bore yields have been estimated from zero to 75 L/s (CSIRO, 2009; Fell-Smith and Sumner, 2011). However, yields in bores sited in the Koolpinyah Dolostone in the Wildman River catchment vary considerably, due to the karstic nature of the aquifer. In this area, high yields are typically associated with the presence of overlying Cretaceous sediments. Other Proterozoic aquifers occur both adjacent to and below the Koolpinyah Dolostone. However, these aquifers may be limited in spatial extent, with high yields limited to fractured or brecciated areas associated with geological faults (Fell-Smith and Sumner, 2011). The regional hydrogeology, particularly Cretaceous sediment aquifers, has been well characterised for the Finniss and Adelaide catchments in the west of the Darwin catchments (Jolly and Yin Foo, 1988; Richardson, 1996; Jolly et al., 2000; EHA Pty Ltd, 2007; Yin Foo et al., 2007; CSIRO, 2009; Fell- Smith and Sumner, 2011). However, the regional hydrogeology of the Mary and Wildman river basins in the east is poorly understood, other than localised investigations in the Wildman by Coffey and Partners (1984) and Australian Groundwater Consultants (1985, 1987). Currently, the Northern Territory Department of Land Resource Management (DLRM) is undertaking a hydrogeological assessment of aquifers in the Wildman catchment, but not in the Mary catchment. The focus of the Assessment’s groundwater hydrology activity in the Darwin catchments will be to: (i) summarise the hydrogeology of the catchments from existing literature and data, (ii) undertake field hydrogeological investigations that complement those of DLRM in the Wildman catchment, (iv) undertake exploratory field hydrogeological investigations in the Mary catchment, (v) estimate groundwater recharge and discharge across the Assessment area, (vi) undertake a targeted assessment of the potential risk of irrigation-induced salinity from surface water developments, (vii) examine the opportunities for managed aquifer recharge, and (viii) evaluate and discuss the potential risks associated with groundwater resource development. 6.3 Field hydrogeological investigations Field investigations in both the Wildman and Mary catchments will comprise several tasks to better characterise the groundwater flow systems in these areas, and thereby identify the opportunities for, and risks associated with, groundwater resource development. 6.3.1 WILDMAN CATCHMENT DLRM has an established field hydrogeological investigation under way in the Wildman catchment, with a focus on characterising the regional hydrogeology. Components of that investigation include: (i) mapping the spatial extent of the main aquifers through exploratory drilling, (ii) estimating aquifer hydraulic properties using downhole geophysical methods, (iii) mapping groundwater-dependent ecosystems (GDEs), (iv) estimating the scale of the flow system from water-level mapping, (v) quantifying recharge from watertable fluctuation and groundwater chloride mass balance, and (vi) quantifying spring discharge. The Assessment’s groundwater hydrology activity intends to undertake a range of complementary activities, including: (i) installing groundwater level loggers to assist groundwater level mapping; (ii) sampling for environmental tracers, to quantify the scale of the flow system, and to understand the source and magnitude of groundwater recharge; (iii) estimating groundwater use of key groundwater-dependent vegetation (i.e. wet monsoon rainforests) (see Chapter 7); and (iv) applying simple analytical groundwater models to investigate the potential impacts of pumping on groundwater levels and, in conjunction with the ecology activity, undertake a qualitative or semi- quantitative analysis of the potential impacts on the wet monsoon rainforests. Collectively, the groundwater hydrology activity will work closely with DLRM to develop a hydrogeological conceptual model of the groundwater flow systems, including developing preliminary groundwater balance estimates and discussing potential vulnerable ecological assets in the basin. Groundwater level mapping A combination of automated and manual water level data sources will be used to generate a regional groundwater level map, to indicate the direction and scale of the groundwater flow system. Depending on the amount, quality and location of data, products to be produced include: (i) up to 25 hydrographs; (ii) a select number of piezometric cross-sections at the kilometre scale; and (iii) regional-scale water level maps for both wet and dry season conditions. Groundwater level maps will be produced using spatial analysis methods, which will also provide estimates of uncertainty associated with interpolation. Environmental tracers Environmental tracers will be sampled from groundwater in the regional aquifers where the construction of existing bores has been validated and is suitable (i.e. discretely screened in one aquifer). Tracers are useful tools when coupled with traditional hydrogeological data (water level, chemistry, aquifer geometry, hydraulic properties) for developing a hydrogeological conceptual model (Cook and Böhlke, 2000), particularly in data-sparse regions. Briefly, the composition of groundwater with respect to major ions and stable isotopes (d2H and d18O) is useful for identifying water sources and recharge mechanisms for an aquifer, and the strontium isotope ratio (87Sr/86Sr) in groundwater can be used to identify the aquifer material (host rocks) through which groundwater has flowed. Additionally, a number of age-dating environmental tracers – including anthropogenic trace gases (chlorofluorocarbons and SF6), radioactive isotopes (3H, 14C and 36Cl) and noble gases (222Rn and 4He) – are useful for determining timescales associated with groundwater flow, allowing estimation of recharge rates and flow velocities. Previous uses of environmental tracers to study groundwater recharge in northern Australia include Cook et al. (1998, 2001) and Jolly et al. (2013). The groundwater hydrology activity will work closely with DLRM hydrogeologists to identify a suitable suite of environmental tracers for characterising groundwater flow processes in the Wildman catchment. Numerous methods exist to interpret environmental tracer concentrations observed in groundwater, from simple one- and two- dimensional analytical solutions (e.g. Vogel, 1967) to complex numerical groundwater flow and solute transport models (e.g. Salmon et al., 2015). In this Assessment, the hydrogeological conceptual model derived from field investigations will dictate the interpretation approach to be used, starting with a simple model and building in complexity as knowledge and process understanding are acquired. 6.3.2 MARY CATCHMENT The Mary catchment contains very little hydrogeological information (CSIRO, 2009) – considerably less than the Wildman catchment. The Assessment groundwater hydrology activity will undertake exploratory field hydrogeological investigations in the Mary catchment to improve the current understanding of Cretaceous sediment and Koolpinyah Dolostone aquifers east of the Adelaide River. Specific components of the investigations include: (i) reprocessing and reinterpretation of existing airborne electromagnetic data (AEM) (Figure 6-1) to identify potential aquifers, (ii) a targeted drilling program to characterise the vertical and lateral extent of regional aquifers in selected locations, (iii) estimation of aquifer hydraulic properties, (iv) water level monitoring and environmental tracer sampling to determine the source and magnitude of recharge, and (v) field validation of remotely sensed mapping of the main GDEs and estimation of their water usage (Chapter 7). For more information on this figure please contact CSIRO on enquiries@csiro.au Figure 6-1 Example of the spatial variability in the mean conductivity and resistivity of the subsurface at a depth of 40 to 60 m for existing airborne electromagnetic data in the Darwin catchments Airborne geophysics Existing airborne geophysical datasets, particularly AEM data (Figure 6-1), will be analysed to better understand the nature and variability of aquifer systems within the Assessment area. Current coverages include data from the TEMPEST fixed-wing time domain airborne electromagnetic system, which was collected for identifying areas of potential uranium or thorium mineralisation (Costelloe and Hutchinson, 2010), and mapping seawater intrusion in coastal aquifers (Tan et al., 2012). This system, like other electromagnetic surveying techniques, measures varying response of the subsurface to electromagnetic fields. However, from conductivity observations alone it is generally not possible to distinguish between (1) high permeability rocks and the presence of freshwater, or between (2) low-permeability rocks and the presence of saline water. The interpretation of conductivity observations therefore typically requires the use of ancillary data such as groundwater level and quality observations and lithological logs obtained from bore drilling. The processing and interpretation of existing AEM data will employ the approach shown schematically in Figure 6-2. The resulting conductivity maps and sections will be interpreted against available hydrogeological datasets to identify the spatial extent and thickness of Cretaceous sediment and Koolpinyah Dolostone aquifers. These results will subsequently be validated through targeted exploratory drilling. For more information on this figure please contact CSIRO on enquiries@csiro.au Figure 6-2 Schematic representation of acquisition of airborne electromagnetic data (in this example, a time domain electromagnetic system) and interpretation With reference to Figure 6-2, data are acquired along parallel flight lines (a), with data recorded at regular intervals along each flight line. The electromagnetic transmitter is slung around the airframe in a fixed-wing system, and the receiver is towed beneath the aircraft. The receiver measures secondary magnetic fields from the subsurface as a function of time (b). The measured response is used to determine the conductivity-depth function by transformation or inversion (d). Conductivity-depth values can be calculated for each observation, taking account of the elevation of the system above the ground, and then combined into sections to provide a representation of the two-dimensional variation of conductivity, sometimes referred to as a ‘parasection’. Conductivity-depth profiles can be combined into a three-dimensional gridded volume from which arbitrary sections, horizontal depth slices (or interval conductivity images) and isosurfaces can be derived, showing the spatial distribution of conductivity as it varies with depth. These maps can be shown as elevations (in metres above Australian Height Datum [mAHD]) or as depth intervals below the ground surface. Drilling and piezometer installation The locations and nature of the drilling program will be determined following evaluation of both AEM data and lithological data contained in the DLRM groundwater database. Discussions will be undertaken with local drilling companies in the Darwin catchments to identify appropriate drilling methods, and DLRM will install bores. A technique for drilling will be selected based on previous successful drilling in areas with a similar variety of hydrogeological settings. Previous drilling in the Darwin catchments has used two preferred options: conventional mud drilling and a more advanced technique such as sonic drilling. Exploratory drilling will be used to characterise the vertical and lateral extent of the aquifers in selected locations, with the most promising sites chosen for the installation of monitoring bores. If suitable, a limited number of production bores may also be installed and developed for aquifer pumping test purposes. Installed bores will be used to determine aquifer hydraulic properties, as well as for water level monitoring and environmental tracer sampling to characterise recharge (see ‘Environmental tracers’, above). In addition, multiple downhole geophysical logging methods (e.g. gamma ray) will be applied to all newly installed bores, which will assist in the estimation of aquifer thicknesses and hydraulic properties. Aquifer hydraulic properties A range of methods will be used to estimate aquifer hydraulic properties. If undertaken, analytical solutions (e.g. Theis, 1935; Cooper and Jacob, 1946) will be used to interpret changes in groundwater levels and/or pressures resulting from aquifer pumping tests, while slug tests will also be interpreted using conventional analytical solutions (e.g. Hvorslev, 1951). These will be implemented using either Python language scripts or specialised software such as AQTESOLV (Duffield, 2007). Where possible, harmonic analyses may also be used to provide independent estimates of aquifer hydraulic properties. Such analyses involve the comparison of aquifer water level data with Earth tides (Bredehoeft, 1967; Cutillo and Bredehoeft, 2011) or other aquifer water level data (Boldt-Leppin and Hendry, 2003). The harmonic analysis of time series data will be undertaken using Python language scripts. Where data are available, the processing of downhole geophysical log data may also be used to provide estimates of aquifer hydraulic conductivity and porosity. Validation of remotely sensed mapping of the main GDEs and estimations of their discharge Remote sensing techniques will be used to develop an initial understanding of potential groundwater-dependent assets in the Mary catchments. For more detail, see the remote sensing activity (Chapter 7). The regional-scale GDE map produced by the Earth observation activity will be validated through closer inspection with a helicopter survey, during which aerial photographs will be taken and geographical locations will be recorded. In addition, where access is available, field measurements of spring discharge will be undertaken, as well as hydrogeochemical and environmental tracer sampling, to determine sources of groundwater discharge. 6.4 Recharge and discharge estimation Understanding the groundwater recharge and discharge components of the water balance is important for determining aquifer sustainable yields – a crucial component of effective water resource management. However, groundwater recharge and discharge are difficult to measure, and exhibit considerable variability, both spatially and temporally. For these reasons, a variety of independent methods will be used to estimate groundwater recharge and discharge in the Darwin catchments. Region-scale modelling and point-scale field measurements will be used to estimate groundwater recharge and discharge at a variety of spatial scales. 6.4.1 RECHARGE Methods of region-scale recharge estimation (to be applied across the all Darwin catchments) will include: (i) a catchment water balance analysis using remotely sensed actual evapotranspiration (ETa) data (e.g. Crosbie et al., 2015), (ii) the Australian Water Resources Assessment – Landscape (AWRA-L) model (Vaze et al., 2013), (iii) soil vegetation atmosphere transfer (SVAT) modelling using the WAVES model (Zhang and Dawes, 1998; see below), and (iv) published regression relationships (e.g. Petheram et al., 2002; Crosbie et al., 2010). Regional estimates of recharge will subsequently be constrained using point-scale estimates of recharge inferred from field data in the Wildman and Mary catchments. The catchment water balance method relies on a water balance where net recharge can be estimated as the difference between rainfall and ETa. On an annual time scale (September to August), the change in soil water can be ignored in northern Australia. The non-transpired component of rainfall, also referred to as the ‘excess water’, then needs to be partitioned into recharge and surface runoff, based on supplementary information (e.g. runoff modelling, point estimates of recharge). MODerate-resolution Imaging Spectroradiometer (MODIS) data will be used to produce scaled ETa estimates on a 250 m resolution grid using an algorithm developed by Guerschman et al. (2009) (CSIRO MODIS Reflectance-based Scaling EvapoTranspiration; CMRSET) that incorporates a relationship derived from the enhanced vegetation index (EVI) and the global vegetation moisture index (GVMI). The MODIS reflectance, EVI and GVMI datasets will be sourced from the remote sensing activity; the CMRSET data will be calibrated locally to flux towers in northern Australia. Precipitation data will be sourced from the climate activity. The AWRA-L model is a regional-scale rainfall-runoff model which is used to simulate changes in soil water storage. One component of soil water storage calculations is the loss of water to the subsurface (i.e. deep drainage), which is considered an estimate of potential recharge to groundwater. The AWRA-L model is described in more detail in the surface water hydrology chapter (Chapter 5). The WAVES model (Zhang and Dawes, 1998) is a physically based model that achieves a balance in its modelling complexity between soil physics, plant physiology, energy and solute balances. WAVES has previously been used on multiple occasions to model recharge across northern Australia, including modelling of the impacts of a future climate (Crosbie et al., 2009b). The WAVES model will be parameterised using pedotransfer functions based on grids of parameters generated by the digital soil mapping undertaken in the land suitability activity (Chapter 4). Default vegetation parameters will be assigned based on available vegetation mapping. The WAVES recharge modelling will be constrained using MODIS-derived leaf area index data and point estimates of recharge derived from field data. Field estimates of recharge will be inferred using several different methods, and are most likely to be applicable at the paddock scale or to a linear integration along groundwater flow paths (Harrington et al., 2002). These methods will include the watertable fluctuation method where monitoring bores have a sufficient amount of data available. An updated assessment of recharge using the chloride mass balance method (Crosbie et al., 2009a) will use historical information on chloride in groundwater collected by the jurisdictions and the new chloride deposition grids (Davies and Crosbie, 2016). The most rigorous estimates of recharge will be determined using environmental tracers sampled as part of the hydrogeological field investigations. A clear explanation will be provided of which methods provide estimates of gross recharge (i.e. excluding discharge) or net recharge. Recharge estimates will also be discussed in terms of potential future climate and development scenarios developed by the climate activity (Chapter 3). 6.4.2 DISCHARGE In addition to the estimation of regional-scale ETa based on remotely sensed data, estimation of local-scale groundwater discharge will be undertaken. Here, ‘local scale’ refers to scales of 10-100 m, while ‘regional scale’ refers to scales of 10-100 km (Dagan, 1986). Methods for identifying local groundwater discharge include: (i) use of remotely sensed data for indicating ETa from terrestrial GDEs (Chapter 7); (ii) gauged spring flow data, and hydraulic estimation of discharge from ungauged springs; and (iii) analysis of comparisons between variations in groundwater levels and corresponding changes in stream stage throughout the dry season, to estimate the potential of baseflow to rivers. The groundwater hydrology activity will work closely with the surface water activity (Chapter 5) where complementary groundwater and surface water hydrographs are available, to evaluate and identify the best methods for inferring baseflow. Estimates of regional groundwater recharge and discharge from each of the methods will be compared, and the appropriateness of the different methods for various catchment and hydrogeological conditions will be evaluated and discussed. Discharge estimates will also be discussed in terms of potential future climate and development scenarios developed by the climate activity (Chapter 3). Estimates of tree water use Evaporation of groundwater by vegetation represents a significant proportion of total groundwater discharge in northern Australia. The primary terrestrial GDEs of interest in the Darwin catchments are tropical ‘jungle patches’. These areas are generally less than 5 ha in spatial extent and are surrounded by Eucalyptus savannah vegetation. Rates of transpiration from the jungle patches in the Finniss and Adelaide catchments have previously been estimated using sap flow sensors and eddy covariance flux towers (Cook et al., 1998). Analyses of the stable isotopes of water have also been used to partition vegetation water use into soil water and groundwater components. Recent research indicates that groundwater typically provides around 40% of total vegetation water use by the jungle patches (Boggs et al., in review). Remote sensing techniques (Chapter 7) will be used to spatially estimate evaporation from jungle patches in the Darwin catchments over time. Evaporative fluxes will be portioned based on previous studies. 6.5 Assessing the potential impacts of groundwater development The potential impacts of groundwater development on water-related ecological assets in the Wildman and Mary catchments will be determined using analytical and numerical modelling methods, in conjunction with the Earth observation (Chapter 7) and ecology (Chapter 12) activities. For all modelling, a sensitivity analysis will be undertaken to explore how sensitive the results are to variations in key input parameters. Estimates of recharge and discharge estimates will also be discussed in terms of potential future climate and development scenarios developed by the climate activity (Chapter 3). 6.5.1 ASSESSING POTENTIAL IMPACTS OF GROUNDWATER DEVELOPMENT ON GROUNDWATER LEVELS AND SPRING DISCHARGE Analytical and/or numerical models will be used in sensitivity analyses to assess the depth and rate of groundwater drawdown as a result of different levels and patterns of groundwater extraction at various locations. Where sufficient data are available, these models will also be used to estimate ‘buffer zones’. These will represent distances beyond which sustainable extraction may take place. The conceptualisation and parameterisation of these models will be based on hydrogeological data collected during the Assessment (e.g. aquifer hydraulic properties and water levels, and recharge and discharge fluxes). The robustness of buffer zone estimates will be examined through model sensitivity analyses. A suite of modelling tools will be used to estimate changes in spring discharge due to groundwater extraction. These tools range in complexity from analytical solution–based estimates of streamflow depletion (e.g. Glover and Balmer, 1954; Hunt, 1999, 2003) to the numerical modelling of changes in evaporation fluxes from wetland features resulting from watertable drawdown (e.g. Turnadge and Lamontagne, 2015). Simpler approaches will be implemented using Python language scripts. If there are sufficient data, more complex numerical groundwater flow modelling will be undertaken. 6.5.2 ASSESSING THE IMPACTS OF GROUNDWATER DEVELOPMENT ON GROUNDWATER-DEPENDENT ECOSYSTEMS To investigate how the condition of key GDEs (monsoon vine forests) in the catchments may change with groundwater drawdown, a version of the WAVES model with a modified lower boundary condition will be employed. Here, a fractional leakage rate (which can be varied daily) will be introduced at the lower boundary flux, which can be considered analogous to groundwater pumping. This modified boundary condition enables the internal saturated level to vary in response to rainfall and transpiration demand, allowing jungle patch condition to be evaluated by plotting the modelled vegetation leaf area index trace. If the Assessment is unable to find an appropriate dynamic root growth model, the extent to which jungle patches can adaptively modify their root architecture in response to falling groundwater levels will be ascertained from expert knowledge. This may then enable consecutive simulations using WAVES with varying rooting parameters, to further explore how vegetation condition may respond to different depths and rates of groundwater drawdown. 6.6 Risk of irrigation-induced salinity The introduction of irrigation developments can potentially result in increased drainage below the root zone and recharge to underlying aquifers. Enhanced recharge as a result of irrigation can lead to the formation of groundwater mounds in aquifers and, in extreme cases, waterlogging and secondary salinisation. Numerous studies have demonstrated the potential risk of watertable rise (Cook et al., 2008; Petheram et al., 2008; Paydar et al., 2011) and increased discharge to adjacent rivers (Rassam et al., 2004, 2005; Knight et al., 2005; Jolly et al., 2013) as a result of increased root- zone drainage beneath areas under irrigation. For secondary salinisation to occur, there needs to be (i) a source of water, (ii) a source of salt, and (iii) a mechanism by which the salt can be redistributed in the soil profile. The risk of irrigation-induced salinity will be assessed in the Darwin catchments in areas where irrigation developments may occur, based on opportunities identified by the coexistence of surface water availability and land suitability, as well as areas where shallow watertables exist. This will involve three tasks: (i) estimating likely ranges in deep drainage under different crops and management – on the more suitable agricultural soils in the Darwin catchments, deep drainage will be estimated using the Agricultural Production Systems Simulator (APSIM) model, which is being run as part of the agriculture and aquaculture viability activity (Chapter 8); (ii) undertaking a sensitivity analysis of the rate of watertable rise and groundwater discharge, with and without a river boundary condition, using analytical solutions developed by Jolly et al. (2013). These solutions can be used to calculate watertable rise and discharge to rivers based on the size and spatial location of one or more irrigation developments, amount of deep drainage, and hydraulic properties of the unsaturated zone and depth to groundwater; and (iii) an evaluation and discussion of the risk of irrigation-induced salinity, based on a high risk of watertable rise indicated by the sensitivity analysis and the occurrence of existing high salinity in soil (spatial soil salinity grids) from the land suitability activity and groundwater salinity data (estimated as part of this activity). In extreme cases, there may be a need for estimating total salt loads (including unsaturated zone salinity) with input data from various sources (listed below). If required, total salt loads will be estimated using datasets sourced from: (i) the surface water activity (irrigation water quality), (ii) the land suitability activity (spatial soil salinity grids), and (iii) data on spatial groundwater salinity and depth to groundwater (estimated as part of this activity). Estimates of salt in the unsaturated zone will be obtained from drilling (conventional or direct push) to collect sediment profiles from the unsaturated zone, with salt estimates generated by 1:5 electrical conductivity soil water suspensions. Potential areas where irrigation-induced salinity are likely to occur will be identified by the coincidence of areas with potential for watertable rise and areas where salt from the unsaturated zone may be mobilised as a result of irrigation. 6.7 References Australian Groundwater Consultants (1985) Wildman River Station water source investigation. Report EXTD0654. Water Resources Division, Northern Territory Department of Mines and Energy, Darwin. Australian Groundwater Consultants (1987) Wildman River Station water source investigation Stage 2 groundwater studies – final report. Water Resources Division, Northern Territory Department of Mines and Energy, Darwin. Batlle-Aguilar J, Harrington GA, Leblanc M, Welch C and Cook PG (2014) Chemistry of groundwater discharge inferred from longitudinal river sampling, Water Resources Research 50(2): 1550- 1568. Boggs D, Hutley LB, Liddle DT, Boggs GS and Cook PG (in review) Partitioning water sources in monsoon vine forest trees in the seasonal tropics of northern Australia. Ecohydrology Journal. Boldt-Leppin BE and Hendry MJ (2003) Application of harmonic analysis of water levels to determine vertical hydraulic conductivities in clay-rich aquitards. Ground Water 41(4), 514– 522. Bredehoeft JD (1967) Response of well-aquifer systems to earth tides. Journal of Geophysical Research 72(12), 3075–3087. Close PG, Wallace J, Bayliss P, Bartolo R, Burrows D, Pusey BJ, Robinson CJ, McJannet D, Karim F, Byrne G, Marvanek S, Turnadge C, Harrington G, Petheram C, Dutra LXC, Dobbs R, Pettit N, Jankowski A, Wallington T, Kroon F, Schmidt D, Buttler B, Stock M, Veld A, Speldewinde P, Cook BA, Cook B, Douglas M, Setterfield S, Kennard M, Davies P, Hughes J, Cossart R, Conolly N and Townsend S (2012) Assessment of the likely impacts of development and climate change on aquatic ecological assets in Northern Australia, report for the National Water Commission of Australia, Tropical Rivers and Coastal Knowledge project, Commonwealth Environmental Research Facility, Charles Darwin University, Darwin, 621p. Coffey and Partners (1984) Wildman River groundwater investigation: stage 1, final report. Water Division Report WRD84071. Northern Territory Department of Transport and Works, Darwin. Cook PG and Böhlke JK (2000) Determining timescales for groundwater flow and solute transport. In: Cook PG and Herczeg AL (eds) Environmental tracers in subsurface hydrology. Kluwer, London, 1–30. Cook PG, Hatton TJ, Pidsley D, Herczeg AL, Held A, O’Grady AP and Eamus D (1998) Water balance of a tropical woodland ecosystem, northern Australia: a combination of micro- meteorological, soil physical and groundwater chemical approaches. Journal of Hydrology 210, 161–177. DOI: 10.1016/S0022-1694(98)00181-4. Cook PG, Herczeg AL and McEwan KL (2001) Groundwater recharge and stream baseflow: Atherton Tablelands, Queensland. CSIRO Land and Water Technical Report 08/01. CSIRO, Australia. Cook FJ, Xevi E, Knight JH, Paydar Z and Bristow KL (2008) Analysis of biophysical processes with regard to advantages and disadvantages of irrigation mosaics. CRC for Irrigation Futures Technical Report 07/08 and CSIRO Land and Water Science Report 14/08, Canberra. Cooper HH and Jacob CE (1946) A generalized graphical method for evaluating formation constants and summarizing well-field history. Transactions of the American Geophysical Union 27(4), 526–534. Costelloe MT and Hutchinson DK (2010) Woolner Granite and Rum Jungle TEMPESTTMAEM survey: inversion report. Geoscience Australia internal report. Crosbie RS, McCallum JL and Harlington GA (2009a) Estimation of groundwater recharge and discharge across northern Australia. In: Anderssen RS, Braddock RD and Newham LTH (eds) Interfacing modelling and simulation with mathematical and computational sciences, 18th World Imacs Congress and MODSIM09 International Congress on Modelling and Simulation. Modelling and Simulation Society of Australia and New Zealand, and International Association for Mathematics and Computers in Simulation, 3053–3059. Crosbie RS, McCallum JL and Harrington GA (2009b) Diffuse groundwater recharge modelling across northern Australia. A report to the Australian Government from the CSIRO Northern Australia Sustainable Yields Project. CSIRO Water for a Healthy Country Flagship, Canberra. Crosbie R, Davies P, Harrington N and Lamontagne S (2015) Ground truthing groundwater- recharge estimates derived from remotely sensed evapotranspiration: a case in South Australia. Hydrogeology Journal 23, 335–350. DOI: 10.1007/s10040-014-1200-7. CSIRO (Commonwealth Scientific and Industrial Research Organisation) (2009) ‘Water in the Timor Sea Drainage Division’. A report to the Australian Government from the CSIRO Northern Australia Sustainable Yields Project. CSIRO Water for a Healthy Country Flagship. Cutillo PA and Bredehoeft JD (2011) Estimating aquifer properties from the water level response to earth tides. Ground Water 49(4), 600–610. Dagan G (1986) Statistical theory of groundwater flow and transport: Pore to laboratory, laboratory to formation, and formation to regional scale, Water Resources Research 22(9S), 120S-134S. Davies PJ and Crosbie RS (in prep.) Mapping the spatial distribution of chloride deposition across Australia. Duffield GM (2007) AQTESOLV for Windows version 4.5 user’s guide. HydroSOLVE, Reston, Virginia, United States. EHA (2007) Integrated hydrologic modelling of the Darwin rural area and development of an integrated water resource monitoring strategy: report for hydrogeological conceptualisation & groundwater modelling. A report to the Department of Natural Resources, Environment and the Arts and Natural Systems. EHA Pty Ltd. Fell-Smith SA and Sumner J (2011) Technical report – Koolpinyah Dolomite aquifer characteristics project. Report no. 26/2011D. Northern Territory Department of Land Resource Management, Palmerston. Gardner WP, Harrington GA, Solomon DK and Cook PG (2011) Using terrigenic 4He to identify and quantify regional groundwater discharge to streams, Water Resources Research 47(6). Glover RE and Balmer GG (1954) River depletion resulting from pumping a well near a river. Transactions of the American Geophysical Union 35(3), 468–470. Grice AC, Watson I and Stone P (2013) Mosaic Irrigation for the Northern Australian Beef Industry: An assessment of sustainability and potential: Technical Report, report prepared for the Office of Northern Australia, CSIRO, Brisbane, 271p. Guerschman JP, Van Dijk AIJM, Mattersdorf G, Beringer J, Hutley LB, Leuning R, Pipunic RC and Sherman BS (2009) Scaling of potential evapotranspiration with MODIS data reproduces flux observations and catchment water balance observations across Australia. Journal of Hydrology 369(1–2), 107–119. Harrington GA, Cook PG and Herczeg AL (2002) Spatial and temporal variability of groundwater recharge in central Australia: a tracer approach. Ground Water 40(5), 518–527. Harrington GA, Stelfox L, Gardner WP, Davies P, Doble R and Cook PG (2011) Surface water – groundwater interactions in the lower Fitzroy River, Western Australia, CSIRO Water for a Healthy Country National Research Flagship, Australia, 54p. Hunt B (1999) Unsteady stream depletion from ground water pumping. Ground Water 37(1), 98– 102. Hunt B (2003) Unsteady stream depletion when pumping from semiconfined aquifer. Journal of Hydrologic Engineering 8(1), 12–19. Hutley LB, O’Grady AP and Eamus D (2000) Evapotranspiration from Eucalypt open-forest savanna of Northern Australia, Functional Ecology 14(2): 183-194. Hutley LB , O’Grady AP and Eamus D (2001) Monsoonal influences on evapotranspiration of savanna vegetation of northern Australia, Oecologia 126(3): 434-443. Hutley LB, Leuning R, Beringer J and Cleugh HA (2005) The utility of the eddy covariance techniques as a tool in carbon accounting: Tropical savanna as a case study, Australian Journal of Botany 53(7): 663-675. Hvorslev MJ (1951) Time lag and soil permeability in ground-water observations. Bulletin 36. Waterways Experimental Station, Corps of Engineers, US Army, Vicksburg, Mississippi, 1–50. Jolly I, Taylor AR, Rassam D, Knight J, Davies P and Harrington G (2013) Surface water – groundwater connectivity. A technical report to the Australian Government from the CSIRO Flinders and Gilbert Agricultural Resource Assessment, part of the North Queensland Irrigated Agriculture Strategy. CSIRO Water for a Healthy Country and Sustainable Agriculture flagships, Australia. Jolly P and Yin Foo DA (1988) Lambells Lagoon groundwater resource investigations. Report 5/1988. Power and Water Authority, Water Directorate, Darwin. Jolly P, Sutcliffe E and Jolly I (2000) Prediction of springflows at GS 8150027 on Berry Creek for the period 1870 to 1999. Report no. WRD00023. Department of Lands, Planning and Environment, Darwin. Knight JH, Gilfedder M and Walker GR (2005) Impacts of irrigation and dryland development on groundwater discharge to rivers – a unit response approach to cumulative impacts analysis. Journal of Hydrology 303, 79–91. McNeil JD (1980) Electromagnetic terrain conductivity measurement at low induction numbers. Technical Note TN-6. Geonics Ltd, Ontario, 15. McNeil JD (1990) Use of electromagnetic methods for groundwater studies: investigations. In: Ward SH (ed), Geophysics No.5, Geotechnical and Environmental Geophysics, SEG, 191–218. NRMMC, EPHC and NHMRC (2009) Australian guidelines for water recycling: managing health and environmental risks (phase 2) – managed aquifer recharge. National Water Quality Management Strategy document 24. Natural Resource Management Ministerial Council, Environment Protection and Heritage Council, and Australian Health Ministers’ Conference, National Health and Medical Research Council, Canberra. Viewed 22 February 2016, http://www.environment.gov.au/system/files/resources/d464c044-4c3b-48fa-ab8b- 108d56e3ea20/files/water-recycling-guidelines-mar-24.pdf. Palacky GJ (1983) Tutorial: research, applications and publications in electrical and electromagnetic methods. Geophysical Prospecting 31, 861–872. Palacky GJ (1987) Resistivity characteristics of geologic targets. In: Nabighian MN (ed) Electromagnetic methods in applied geophysics. Society of Exploration Geophysics 1(3), 53– 129. Paydar Z, Cook F, Xevi E and Bristow K (2011) An overview of irrigation mosaics. Irrigation and Drainage 60, 454–463. Petheram C and Bristow KL (2008) Towards an understanding of the hydrological opportunities and constraints to irrigation in northern Australia, CSIRO Land and Water technical report no. 14/08, 111p. Petheram C, Bristow KL and Nelson PN (2008) Understanding and managing groundwater and salinity in a tropical conjunctive water use irrigation district. Agricultural Water Management 95(10), 1167–1179. Rassam D, Walker G and Knight J (2004) Applicability of the unit response equation to assess salinity impacts of irrigation development in the Mallee region. CSIRO Land and Water Technical Report No. 35/04. CSIRO, Canberra. Rassam D, Walker G and Knight J (2005) Applicability of the unit response equation to assess salinity impacts of irrigation development in the Mallee region: supplementary analyses. Report for CSIRO Water for a Healthy Country Flagship, Townsville. Richardson K (1996) Groundwater resources of La Belle Downs. Report no. WRD96022. Power and Water Authority, Darwin. Salamon SU, Prommer H, Park J, Meredith KT, Turner JV and McCallum JL (2015) A general reactive transport modelling framework for simulating and interpreting groundwater 14C age and d13C. Water Resources Research 51(1), 359–376. Smerdon BD, Gardner WP, Harrington GA and Tickell SJ (2012) Identifying the contribution of regional groundwater to the baseflow of a tropical river (Daly River, Australia), Journal of Hydrology 464: 107-115. Tan KP, Clarke JDA, Halas L, Lawrie KC, Apps HE, Brodie RC, Costelloe M, Sumner J, Fell-Smith S, Christensen NB, Schoning GL, Halas V and Lewis SJ (2012) Northern Territory coastal plain: mapping seawater intrusion in coastal plain aquifers using airborne electromagnetic data. Geoscience Australia Record, 2012/41. Geoscience Australia, Canberra. Theis CV (1935) The relation between the lowering of the piezometric surface and the rate and duration of discharge of a well using ground-water storage. Transactions of the American Geophysical Union 16(2), 519–524. Turnadge C and Lamontagne S (2015) A MODFLOW-based approach to simulating wetland– groundwater interactions in the Lower Limestone Coast Prescribed Wells Area. Goyder Institute for Water Research Technical Report Series 15/12. Vaze J, Viney N, Stenson M, Renzullo L, Dijk AV, Dutta D, Crosbie R, Lerat J, Penton D, Vleeshouwer J, Peeters L, Teng J, Kim S, Hughes J, Dawes W, Zhang Y, Leighton B, Perraud J-M, Joehnk K, Yang A, Wang B, Frost A, Elmahdi A, Smith A and Daamen C (2013) The Australian Water Resource Assessment Modelling System (AWRA). In: Piantadosi J, Anderssen RS and Boland J (eds) MODSIM 2013 Adapting to change: the multiple roles of modelling, 20th International Congress on Modelling and Simulation. Modelling and Simulation Society of Australia and New Zealand, Adelaide, 3015–3021. Vogel JC (1967) Investigations of groundwater flow with radiocarbon. Proceedings of the Symposium on Isotopes in Hydrology. IAEA, Vienna. Yin Foo D, Arunakumaren J and Evans P (2007) Hydrologic modeling of the Darwin rural area and development of a water resource monitoring strategy. CD containing presentations at IAHNT 2007 Conference Groundwater in Carbonate Rocks in the North of Australia, Darwin. Zhang L and Dawes W (1998) WAVES: an integrated energy and water balance model. CSIRO Land and Water, Canberra. 7 Earth observation Challenges of working in northern Australia include the scarcity of data and the remoteness of the landscape. For these reasons, the Assessment will use remotely sensed imagery (i.e. satellite imagery) where it can meaningfully inform the objectives of the Assessment. Many of the Assessment’s activities will make extensive use of readily available remote sensing products. For example, the land suitability activity (Chapter 4) will use existing airborne gamma radiometrics data as a spatial covariate in its digital soil mapping process. This chapter describes tasks that require detailed processing of remotely sensed data as part of the Assessment to produce products that help other activities achieve their objectives. This activity has five tasks in the Darwin catchments, each of which are described in this chapter: Mapping the relative frequency and distribution of inundation, and the distribution and persistence of waterholes 1. Mapping the suspended sediment concentration of waterholes 2. Mapping riparian vegetation and groundwater-dependent ecosystems (GDEs) 3. Increasing the spatial resolution and coverage of vegetation water use maps (actual evapotranspiration, ETa), and 4. Increasing the spatial and temporal resolution of soil moisture maps In conjunction with discipline specialists from other activities, the Earth observation activity seeks to address a range of questions in the Darwin catchments, including: • Where are persistent waterholes located? • Which areas are most prone to widespread and frequent inundation? • Which surface water bodies are replenished in part by groundwater during the dry season? • Which waterholes may be particularly sensitive to nearby development? • Where are vegetation communities likely to be most sensitive to groundwater extraction? • What is the distribution of persistently productive vegetation that may be valuable as drought refuge habitat for Wildlife? • How do evapotranspiration rates vary throughout the Assessment area? The Earth observation activity will draw on a wide range of data sources, primarily satellite remote sensing technologies, that each have unique characteristics in terms of their capture frequency, spatial resolution and coverage and application domain. These attributes are summarised for some of the primary data sources in Table 7-1. Table 7-1 Spatial, temporal and spectral characteristics of selected satellite data sources that will be used in the Earth observation activity PLATFORM/SENSOR DATE FREQUENCY NO. BANDS PIXEL SIZE SWATH WIDTH (km) APPLICATION NOTES Landsat 5 Thematic Mapper (TM) 1987-1999 2003-2011 16 days 7 30m Multi (6) 120m Thermal (1) 185 Terrestrial Earth observation. Data gap 1999-2003 Landsat 7 Enhanced Thematic Mapper Plus (ETM+) Apr 1999 16 days 8 15m Pan (1) 30m Multi (6) 60m Thermal (1) 185 Terrestrial Earth observation Scan line correction failure May 2003 Landsat 8 Operational Land Imager (OLI) Feb 2013 16 days 9 15m Pan (1) 30m Multi (8) 185 Terrestrial Earth observation Terra MODIS Dec 1999 1 day 36 250m (2) 500m (5) 1000m (29) 2,330 Terrestrial Earth observation Aqua MODIS May 2002 1 day 36 250m (2) 500m (5) 1000m (29) 2,330 Terrestrial Earth observation AMSR-2 May 2012 1 day 7 Radar 5km x 3km (1) to 62km x 35km (2) 1,450 “Advanced Microwave Scanning Radiometer 2” Global water cycle observation SMOS Nov 2009 3 days 1 radar ~43km 1,000 “Soil Moisture Ocean Salinity” Hydrology and ocean circulation L-band radar SMAP Jan 2015 3 days 2 radar 30km (Radar) 1-3km (Radiometer) 1,000 “Soil Moisture Active Passive” L-band radar (active) Radiometer (passive) ASCAT Nov 2006 2 days 1 radar 50km 2 x 500 “Advanced Scatterometer” Ocean surface wind C-band WindSat Jan 2003 1 day 5 radar 8km x 13 km 39km x 71km 350 to >1,200 Polarimetric radiometer Ocean surface wind, and water cycle Key elements of the methods that will be used to complete each remote sensing task are presented below. 7.1 Flood inundation and waterhole persistence 7.1.1 INTRODUCTION In northern Australia, the hydrological extremes have particular ecological and agricultural significance. Flooding can be catastrophic to agricultural production in terms of loss of stock, fodder and topsoil, and damage to crops and infrastructure. However, the high biodiversity found in many unregulated floodplain systems in northern Australia is thought to largely depend on ‘flood pulses’, which allow biophysical exchanges to occur between the main channel and wetlands. Also of importance is that many aquatic biota in the Assessment area survive the long dry season by using refugia provided by persistent waterholes. Remotely sensed information on flooding will be used to help calibrate and post-audit the hydrodynamic models (Chapter 5). Information on waterhole persistence can also be used, in conjunction with the river system models (Chapter 5), to develop relationships between flow and waterhole persistence. The information and models may then be used to inform the ecology activity, with respect to the likely perturbations in flooding and waterhole persistence as a result of future climate and development scenarios. 7.1.2 METHODS The key challenge for mapping surface water in northern Australia using optical images is persistent cloud cover in the wet season. Archives of time series images such as from the Landsat and MODIS satellites provide the best opportunity to observe surface water and compare inundation over time. These images will be used to produce maps that will provide fundamental information about the frequency, persistence and distribution of surface water. Inundation will be mapped at two scales. First, historical flood maps will be created using MODIS satellite imagery captured daily at 500 m resolution by the Terra (MOD09GA) and AQUA (MYD09GA) satellites. Inundation will be mapped using the Open Water Likelihood (OWL) algorithm (Guerschmann et al., 2009a), which estimates the probability and/or fraction of standing water in each MODIS pixel. Pixel values range from 0, meaning that there is no water in the pixel during the month, to 100, meaning that the pixel is completely ‘flooded’ at some time. MODIS imagery is best suited to detecting inundation over larger extents, and its high temporal frequency provides the best opportunity to examine the timing and persistence of inundation. MODIS satellite imagery will be used to generate individual flood maps for every day (where available) of selected large flood events for which cloud-free images are available. Second, Landsat TM, ETM+ and OLI satellite imagery (where available) will also be used to assess flood inundation extent. The smaller pixel size of Landsat (25 m × 25 m) is better suited to mapping smaller inundation frequency, but the images are captured only once every 16 days, and the Landsat sensor cannot penetrate cloud cover. Consequently, Landsat provides less information about the persistence and frequency of inundation than the spatially coarser MODIS images, and is less likely to capture flood peaks. In northern Australia, when the dry season is usually cloud free, Landsat is better suited to identifying persistent waterholes. Landsat data have been used by Geoscience Australia to create the Water Observations from Space (WOfS) dataset, which counts – across the entire Australian continent – the number of cloud-free observations of the land surface that show open water, and provides a number of confidence statistics for surface water occurring in each pixel. The WOfS dataset will be used to show relative inundation frequency across the three Assessment areas, including the relative persistence of waterholes. Waterhole persistence mapping will require a threshold to be applied across the WOfS images to highlight the most persistent waterholes in the landscape. In addition, Normalised Difference Water Index (NDWI; Xu, 2006) images will be created from pre-calibrated Landsat images, which will provide flood maps for selected flood events (which currently cannot be extracted from WOfS). Thresholds to separate inundated from dry pixels in the mNDWI images will be determined in comparison with false colour images, as demonstrated in Sims et al., 2014. That study found that classifying mNDWI pixels with a value greater than -0.3 produced relatively contiguous inundation maps and minimised commission and omission errors in southern Australian landscapes. Field validation of inundation maps is likely to be impractical, but comparing inundation maps produced from a range of data sources and processing methods can assist in accuracy assessment and map interpretation. 7.2 Waterhole suspended sediment 7.2.1 INTRODUCTION The turbidity of waterholes, which affects light penetration and nutrient loads, is an important property of their ecological function. Waterholes that are relatively clear are most likely to undergo ecological change as a result of development activities that increase turbidity. Knowledge of waterholes that are naturally more and less turbid may help inform region-scale planning decisions. Furthermore, temporal changes in waterhole turbidity over the course of the dry season may be an indication of groundwater discharge, which may differ in salinity and suspended material concentrations than surface runoff. Such information may be useful to help the groundwater hydrology activity (Chapter 6) prioritise which waterholes and river reaches should be sampled for assessing surface water – groundwater interactions. Remotely sensed turbidity data will also be investigated for its potential to indicate, on a regional basis, the likelihood of clogging of subsurface bedsand dams (Chapter 9). 7.2.2 METHODS Archival multi-temporal Landsat imagery from the Australian Geoscience Data Cube (AGDC) will be used to reconstruct suspended sediment time series for waterholes in the catchment. Based on the relationship between the satellite-derived reflectance and the inherent optical properties of the water column, semi-empirical algorithms can be derived for these optically complex waters to provide estimates of water quality properties. Multi-temporal data from the AGDC have been processed to a consistent surface reflectance product (Li et al., 2010), which then enables large-scale time series analysis of water quality dynamics. First, a semi-analytical inversion algorithm, known as a-LMI (Brando et al., 2012), is parameterised with bio-optical measurements to estimate satellite reflectance from the known absorption and backscattering coefficients. This model then linearly resolves the water constituent concentrations (Brando and Dekker, 2003). The concentrations resolved are chlorophyll-a, non-algal particulates and coloured dissolved organic matter. A suspended sediment product is produced by combining the constituents that scatter the light: non-algal particulates and chlorophyll-a, however the concentration of these fractions can be quantified individually. Turbidity mapping of waterholes where no in-situ data exist will require calibration and validation on nearby areas where in-situ measurements have been made. As part of the Flinders and Gilbert Agricultural Resource Assessment, in-situ measurements of turbidity were made at 19 waterholes (Waltham et al., 2013), and these could be used as a calibration dataset. After calibration, the a- LMI model is applied to the historical time series of images and plots of individual waterholes to estimate suspended sediment levels over time. These will be investigated to indicate which waterholes may be receiving groundwater discharge. 7.3 Riparian vegetation 7.3.1 INTRODUCTION Riparian vegetation performs a range of important functions in riverine and wetland ecosystems, including habitat provision, organic matter supply, shading and bank stabilisation. Understanding the extent and water use requirements of riparian vegetation is necessary for assessing the potential impacts of water development. 7.3.2 METHODS A time series of dry-season Landsat TM images will be collated from cloud-free observations between 1988 and 2011 for each of the study areas. These images will be converted to a vegetation index (NDVI) that is sensitive to plant biomass and growth vigour, and an NDWI (Xu, 2006) that shows relative differences in vegetation and terrain wetness. Previous studies have used a range of processing algorithms including canonical covariates analysis to identify areas of persistently green vegetation from time series images (Canty, 2010; Canty and Neilsen 2006). Recent studies have demonstrated the advantages of principal components analysis (PCA) to identify the variance of vegetation greenness and wetness over time (see Dronova et al., 2015; Neeti and Eastman, 2014) which include its ease of use over larger time series and its ordering of productivity stability groups between the principal component bands. The first principal component (PC1) typically highlights persistently green vegetation areas characterised by high NDVI values throughout the time series, which typically includes riparian vegetation. Areas of more variable greenness are highlighted in the higher order PCs (PCs two and above). PC’s are uncorrelated with one another, so they can be interpreted as showing areas with different patterns of plant growth vigour over time. Thresholding will be used to segregate the riparian vegetation from other land-cover classes. Two approaches – automatic thresholding and image classification – will be applied to define PC1 values distinctive to riparian vegetation. The accuracy of the results will be estimated by comparing classification outputs with the best available ground-truth data of known riparian and non-riparian vegetation throughout the Assessment area, using the kappa and Jaccard coefficients. It is anticipated that some riparian vegetation maps will be available in the Assessment area. Time series Landsat images will be linked to vegetation type; geological, hydrogeological and hydrological data; and available information on mapped vegetation and its typical ecohydrological characteristics. A multivariate classification between these parameters and the NDVI and NDWI layers will be used to classify the wetlands into a range of functional types. Vegetation will be classified on the basis of its likelihood of water dependency type: inflow dependent (surface water), aquifer dependent (groundwater resource that has economic values and may support consumptive water supply) or dependent on other sources of water stored under the surface (perched groundwater, water stored in fractured or weathered zones of bedrock formations, or soil water). 7.4 High-resolution estimates of vegetation water use 7.4.1 INTRODUCTION High–spatial resolution (i.e. 25 m) remote sensing (e.g. Landsat) and gridded meteorological data will be used in a process-based model to estimate actual evapotranspiration (ETa), which is the quantity of water actually removed from the land surface by evaporation and transpiration (Pidwirny, 2006). The output of high–spatial resolution ETa estimates will be used to address ecohydrological questions related to water use by GDEs and their sensitivity to changes in water availability associated with development activities. This task will work closely with the groundwater hydrology activity (Chapter 6). High-resolution data are required because many of the GDE features that need to be assessed are discontinuous and fragmented in the landscape (e.g. monsoon rainforest and riparian vegetation). 7.4.2 METHODS Rates of ETa will be determined during the dry season for key GDEs in areas that could potentially be affected by groundwater pumping. This is likely to be the critical time for which high-resolution estimates of vegetation water use are required. The CSIRO Remote Sensing of ETa algorithm, CRemSET (Guerschman et al., 2009b) will be calibrated and audited using an existing network of ground-based flux towers, maintained by Charles Darwin University, that measure ETa. The CRemSET algorithm combines information about vegetation vigour and environmental moisture to scale potential evapotranspiration (ETp), which is a measure of the ability of the atmosphere to remove water from the surface by evaporation and transpiration assuming no control on water supply (Pidwirny, 2006) to ETa. CRemSET was identified as the most accurate ETa estimation method in a recent Australia-wide intercomparison of eight ETa algorithms against independent catchment water balance and flux tower measurement evaluation datasets (Glenn et al., 2011). The need for a specific northern Australian calibration of CRemSET will be determined following assessment of the initial results. Recent increases in the number of flux towers in northern Australia will likely increase model accuracy by providing spatially denser calibration data in this catchment. 7.5 Soil water availability 7.5.1 INTRODUCTION Soil water levels are a critical control of plant productivity. This task explores the use of multiple satellite (multi-sensor) soil moisture products to provide consistent mapping and monitoring of soil water availability across northern Australia at a higher resolution than has previously been available. This information will be used to inform the land suitability mapping (Chapter 4). Multiple satellite soil water products are considered because they each have particular strengths and weaknesses in mapping soil water across Australia’s varying landscape. These differences make some datasets more sensitive to variations in topography or vegetation cover density than others. In combination, the datasets provide considerably more information than any single product used in isolation. In addition, a multi-sensor approach increases the spatial completeness and temporal density of the satellite data – which are the only observational source of soil moisture information, especially in the sparse data environment of remote northern Australia. 7.5.2 METHODS This task will collate, evaluate and calibrate satellite soil water products at approximately 10km spatial resolution across northern Australia. The products include those derived from the passive microwave sensors AMSR2, SMOS and SMAP, and active (radar) microwave sensors, including ASCAT and WindSat. The combination of passive and active sensors is required to provide accurate and high-resolution soil moisture estimates. The products will be evaluated against in-situ and proximal soil moisture monitoring stations – namely, the OzFlux and CosmOz networks – before recalibration of the data, so that each product reports soil moisture in a consistent set of units. A second part of this task will add value to soil moisture monitoring capability across northern Australia by assimilating the satellite products into a simple water balance model. The proposed model is itself driven by satellite-derived rainfall estimates from the GPM multisatellite precipitation analysis system (IMERG). The integration of data into the model has two benefits: it improves the spatial resolution of the models to 1 km, and it imparts the constraints of the satellite data to the modelled root-zone moisture (approximately the top 1 m of soil), thus making the resulting estimates more suitable for plant water availability assessments. 7.6 References Brando VE and Dekker AG (2003) Satellite hyperspectral remote sensing for estimating estuarine and coastal water quality. IEEE Transactions on Geoscience and Remote Sensing 41, 1378– 1387. Brando VE, Dekker AG, Park YJ and Schroeder T (2012) Adaptive semianalytical inversion of ocean color radiometry in optically complex waters. Applied Optics 51, 2808–2833. Canty MJ (2010) Image Analysis, Classification and Change Detection in Remote Sensing, With Algorithms for ENVI/IDL. Taylor & Francis, CRC Press. 441 pp. Canty MJ and Nielsen AA (2006) Visualization and unsupervised classification of changes in multispectral satellite imagery. International journal of remote sensing, 27(18), 3961-3975. Dronova I, Gong P, Wang L and Zhong L (2015) Mapping dynamic cover types in a large seasonally flooded wetland using extended principal component analysis and object-based classification. Remote Sensing of Environment, 158, 193-206. Glenn EP, Doody TM, Guerschman JP, Huete AR, King EA, McVicar TR, Van Dijk AIJM, Van Niel TG, Yebra M and Zhang YQ (2011) Actual evapotranspiration estimation by ground and remote sensing methods: the Australian experience. Hydrological Processes. 25(26), 4103–4116. Guerschman JP, Sims NC, Warren G, Arthur T and Colloff M (2009a) Remote sensing of inundation. In: Overton IC, Colloff MJ, Doody TM, Henderson B and Cuddy SM (eds) Ecological outcomes of flow regimes in the Murray–Darling Basin. A report prepared for the National Water Commission by CSIRO Water for a Healthy Country Flagship. CSIRO, Canberra, 422. Guerschman J.P., Van Dijk A.I.J.M., Mattersdorf G., Beringer J., Hutley L., Leuning R., Pipunic R. and Sherman B., (2009b) Scaling of potential evapotranspiration with MODIS data reproduces flux observations and catchment water balance observations across Australia. Journal of Hydrology, 369 (1-2), 107-119. Li F, Jupp DL, Reddy S, Lymburner L, Meueller N, Tan P and Islam A (2010) An evaluation of the use of atmospheric and BRDF correction to standardize Landsat data. IEE Journal of Selected Topics in Applied Earth observations and Remote Sensing 3, 257–270. Neeti N and Eastman JR (2014) Novel approaches in Extended Principal Component Analysis to compare spatio-temporal patterns among multiple image time series. Remote Sensing of Environment, 148, 84-96. Pidwirny M (2006) Actual and Potential Evapotranspiration. Fundamentals of Physical Geography, 2nd Edition. Viewed 20 May 2016. http://www.physicalgeography.net/fundamentals/8j.html. Sims NC, Warren G, Overton IC, Austin J, Gallant J, King D, Merrin LE, Donohue R, McVicar TR, Hodgen MJ, Penton DJ, Chen Y, Huang C, Cuddy SM (2014) RiM-FIM floodplain inundation modelling for the Edward-Wakool, Lower Murrumbidgee and Lower Darling River systems. Report prepared for the Murray–Darling Basin Authority. Water for a Healthy Country Flagship, CSIRO: Canberra. https://publications.csiro.au/rpr/pub?pid=csiro:EP143823. Waltham N, Burrows D, Butler B, Wallace J, Thomas C, James C and Brodie J (2013) Waterhole ecology in the Flinders and Gilbert catchments. A technical report to the Australian Government from the CSIRO Flinders and Gilbert Agricultural Resource Assessment, part of the North Queensland Irrigated Agriculture Strategy. CSIRO Water for a Healthy Country and Sustainable Agriculture flagships, Canberra, Australia. https://www.researchgate.net/profile/Nathan_Waltham/publication/260599789_Waterhole_ecology_in_the_Flinders_and_Gilbert_catchments/links/5405bab30cf2bba34c1d7e38.pdf. Xu H (2006) Modification of normalised difference water index (NDWI) to enhance open water features in remotely sensed imagery. International Journal of Remote Sensing 27, 3025– 3033. Part III Economic viability: What are the opportunities by which water resource development may enable regional development? 8 Agriculture and aquaculture viability 8.1 Agriculture viability 8.1.1 INTRODUCTION This chapter provides an overview of approaches used in recent agricultural assessments in northern Australia. Based on the lessons learned from that work, it then outlines the methods to be used in the agriculture viability activity. The activity requires biophysical agricultural production assessments and economic assessments to be undertaken in a fully integrated way. Agricultural production assessments will also depend heavily on information generated by other Assessment activities on soils, land suitability and water resources. Rather than being prescriptive about cropping systems for particular locations, the aim is to provide insights on the issues and opportunities associated with developing integrated cropping or crop–livestock systems, as opposed to individual crops. Key questions that the agriculture sub-activity seeks to address in the Darwin catchments include: • What are the potential yields and costs of production for a range of crops and forages, given the soils and climate constraints? • How different are the opportunities for agriculture in the peri-urban areas currently dominated by horticulture and vegetables compared with the Wildman and Adelaide river catchments? • What are the water demands for a range of crops and forages, assuming that water resources for irrigation are available in the Darwin catchments? • What are the options for cropping systems (double cropping, rotational cropping) in the Darwin catchments? • Can cropping provide yields and net returns capable of servicing the capital costs of development? • How important are risk factors such as extreme weather events, pests and diseases, and management constraints for the viability of crops? • What are the opportunities for less intensive forms of cropping and improved forage production – that is, rainfed systems supplemented by some irrigation? • How are future climate likely to affect crop and forage production? • Can crops and forages be grown sustainably in the longer term in terms of off-site impacts, given the proximity of agriculture to urban settlement and areas of high conservation value? • What are the mitigation measures necessary to manage environmental risks for irrigated agriculture development? 8.1.2 SUMMARY OF PREVIOUS AGRICULTURAL ASSESSMENTS The most recent technical assessment that is highly relevant to the current Assessment was the Flinders and Gilbert Agricultural Resource Assessment (FGARA; Petheram et al., 2013a, b). A largely bottom-up approach was employed in FGARA, whereby crop calendars and crop yields were determined for a range of field and horticultural crops. Crop yields were determined based on crop modelling and expert knowledge, drawing on information from other research activities in FGARA on soils, land suitability, water resources and climate (including climate change scenarios). FGARA included an assessment of both dryland and irrigated crop production, and determined the best time of year for planting of annual crops (Webster et al., 2013). Water demand from crops was calculated; together with information on transmission and storage losses, this helped inform scenarios of sizes of water storages required and capital costs associated with irrigation. Indicative gross margins were then developed for each of the crops. Potential farm profitability of irrigated cropping and irrigated fodder crops for beef production was assessed by taking different scenarios of gross margins (revenue minus variable costs), and combining them with overhead costs and capital costs for a range of irrigation infrastructure options. This provided a good approach for assessing the gross margin needed to achieve profitability under a range of capital infrastructure options for irrigation, rather than introducing the complexity and uncertainty associated with a large range of possibilities for individual crop gross margins. As well as on-farm profitability, the economic analysis included costs and returns at the scheme and regional scales. FGARA also undertook case studies of specific development options within each of the two catchment areas. The case studies took a farming systems approach to examine multiple cropping options and sequences, and implications for region-scale capital costs, such as processing infrastructure (e.g. cotton gin, sugar mill). It is recognised that this case study approach did not (and could not) capture all of the potential system synergies, trade-offs and commercial innovations that may occur – for example, a sugar mill providing co-generation for powering a cotton gin and other infrastructure, or use of by-products and crop residues in livestock systems. The Northern Australia Food and Fibre Supply Chain Project (Ash and Gleeson, 2014) developed a library of gross margins for a wide range of different crops in four different regions across northern Australia (Flinders–Gilbert, Katherine, Ord and Pilbara), which will be used in this Assessment. Associated with this economic analysis, a supply chain model for transport costs was developed that will be of benefit in the Assessment. A broader historical assessment of the successes and challenges of agricultural developments across northern Australia (Ash et al., 2014) will also be used as context in assessing agricultural viability in the Darwin catchments. 8.1.3 METHODS OVERVIEW Rather than commencing with the biophysical drivers (such as climate, soils and environment) from which crop, forage and animal production, and financial analyses are generated, it is proposed that the starting point be the returns required to service capital costs of development i.e. a market driven approach. This information will drive an analysis of the type of cropping systems and/or crop–livestock systems that are capable of delivering required returns, given the constraints of soils, environment, climate, and supply and reliability of irrigation resources. The approach will still require significant individual crop and fodder assessment, using modelling, industry best practice and expert knowledge. The use of expert knowledge and local industry experience will be particularly important in the Darwin catchments because of the high-value horticultural and vegetable production that already takes place. However, it is envisaged that these assessments can be better targeted than apriori assessing a wide range of crops for yield and gross margins and then discovering that many will never provide the required returns on investment. The logic behind this approach is shown in Figure 8-1. For more information on this figure please contact CSIRO on enquiries@csiro.au Figure 8-1 Proposed market driven approach to agricultural viability and economic assessment In the 2.5 years before completion of the Assessment, a small amount of opportunistic field work is planned to help validate crop models. This will depend on having in place ‘crop-ready’ appropriate land resource and irrigation infrastructure, because it will not be feasible – either in time or in financial resources – to initiate any testing of areas. 8.1.4 REQUIRED RETURNS ON INVESTMENT Examining required returns on investment will involve initially assessing the likely costs of development – that is, for land, land preparation, capital costs of irrigation storages (where appropriate), equipment, and legal and regulatory costs. From this, it will be possible to determine the gross margins (gross returns less variable costs) needed to provide an acceptable return on investment, with a deliberate effort made to factor in risks such as failed crops associated with extreme weather events, pests or diseases. These disruptions cannot be explicitly predicted so scenarios with assumptions based on crop failures in 2-3 years in ten will be used. A key driver of profitability of agricultural development in northern Australia relates to transport and supply chains. This task will therefore examine sensitivity to different supply chain options (Chapter 10), using the supply chain modelling approach developed in the Northern Australia Food and Fibre Supply Chain study (Ash and Gleeson, 2014). The approach will be iterative because the costs of development will only become clear as data are provided from other project analyses on water resource development options (i.e. groundwater and surface water storages). It is anticipated that, in addition to in-house CSIRO expertise in economics, the Assessment will draw on expertise from business consultants experienced in agricultural developments in northern Australia. 8.1.5 FARM- AND PADDOCK-SCALE ECONOMICS Recently completed projects such as FGARA and the Northern Australia Food and Fibre Supply Chain study have generated a substantial ‘library’ of crop and forage gross margins, which can be used to generate indicative gross margins that will feed into the economic analyses. A combination of development costs and gross margins will be used to determine farm-scale and scheme-scale profitability under a range of scenarios of price, yield, supply chain costs, and costs of development; this will incorporate risk factors such as extreme weather events and outbreaks of pests or diseases (Table 8-1).Yield variability will be generated from crop analyses and price variability will be based on historical movements and ABARES mid-term projections, realised as part of their quarterly Agricultural Commodity update (ABARES, 2016). Table 8-1 Summary of the economic analytical approach proposed for different scales (paddock, farm, and irrigation scheme) ANALYSIS WHY INFORMATION NEEDS Enterprise gross margins (paddock- scale) analysis Asks if there are viable enterprises to accompany irrigation development List of prospective crops, prices and input costs tailored to geographic context (including transport costs) Modification of gross margin budgets Farm-scale profitability analysis Asks if an on-farm irrigation development is viable after considering all fixed and variable costs of the investment. Variations: What is the most that could be invested in infrastructure based on known gross margins? What is the minimum gross margin that would be needed to break even on a known investment cost? Costs of development (e.g. irrigation and storage infrastructure) Gross margins Review estimates from other projects Scheme-scale profitability analysis* Asks if the costs of providing scheme-scale infrastructure can be exceeded by returns from a regional development Specification of scheme-scale development – e.g. costs of infrastructure, what area is suitable * Scheme-scale economic analysis is described in the socio-economics chapter (Chapter 10). 8.1.6 CROP AND FORAGE ANALYSES Literature review A literature review will be the first step in selecting possibilities for crops and cropping systems, because interest in the agricultural potential of the Assessment area has a long history. Two processes will be used: • Past publications and personal experiences will be sought and reviewed, covering cropping and livestock experience (either actual (trial, commercial) or desktop). This will include the full hierarchy from single-enterprise biophysical studies to bigger-picture agro-economic studies. • Available information will be interpreted and analysed – for example, on optimal climate–crop combinations, soil suitability, available market windows (especially counter-seasonal supply options), infrastructure constraints and opportunities, and human capital issues. The literature review will capture existing agricultural research work in northern Australia to inform the assessment of productivity and water use of irrigated crops and forages. Crop and forage modelling and analysis The cropping systems analysis will be dependent on having estimates of crop, forage and livestock production for individual components of the system. The range of potential crops and forages will be beyond the scope of highly parameterised simulation models such as the Agricultural Production Systems Simulator (APSIM), so a pluralistic approach will be adopted, using simulation models, industry data, best practice and expert knowledge. For the Darwin catchments, a range of crops and forages are potentially suitable, including broadacre crops (e.g. rice, maize, sorghum, pulses), horticultural crops (e.g. mangoes, bananas, melons, Asian vegetables), root crops (e.g. peanuts, cassava) and forages (e.g. sorghum, lablab). Crop calendars will be developed for each of the crops and forage crops assessed. It should be noted that crop modelling will be undertaken with existing varieties. As regions are developed it will become important to develop varieties adapted to the local environment. This will take many years and is outside the scope of this Assessment. APSIM APSIM is a modelling framework that has been developed to simulate biophysical process in farming systems – in particular, where there is interest in the economic and ecological outcomes of management practice in the face of climate risk (Holzworth et al., 2015). The APSIM modelling framework has been used for a broad range of applications, including on-farm decision making, seasonal climate forecasting, risk assessment for government policy making (Keating et al., 2003), and assessment of the impact of changes in cropping systems and agronomic practices on the water balance of dryland regions (Verburg et al., 2003). It has demonstrated utility in predicting performance of commercial crops, provided that soil properties are well characterised (Carberry et al., 2009). Validated crop models have been used in previous assessments of cropping potential for a range of prospective crops in northern Australia (Carberry et al., 1991; Yeates, 2001; Pearson and Langridge, 2008; Webster et al., 2013; Ash and Gleeson, 2014). The APSIM simulation framework comprises several modules: biological modules that simulate biophysical farming processes, management modules that define ‘management rules’ that characterise a simulated scenario(e.g.crop, sowing date,application ofirrigationand nitrogenfertiliser),environmental modulesthat commandthe rate at which biological modules operate(maximum and minimum temperature, rainfall,solar radiation),program management modules tofacilitatedata flow within the simulations and produce outputs (e.g.grain yield,biomass, irrigationrequirement, wateruseefficiency, nitrogen use efficiency),and a central ‘simulation engine’ thatdrives the processes andpasses messages betweenindependent modules.The soil remainscentral to thefarming system,with modules specifying individual cropsor management actions‘plugged in’ or ‘pulled out’, depending on a specific simulation’s requirements. The key inputsrequired by APSIM are long-term daily climate records, characterised soilsdescribingplantavailablewater capacity,and agronomicpracticefor managing irrigation and crop agronomy. Crops currently available in APSIMand importanttothe Assessment areshownin Table8-2. Althoughthefocusof this work will beon irrigated crop andforage production, it will alsobeimportant to examineopportunistic dryland farming options,aswellas dryland farmingsupplemented by small amounts of irrigation. Table8-2Land use category and crop/forage modules currently availablein APSIM that can beusedin thisAssessment Crops/forageswithanasterisk have been validated for thetropics of Australia. LAND USECROP Cereal cropMaize* Rice* Sorghum (grain)* Food legumeMungbeanSoybean Forage grass, hay, silageBambatsiRhodes grassMaize Millet (forage) Sorghum (forage) Forage legumesLablab Cowpea LucerneIndustrialcropCottonSugarcaneRoot cropPeanuts Model characterisation for some of the important crops andforages for northern Australia identified in FGARAcould be improved during thisAssessment. Cropand forages that cannot be modelled Anumber of irrigated agricultural systems may be of interest inthe catchments that APSIM doesnot have the capacity tosimulate–for example,some cucurbits, tree crops, vegetables and somefodder or pastures. For these crops,expert and local experiencefrom northern Australia will be Agriculture and aquaculture viability|77 used to develop an assessment of production potential and water use. For some of these cropping systems, simple day degree models exist that have been designed to estimate harvest date and potential yield. These simple models will be compared with available data in each catchment to determine their utility. Where existing cropping systems operate in northern Australia – for example, mango, melon and Asian vegetable industries – production and water use data will be collected from the existing farming systems to inform the assessment. Climate change and crop, forage and livestock production It is possible to incorporate climate change projections (Chapter 3) in APSIM and NABSA to assess climate change impacts and potential effectiveness of adaptation options. This will be undertaken for a few key crops and forages to illustrate the production risks and opportunities associated with climate change. Given the projections are well into the future, economic analysis of climate change will not be assessed in this Assessment. Representing uncertainty Economic outcomes will depend on a range of variables; climate uncertainty now and into the future, yields, pests and diseases, debt, prices, variable costs such as transport and market access. In the scope of the Assessment it is impossible to adequately represent all of these variables and their interactions. However, uncertainty will be examined in a few key variables such as; climate and its impact on yields as represented in probabilistic outcomes from the APSIM model, price and transport variability and its effect on gross margins, and capital costs. Dependencies on other activities in the Assessment The crop, forage and livestock assessments, especially for simulation modelling, will depend on data and outputs from the climate, land suitability and water storage activities. Initial estimates of production will be based on generic soils selected to match as closely as possible existing soils information in the target regions, but these will be adjusted as new information becomes available. The modelling work should also provide estimates of water used for different crops and forages. Ultimately, this will link to water requirements for cropping systems, with feedback to water resource requirements, availability and reliability. As stated through this chapter, the agricultural viability activity will link closely to the work on economics (Chapter 10) to integrate methods and results. Losses of water from irrigation land may impacts on water quality in streams and aquifers. It will be important to link with the ecology activity (Chapter 12) and the groundwater activity (Chapter 6) to assess possible off-site impacts. Irrigated agriculture may also have an effect on Indigenous water values (Chapter 11), which will need to be considered. Fieldwork Limited field work will be undertaken to assist in validating crop and forage models, and estimates of crop and forage production. The field work will be largely opportunistic and tailored to the catchment, drawing on existing areas where it will be possible to use irrigation. There is already an active cropping sector in the Darwin catchments, particularly in horticulture (mangoes) and Asian vegetables. The Northern Territory Government has very good research and industry data on horticultural crops, and the Northern Territory Department of Primary Industry and Fisheries (DPIF) will be closely worked with to draw in data from its field work. Less is known about Asian vegetables, and the Northern Territory Farmers Association will be worked with - it has experience in working with Asian vegetable growers. DPIF is also working on broadacre crops (e.g. rice), and this activity will contract DPIF to assist with data collection. 8.1.7 CROPPING SYSTEMS AND CROP–FORAGE SYSTEMS This task will focus on analysing the type of cropping systems and crop–forage systems that are capable of delivering required returns, given the constraints of soils, environment, climate, and supply and reliability of water for irrigation. The number of different possibilities for crops and crop–forage systems will be very large, given the range of field crops, horticultural crops, forages, and crop and crop–forage rotations that can be grown in northern Australia. The first stage of this task will be to develop an approach to reducing the range of possibilities into a discrete set of agricultural system typologies that can be integrated with the economic analysis outlined above. Part of this analysis will be the scale needed to justify investment in processing facilities for industrial crops such as cotton or sugarcane. As information on returns required becomes available, testing of the different cropping system typologies will commence. The systems analysis will include a risk assessment of off-site impacts, and in particular, losses of nutrients, sediments and pesticides as a result of intensive crop–forage systems. This assessment will be based on case studies for major crops examining the risks in relation to the interacting factors of climate, soil, landscape location, irrigation type and management practices. 8.2 Aquaculture viability The purpose of the aquaculture viability sub-activity is to assess the viability of irrigated and near- shore land-based aquaculture in the Darwin catchments, and examine the opportunities and circumstances under which aquaculture enterprises could complement irrigated cropping and grazing enterprises. There will be a number of species that will be of interest in the catchment. To differentiate from agriculture production, the term “Pond crop” will be used to describe aquaculture production. This section is in three parts. The first part provides a brief background to aquaculture in northern Australia and summarises recent relevant studies. The second part describes the general method for assessment of the suitability of the Darwin catchments for different types of aquaculture enterprises. The third part describes tasks that will be undertaken to assess the viability of aquaculture in the Darwin catchments. The key questions that the aquaculture sub-activity seeks to address in the Darwin catchments include: • What are the opportunities for intensive and extensive aquaculture farming? • What are the key environmental and regulatory obstacles that will affect aquaculture development? • What are the potential yields and costs of production for a range of pond crops, given the water, soils and climate? • Can candidate species provide yields and net returns capable of servicing the capital costs of development? • Under what circumstances could irrigated crop waste streams complement aquaculture enterprises? • What are the opportunities to provide nutrient-rich wastewater from freshwater aquaculture to a range of agricultural crops? • How important are risk factors such as extreme weather events on the viability of pond crops? • How can environmental risks of aquaculture development be mitigated? • How is future climate likely to affect pond crop production? • What are the opportunities for Indigenous community engagement? 8.2.1 INTRODUCTION In the past 20 years, aquaculture has been the fastest growing food sector in the world. During this period, Australian aquaculture production has increased by an average of 11% annually (Skirtin et al., 2013) to a production value of $994 million in 2013–14. Culture of temperate water salmonids accounts for 55% of the total value of Australian aquaculture production. The aquaculture production value for Western Australia, Queensland and the Northern Territory combined accounts for less than 25% of the Australian industry (Savage and Hobshawn, 2015). There is significant untapped potential for the economic and sustainable expansion of aquaculture in northern Australia. Preliminary broadscale analysis by Mcleod et al. (2002) indicated that 1.2 million hectares of Australia’s northern coastline is potentially suitable for pond-based aquaculture. To date, significant aquaculture development in northern Australia has in part been constrained by complex legislation and the absence of aquaculture-specific policy. The most recent relevant assessment is the Aquaculture futures for coastal northern Australia final report (Preston et al., 2015). The study concluded that there is enormous potential for the development of economically, socially and environmentally sustainable aquaculture enterprises in coastal areas of northern Australia. The focus for the study was large-scale intensive farming of high-value marine species. In Preston et al. (2015), GIS methods developed by Mcleod et al. (2002) were applied to identify optimal biogeographic locations for land-based finfish and crustacean aquaculture in coastal northern Australia. The first step was a coarse preselection using low- resolution data, which are typically cheap, extensive and easily available. The second step involved fine-scale analysis with high-resolution data. A set of refined physical parameters defined by Mcleod et al. (2002) was used to select potential sites. For every land parcel, each parameter was assigned an output (score), and the overall rating of the block was a function of the sum of the parameter outputs.Preliminary site selection focusedon distanceto coast and general landtopography, ignoring soil type and composition. Preston et al. (2015) also undertook a case studywithin the Archer River catchment in northQueensland. The catchment was located in a region identified bybroadscale analysisas suitablefor land-based aquaculture.In close consultationwith theTraditional Owners, sites wereinspected byaerialand on-ground surveysto investigate significant constraints,includingseasonalflooding, infrastructure and transport logistics. The study estimated economic returnsfor thetwospecies that offer thebest potential for return on investment:black tigerprawn (Penaeusmonodon) andbarramundi (Latescalcarifer). 8.2.2AQUACULTUREPONDSUITABILITY ANALYSIS Pond-based aquaculture development in northern Australiahasthe potential to besuitable forfinfish and crustaceans of marine and freshwaterorigin. For theAssessment, a rule set of limitations will be developed to enable siteselection for both marineand freshwaterspecies, using some of the fine-scaledata that will be produced bythe Assessment (e.g.fine-scalesoil data produced as part of the land suitability activity–Chapter4). For the purpose of this analysis,each pond type willbetreated as a crop, irrespectiveof the speciesto be cultured in the pond. In theAssessment,fourpond types will be assigned a set of limitations to identify areas suitable formarineand freshwater aquaculture in earthenor plastic-linedponds (Table8-3). Table8-3Pond types POND TYPEDESCRIPTION EarthenmarineEarth-based pond suitable for marine finfish and crustaceans LinedmarinePlastic-lined pond suitable for marine finfish and crustaceansEarthenfreshwaterEarth-based pond suitable for freshwater finfish and crustaceans LinedfreshwaterPlastic-lined pond suitable for freshwater finfish and crustaceans A series of ruleswill bedeveloped for each pond type,based on a set of potential limitations.An initial analysishas identified11potential limitations (Table8-4). Table8-4Limitations identified in preliminary aquaculture pond suitability analysis LIMITATIONRATIONALE Soil pHSoil pHlevels outside tolerance range can be detrimental to animal health andreduce crop production efficiency Soil typeConstruction of earthen ponds requireslow-permeability characteristics (e.g.high clay content) Acid sulfate soilsAcid sulfate soils can bedetrimental to animal health and reducecrop production efficiency Soil salinityHigh soil salinitycanbe detrimental to animal health and reducecrop production efficiency SlopeLand slope has a significantbearing on required capital investment to construct ponds ElevationLand elevation has significant bearing on ability to drain ponds and crop production efficiency Agriculture and aquaculture viability|81 Distance to marine water Pond distance from marine water source has a large bearing on required capital investment and ongoing crop production efficiency Rainfall High rainfall have potential to significantly reduce source water salinity, which can be detrimental to crop production Flooding Flooding events have potential to significantly reduce source water salinity, which can be detrimental to crop production Cyclone Associated infrastructure damage and power losses can be detrimental to animal health, and reduce crop production efficiency Zoning Current land zoning (including native title) will have a significant bearing on the probability of obtaining aquaculture licence approval LIMITATION RATIONALE 8.2.3 AQUACULTURE VIABILITY ASSESSMENT The aquaculture viability assessment will comprise a series of interrelated tasks, some of which will be undertaken in conjunction with other activities. These are briefly discussed below. Economic analysis An economic analysis of exiting or prospective aquaculture enterprises in the Darwin catchments will be undertaken in conjunction with the socio-economics activity (Chapter 10) and, as much as possible, consistent with the approach taken in the agricultural viability analysis (Section 8.1). As with the agricultural viability analysis, the focus of the analysis will be to provide insights on the issues and opportunities associated with developing different marine and freshwater aquaculture operations As with the agricultural analysis, it is proposed that the starting point of the analysis be the returns required to meet investment imperatives. This will involve initially assessing the likely capital costs involved in the development of different aquaculture activities, and the normal return that would be expected to make such an investment viable. The necessary output prices and production levels to achieve this given a range of likely input prices and availability will be assessed. The likelihood of these prices and/or production levels being achieved will be examined based on existing market information. Similarly, the potential available of suitable inputs to achieve the required output levels will also be assessed. To achieve this, the approach will still require significant individual aquaculture business assessments, using modelling and expert knowledge. Detailed production models for a range of different aquaculture businesses have been developed by the Queensland Government (https://publications.qld.gov.au/dataset/agbiz-tools-fisheries-aquaculture) which are applicable to the Darwin catchments and can underpin these analysis. Information on capital costs and input prices are limited, so the use of expert knowledge and local industry experience will be particularly important. The logic behind this approach is shown in 8-2. For more information on this figure please contact CSIRO on enquiries@csiro.au Figure 8-2 Simplified schematic of the proposed top-down approach to aquaculture viability and economic assessment Exploring Indigenous and community stakeholder perspectives Interviews conducted as part of the socio-economics activity (Chapter 10) will scope key issues for local and community stakeholders with respect to water resource development, including aquaculture development. Stakeholders include potential smaller-scale investors in new enterprises, and stakeholders from the wider community who are interested in existing goods and services provided by the natural resource base that may be affected by aquaculture development. Further information on Indigenous perspectives of aquaculture will be derived from the Indigenous aspirations and water values activity (Chapter 11). Collectively, this analysis will assist in understanding and overcoming barriers: . the enablers of, and barriers to, local and Indigenous participation in aquaculture development, such as – knowledge and skills – compatibility with existing business operations – social and cultural issues . key issues for community stakeholders in wider social licence-to-operate for aquaculture development, such as – planning and governance requirements – perceptions of impacts – distribution of benefits – sustainability of the opportunity. Risk analysis Water temperature is the most critical influence on aquaculture production duration, from pond to plate. Warm-water aquaculture species generally have a broad thermal tolerance, although optimal production occurs within a narrow bandwidth of this range. A report by Hobday et al. (2008) suggested that a rise in seawater temperature in northern Australia is likely to increase the area of suitable land and improve production efficiency of prawn aquaculture. In coastal areas, water salinity typically ranges from 37 ppt in oceanic waters to less than 1 ppt in fresh water in the upper catchment. The water salinity of a catchment is likely to vary widely, depending on location, tidal flow, precipitation and freshwater runoff. In aquaculture, the capacity for a species to osmoregulate (maintain internal salt concentrations) has a large bearing on its suitability for culture in a specific location. Euryhaline species such as barramundi have a broad salinity tolerance and are able to thrive in locations affected by variable salinity. The majority of established species, such as freshwater crayfish and to a lesser extent marine prawns, have a narrow salinity tolerance range for optimal production. Policy and legislation Northern Australia has the potential to support diverse economic activity. However, some of the rules governing land use and effluent discharge, for example, are likely to constrain aquaculture development in these catchments. Together with legal support provided by the socio-economics activity (Chapter 10), a review of current land arrangements (such as native title) and aquaculture- related policy will seek to identify development constraints in each jurisdiction. 8.3 References ABARES (2016) Agricultural commodities: March quarter 2016 CC BY 3.0. Ash, A (2014) Factors driving the viability of major cropping investments in northern Australia – an historical analysis. Northern Australia: food and fibre supply chain synthesis study: Appendix 3. CSIRO and Australian Bureau of Agricultural and Resource Economics and Sciences, Canberra/ABARES, Australia. Ash A and Gleeson T (2014) Northern Australia: food and fibre supply chain synthesis study. CSIRO and Australian Bureau of Agricultural and Resource Economics and Sciences, Canberra. Carberry PS, Cogle AL and McCown RL (1991) Future prospects for cropping in the semi-arid tropics of north Queensland. Land Management Research, CSIRO Division of Tropical Crops and Pastures, St Lucia, Queensland. Carberry PS, Hochman Z, Hunt JR, Dalgleish NP, McCown RL, Whish JPM, Robertson MJ, Foale MA, Poulton PL and Van Rees H (2009) Re-inventing model-based decision support with Australian dryland farmers. 3. Relevance of APSIM to commercial crops. Crop & Pasture Science 60, 1044–1056. Hobday, A. J., E. S. Poloczanska, and R. J. Matear (eds) (2008) Implications of Climate Change for Australian Fisheries and Aquaculture: a preliminary assessment. Report to the Department of Climate Change, Canberra, Australia. August 2008. Holzworth D, Huth N, deVoil P, Zurcher E, Herrman N, McClean G, Chenu K, van Oosterom E, Snow V, Murphy C, Moore A, Brown H, Whish J, Verrall S, Fainges J, Bell L, Peake A, Poulton P, Hochman Z, Thorburn P, Gaydon D, Dalgliesh N, Rodriguez D, Cox H, Chapman S, Doherty A, Teixeira E, Sharp J, Cichota R, Vogeler I, Li F, Wang E, Hammer G, Robertson M, Dimes J, Whitbread A, Hunt J, van Rees H, McClelland T, Carberry P, Hargreaves J, MacLeod N, McDonald C, Harsdorf J, Wedgwood S and Keating, B (2014) APSIM — evolution towards a new generation of agricultural systems simulation. Environmental Modelling & Software 62, 327-350. Keating BA, Carberry PS, Hammer GL, Probert ME, Robertson MJ, Holzworth D, Huth NI, Hargreaves JNG, Meinke H, Hochman Z, McLean G, Verburg K, Snow V, Dimes JP, Silburn M, Wang E, Brown S, Bristow KL, Asseng S, Chapman S, McCown RL, Freebairn DM and Smith CJ (2003) An overview of APSIM, a model designed for farming systems simulation. European Journal of Agronomy 18, 267–288. McKeon G, Ash A, Hall W and Stafford Smith M (2000) Simulation of grazing strategies for beef production in north-east Queensland. In: Hammer GL, Nicholls N and Mitchell C (eds) Applications of seasonal climate forecasting in agricultural and natural ecosystems, Kluwer Academic Publishers, The Netherlands, 227–252. Mcleod I, Pantus F and Preston N (2002) The use of geographical information system for land- based aquaculture planning. Aquaculture Research 33, 241–250. Pearson L and Langridge J (2008) Climate change vulnerability assessment: review of agricultural productivity. CSIRO Climate Adaptation Flagship Working paper no. 1. Petheram C, Watson I and Stone P (eds) (2013a) Agricultural resource assessment for the Flinders catchment. A report to the Australian Government from the CSIRO Flinders and Gilbert Agricultural Resource Assessment, part of the North Queensland Irrigated Agriculture Strategy. CSIRO Water for a Healthy Country and Sustainable Agriculture flagships, Australia. Petheram C, Watson I and Stone P (eds) (2013b) Agricultural resource assessment for the Gilbert catchment. A report to the Australian Government from the CSIRO Flinders and Gilbert Agricultural Resource Assessment, part of the North Queensland Irrigated Agriculture Strategy. CSIRO Water for a Healthy Country and Sustainable Agriculture flagships, Australia. Preston N, Chalmers I, Moore A, Zurcher E, Verrall S, Gaydon D, Coman G, and Arnold S (2015) Sustainable Development of Northern Australia. Aquaculture futures for coastal Northern Australia Final Report December 2015. CSIRO Agriculture. Savage J and Hobshawn P (2015) Australian fisheries and aquaculture statistics 2014. FRDC project 2014/245, Australian Bureau of Agricultural and Resource Economics and Sciences, Canberra. Skirtin M, Sahlqvist P and Vieira S (2013) Australian fisheries statistics 2012. FRDC project 2010/208. Australian Bureau of Agricultural and Resource Economics and Sciences, Canberra. Verburg K, Bond WJ and Smith CJ (2003) Use of APSIM to simulate water balances of dryland farming systems in south eastern Australia. CSIRO Land and Water, Canberra. Webster T, Poulton P, Yeates SJ, Cocks B, Hornbuckle J, Gentle J, Brennan McKellar L, Mayberry D, Jones D and Wixon A (2013) Agricultural productivity in the Flinders and Gilbert catchments. A technical report to the Australian Government from the CSIRO Flinders and Gilbert Agricultural Resource Assessment, part of the North Queensland Irrigated Agriculture Strategy. CSIRO Water for a Healthy Country and Sustainable Agriculture flagships, Australia. Yeates SJ (2001) Cotton research and development issues in northern Australia: a review and scoping study. Australian Cotton Cooperative Research Centre, Darwin. 9 Water storage The purpose of the water storage activity is to provide a comprehensive overview of the different water storage options in the Darwin catchments, to enable decision makers to take a long-term view of water resource development and to inform future allocation decisions. In this chapter, methods are described by which different water storage options will be assessed in the Darwin catchments. The key questions that this activity seeks to address in the Darwin catchments include: • Where are the highest yielding and most geologically suitable dam sites? • How much water could large instream and offstream dams yield, and at what cost? • Are there viable opportunities for dams to supply water for multiple purposes in the Darwin catchments? • Would the construction of a large dam(s) result in ecological change? • After how many years would large dam(s) in the Darwin catchments infill with sediment? • Where are the best opportunities for hydro-electric power generation? • What is the opportunity for storing water in offstream farm-scale water storages such as ring tanks? • How much water is contained in naturally occurring wetlands and waterholes? • Where are the better opportunities for managed aquifer recharge? This chapter is in four parts. The first part details the methods that will be undertaken as part of a pre-feasibility assessment of large (i.e. >10 GL) instream and offstream storages. The second part discusses methods for assessing the opportunities for farm-scale water storage structures (i.e. <10 GL), such as hillside dams and ring tanks. The third part describes methods for assessing the opportunity to use water from natural wetlands and waterholes, and the fourth part examines methods for assessing the opportunities for managed aquifer recharge. 9.1 Introduction In a highly seasonal climate such as the Darwin catchments, and in the absence of a suitable groundwater resource, industries that require year-round use of water will need to invest in water storage infrastructure. Before investigating potential dam sites, it is instructive to assess the existing storages within the Darwin catchments. 9.1.1 EXISTING MAJOR STORAGES In the Darwin catchments, there is limited opportunity to use additional water from existing water storages. The only large dams in the Darwin catchments are the Darwin River Dam (265 GL capacity) in the Finniss River catchment and Manton Dam (16 GL capacity) in the Adelaide River catchment. Darwin River Dam was constructed in 1972 and supplies approximately 90% of Darwin’s water. Manton Dam was constructed in 1942; although it was Darwin’s first reliable water supply,today it isa popular recreational area,and water isusedonlyfor consumptivepurposes in emergencies. Table9-1Major damsin the Darwincatchments(CSIRO 2009) STORAGE NAMERIVERCAPACITY(GL) Darwin River DamDarwin River265Manton DamManton River16 9.1.2PROPOSED STORAGES The Northern Territory Powerand WaterCorporation(PWC), which isresponsible for water services across theNorthern Territory,has commissioned numerous studies onpotential dam sitesin theDarwin catchmentsover thepastcouple ofdecades. In 1979,a pre-feasibility analysisoffuturesources of water for Darwin(SMEC, 1979)identified 12potentialdam and weir sites in thevicinity of the cityof Darwin. On thebasis of these investigations,three of these sites were short- listed for more detailedinvestigations inthe early 1990s (Paiva,1991a,b):Marrakai (Stewart andBaker, 1987; GHD,1990;Paiva, 1991a),Mount Bennett (Ullman and Nolan, 1990;Paiva,1992) andWarrai (Paiva, 1990,1991b). Inrecent years,thePWChasbeen undertakingdetailed investigations of alargeoffstream storage location along themid-reaches ofthe Adelaide River. The PWCanalysis hasfocusedexclusively on the supply of water to Darwin, notfor agricultural uses. 9.2Large instreamand offstream storages This section describes the methods by whichpotential dam sites will beselected (Section9.2.1) forpre-feasibility analysisin theDarwin catchments(Section9.2.2).The pre-feasibility analysiswill beundertaken acrossthe entireAssessment area,not just in the vicinity of Darwin city.In thisanalysis,potential dam sites willbe examined fortheir potential to supplywater for multiplepurposes.Two orthree sites will be short-listedfor more detailed analysis (Section9.2.4).Ifone oftheshort-listedsites isthe PWC offstream storage (notingthat the Assessment’s remit isbroaderthan supplying water toDarwin, and sothe criteria for short-listing sites in the Assessmentmaybedifferent from those used bythePWC),the Assessment will seek to complement the detailedstudiesalready undertaken bythePWC,rather thanduplicatethem. Opportunities for hydro- electric power generation will also be investigated. 9.2.1INITIAL DAM SITE SCREENING AND SELECTION Instream storages are highly contentious, because they can affect existing environmental, culturaland recreationalvalues. The process bywhich largedams areselected for investigationhas often been unclear or seeminglysubjective,and the decision-making processis not always transparentto all stakeholders. This section presents an openand transparent method by which sites willbeselected for a pre-feasibility analysis. This first phaseof investigationwill involve identifying all large dam proposals thathavebeen thesubject of earlier or currentinvestigations. However, it is likely that theseprevious studies wereundertaken by a rangeof organisations, at different timesand todifferentdegrees of detail. These 88|Proposed project methods factors make it very difficult to compare the outcomes of one study with another. Hence, all previously studied potential dam sites will be reassessed as part of the pre-feasibility analysis, using a consistent set of models and methods, and the results will be made publicly available. Concurrently, the DamSite model will be run across the Darwin catchments to ensure that no promising potential dam sites were overlooked by previous studies. The DamSite model is a series of algorithms that automatically assesses every location in a catchment for its potential as a dam, based on having suitable topography and hydro-climatology. DamSite model The DamSite model (Read et al., 2012; Petheram et al., 2013) uses a digital elevation model to assess all locations on a river network and test simulated dam walls of varying heights, to produce a comprehensive dataset of sites with relevant attributes, including catchment area, runoff, reservoir volume, reservoir surface area, dam height, dam width and dam face area. Saddle dams are included in the assessment if required by the terrain. This dataset will include an exhaustive set of potential dam locations – more than 100,000 potential sites in each catchment (based on previous experience with DamSite). These sites are then filtered to identify approximately 5000 of the most suitable sites, using a combination of criteria including yield, construction cost, and yield per dollar construction cost. Yield is initially calculated at every location and every metre increment height using the Gould– Dincer Gamma method, and then refined for the subset of around 5000 sites using a behaviour analysis model. Construction cost is calculated using a cost algorithm that serves to penalise higher and longer dam walls. The cost algorithm will be refined as part of the Assessment to improve its accuracy. Key input data to the DamSite model are the SRTM-H (Shuttle Radar Topography Mission; the best available digital elevation model across the Darwin catchments), gridded climate data generated by the climate activity and gridded runoff data from the surface water hydrology activity. 9.2.2 PRE-FEASIBILITY ANALYSIS The pre-feasibility analysis is largely a detailed desktop analysis of a selection (~10) of the more promising potential dam sites in the Darwin catchments. It involves a comprehensive review of past studies, a reassessment of each site using a consistent set of methods and models, and a site investigation by an experienced infrastructure planner and engineering geologist. Each site will be evaluated and the results reported against a consistent set of criteria. The criteria and the methods by which each criteria will be evaluated are described in Table 9-2. Table9-2Proposed methodsfor assessing potential dam sites in theDarwin catchments PARAMETERDESCRIPTION Previous investigationsLiteraturedocumenting previous dam site investigations will beobtained from a variety ofsources,includingterritoryagency libraries Description of proposalBased on review of past reports.Where no documents areidentified,thiswill benoted. Forthe short-listed potentialdam sites,the original proposalswill bemodified to reflect morerecent dataandmethods,and contemporary thinking Regional geologyThe regionalgeology for each dam sitewill beassessed using thebest available geologicalmapping data, previous dam studies and literature sourced fromterritoryagency libraries Site geologyThe site geology for each dam sitewill beassessed using the bestavailable geology data and a site visit by anexperienceddam geologist Reservoir rim stability andThese parameters will beassessed by overlaying inundated area atfull supplylevel (FSL)on leakage potential1:250,000 or 1:100,000 geology data, and a site visit by adam geologist Proposed structuralBased on review of past reports.Where no documents areidentified,thiswill benoted. Forarrangementthe short-listed potentialdam sites,new conceptual arrangementswill bedeveloped, which better reflect contemporary thinking and more recent data Availability of constructionBased on review of available literature, site visits andproximity toquarry locationsmaterials Catchment areaCatchment areas will bederived from SRTM-H. In the majority of cases,the SRTM-H data areconsidered to be superior to historical topographic data for deriving catchment areas and computing reservoir volumes Flow dataMean and median flowswill becomputedusing observed data from the neareststreamflowgauging station CapacityDam capacitywill bederived from SRTM-H, unless stated otherwise.For potential dams,the dead storage volume willbe assumed to be 2% of the reservoir capacity at FSLReservoiryield assessmentA behaviour analysis model willbe used to assess the reliability of different yields.Threeassessments will be undertaken at each dam site: 1) under Scenario A (historical daily climatedata) for a range of dam wall heights and a perennial crop demand pattern usingthe baseline river model; 2) under Scenario Ausing the proposed structural arrangement, the baseline river model and i) a perennial, ii)a dry season,and iii)a wet-season planting crop demand pattern; and3) under Cwet, Cmid and Cdry (i.e. future climate data) for the proposed structural arrangement,baseline river model and a perennial crop demand pattern. The performance of each reservoir will be reportedin terms of the annual time reliability and thevolumetric reliabilityOpen water evaporationMorton’s wet environment areal potential evaporation(Morton1983)and a stabilitycorrected bulk aerodynamic formula (Liuet al.,1979) Opportunities for hydro- electric power generationHydro-electricpower calculationswill bebased on head and flow,and proximity to connect into electricitygridImpacts of inundationonexisting property andinfrastructureBased on review of past studies,satellite imagery, GIS overlays and site visitEcological and culturalconsiderations raised byprevious studiesBased on review of past studies Estimated rates of reservoirsedimentation Sedimentation rates will becalculated using estimatedsediment yields and the FSL damcapacity for each site. Sediment yieldswill becomputed from an empirical relationship derived from 10 sediment yieldstudies across northernAustralia.The rates of reservoirsedimentation will bepresentedfor 1, 10, 30, 100 and 1000 years, as well as the number ofyears taken to 100% infill.Minimum (best-case), expected and maximum (worst-case) estimates will beprovided Water quality and stratificationAssessed using aone-dimensionalhydrodynamic reservoir modelconsiderations 90|Proposed project methods Environmental considerations Barrier to fish movement Mapped data on the ecological assets and the fish species distribution in the Darwin catchments will be sourced from the aquatic ecology activity. Data on the persistence of waterholes in both catchments will be sourced from remotely sensed imagery. Ecological implications of inundation The latest available vegetation mapping data will be used to assess the potential implications of inundation on vegetation communities. No field ecological surveys will be undertaken as part of the Assessment Cultural heritage considerations A desktop Indigenous cultural heritage review will be undertaken by searching the Department of Aboriginal and Torres Strait Islander Partnership databases. This will only be undertaken for the short-listed potential dam sites Estimated cost For all potential dam sites that were previously investigated, the cost estimate reported in the literature will be adjusted for inflation using the Australian Consumer Price Index. This will be compared with an estimate provided by the DamSite model automated dam cost algorithm, informed by local preliminary geology assessment. The uncertainty associated with these estimates is likely to be between –10% and +50% For the short-listed potential dam sites, more detailed cost estimates will be calculated by developing conceptual arrangements for each of the dams, informed by flood design modelling. Cost rates applied for each item of work may be derived from recent relevant construction activity or from local suppliers. The uncertainty in cost of the short-listed sites is likely to be between about –10% and +30% Estimated cost per ML of supply Estimated capital cost divided by the yield at 85% reliability, as computed by the Assessment under the proposed structural arrangement Potential costs and benefits Based on reviewed literature Summary comment Provided by Assessment personnel PARAMETER DESCRIPTION 9.2.3 ASSESSMENT OF SYSTEM YIELD The system yield from cascades of two or more of the better reservoirs in a catchment will be investigated as part of the river system scenario modelling. 9.2.4 SHORT-LISTED DAM SITES Based on the pre-feasibility analysis, a short list will be compiled of approximately three of the more promising sites for the multipurpose supply of water. Short-listed sites will be primarily selected based on topography of the dam axis, geological conditions, proximity to suitable soils, water yield, opportunities for hydro-electric power generation, supply of water to other non- agricultural industries (e.g. cities, mining), and ecological and cultural considerations. Additional studies undertaken for the short-listed dam sites will include a flood design study, a detailed cost estimate, and a desktop cultural heritage assessment. High-resolution data will be acquired for the short-listed dam sites using laser altimetry or photogrammetry methods. Depending on the short-listed sites selected, studies could also potentially include more detailed hydrodynamic reservoir modelling to better estimate dead supply levels (i.e. based on water quality constraints), or accurate evaporation measurements from water bodies using laser scintillometers or submerged evaporation pans. For any of the short-listed options to advance to construction, a feasibility analysis (as per the Nullinga dam feasibility analysis announced in the Northern Australia White Paper) would need to beundertaken, which would involve several iterations of detailed (and expensive) studies. Studiesatthis level of detail arebeyond the scope ofthe current region-scale resource assessment. 9.3Farm-scale instreamand offstream storages This section describes the methods for assessingthe opportunities for farm-scale instream andoffstream water storagestructures (i.e.<10GL). Instream storages include gully dams andhillsidedams, while offstream water storage facilities can takethe form of ringtanks, turkey nesttanksand excavated tanks (described in moredetail in Table7.4). Weirs can alsobeused in conjunction with some offstream water storages, wherethe weir isused to raise theupstream water level toallow diversion into an offstream storage or the creation of a pumpingpool. The most suitabletype of farm-scale waterstoragedepends on a numberof factors, including topography,theavailabilityof suitable soils, excavation costs andthe source of water (e.g.groundwater or surfacewaterpumping, flood harvesting). Table9-3Types of offstream water storages (Lewis, 2002) TYPE OF FARMSCALE STORAGEDESCRIPTIONSTORAGE TO EXCAVATIONRATIO Gully damAn earth embankmentbuilt across a drainage line. Dams are normallybuilt from material located in the storage area upstream ofthe dam site10:1 (favourable conditions) Hillside damAn earth dam located on a hillside or slope,and not in a defined depression or drainage line5:1 (on flatter terrain) 1:1 (on steeper slopes) RingtankA storage confined entirely within a continuousembankmentbuilt frommaterial obtained within the storage basin1.5:1 (small tank) 4.5:1 (large tank) Turkey nest tankA storage confined entirely within a continuousembankmentbutbuiltfrom material borrowed from outside thestorage area. All water isthereforeheld above ground levelUsually smaller than ringtanks,and lower storage toexcavation ratioExcavated tankRestrictedto flat sites and comprise excavations below the naturalsurface. Excavated material is wasted. Generally limited to stock anddomestic use,and irrigation of high-value cropsLow storage to excavationratio Thefollowing analysis will be undertaken to assess the opportunities forfarm-scale water storagesin theDarwin catchments: •The soil attribute grids (to adepth of 1.5 m) generated aspart of theland suitability activity and locally specific ruleswillbe usedto identifythose parts of the Assessment area that are moreand less suitable forfarm-scale water storages.The Assessment will draw on borelithology logs, expert and local knowledge, and electromagneticdatato make assessments below 1.5m. •The DamSitemodel will be usedto identifythose parts of the Assessment area that are likely tobehydrologicallyand topographically favourable forhillsidefarm dams. •Likely physical constraints to water pumping in key river reaches (i.e.minimum pumpingthresholds)will beestimated. In assessing region-scaleeconomics of water harvesting schemes, local variations in scale and site- specific nuances can present challenges. These can result in considerablydifferent construction 92|Proposed project methods and ongoing operational costs from one site to another (e.g. costs for different amounts of diesel required for pumping, removal of sediment deposited in diversion channels, replacement of worn and damaged equipment). Hence, operationally, each site would require its own specifically tailored engineering design. As a result, the Assessment will not produce individual engineering designs for water harvesting infrastructure for each landholder in the Assessment area; this is beyond the scope and resources of the Assessment. Besides, most landholders will have observed the way in which water moves across their land and will have given considerable thought to their most suitable water harvesting configurations. However, the Assessment will provide some overarching principles that could be used by individual landholders in designing, siting and costing water harvesting infrastructure in the Darwin catchments, as well as a relevant list of references on farm dam planning, construction and maintenance. 9.4 Quantifying the volume of natural wetlands and waterholes The possibility of using wetlands to supply water for irrigation while maintaining their ecological functioning has been the focus of research in southern Australia in recent years (Ning et al., 2012; Watkins et al., 2012; Sammond et al., 2012), and the ecological aspects are being investigated as part of the CSIRO project exploring the multiple use of wetlands in northern Australia (Table 1-2). This section describes methods to measure and estimate the storage volume of natural wetlands and waterholes. These include permanent pools in rivers and small water bodies outside of river channels that persist through the dry season. Ecological assessments of how wetlands in northern Australia could potentially be ‘operated’ for multiple purposes will not be included in the Assessment, but will be investigated as part of a companion CSIRO project on the multiple use of wetlands in northern Australia. The total storage volume in natural surface water storages will be estimated by mapping the extent of surface water bodies in the catchments and applying a model relating stored water volume to the surface geometry. The extent will be derived from Geoscience Australia’s Water Observations from Space (WOfS) data, which are based on the Landsat time series archive. These data support the detection and measurement of water bodies larger than about 50 m × 50 m. A model linking water volume to surface geometry will be developed using measured bathymetry in a variety of small water bodies throughout the catchments. The model will include regional variables such as topographic relief and geology, to capture regional differences in the relationship between surface shape and water volume. The actual storage volumes of a number of representative wetlands and waterholes will be measured using a sonar depth sounder mounted on an electrically powered radio-controlled boat. Using a remotely operated boat avoids the hazards of operating manned watercraft in waters where crocodiles may be present, and reduces the logistical difficulty of accessing water bodies and operating in shallow waters. The sonar measurements will be collected using a commercial sonar that records measured depths and GPS position. The equipment will have the capacity to measure depths of less than 0.5 m to greater than 50 m and will collect data at rates of 2 to 8 ha per hour of operation. 9.5 Managed aquifer recharge 9.5.1 INTRODUCTION Managed aquifer recharge (MAR) is the intentional recharge of water to aquifers for subsequent recovery for economic productivity and environmental benefit (NRMMC-EPHC-NHMRC, 2009). Where suitable aquifers are present, MAR or subsurface storage offers the advantage of reducing significant evaporative losses (Hostetler, 2007) and the capital costs associated with dams. Aquifers can be recharged using a variety of techniques, such as by infiltration of water through permeable sediments beneath a dam, river or basin; via injection wells into deeper confined aquifers; or using impermeable barriers to create underground dams or sand dams (Figure 9-1). Terrain and hydrogeology determine the most efficient form of MAR in an area, and will influence the economic viability of the scheme (e.g. infiltration schemes are generally lower cost than injection techniques). MAR can contribute to planned conjunctive use, whereby excess surface water can be stored in an aquifer, for subsequent use when the surface water resource is unavailable (Evans et al., 2013; Lennon et al., 2014). It can also be used to manage potential risks of groundwater resource development, such as creating a hydraulic barrier to prevent seawater intrusion, or providing protection to riparian vegetation and groundwater-dependent ecosystems (Ishida et al., 2011). Importantly, dams and aquifers can be used in combination where shallow surface dams can detain water, allowing greater time for infiltration to occur, or release water to infiltrate into the river bed downstream of the dam wall (e.g. Opthmalia Dam in Western Australia, Little Para Reservoir in South Australia) (Dillon, 2016). A number of infiltration techniques may be suitable for use in northern Australia to support increased economic productivity such as agriculture (Figure 9-1): • Underground dams: a trench constructed and filled with low-permeability material across a streambed in an ephemeral stream, where flow is constricted by a basement high to retain flood flow and allow more time for infiltration – for example, in Brazil (UNEP, 1997). This MAR technique has been applied on the Ashburton River at Minderoo in the Pilbara region of Western Australia, using several metres of concrete and steel to construct an ‘upside-down weir’ and underground dam (Schwartz, 2015). More broadly, use of a mixed-in-place slurry-wall method since the 1990s has allowed the size of subsurface dams to increase, with storage of up to 10 GL in Japan. In southern China, 52 underground dams in karstic limestone store 40 GL, which is used to irrigate an area of 8900 km2 (Ishida et al., 2011). • Percolation tanks and recharge weirs (check dam, gully plug): dams built in ephemeral streams to detain water that infiltrates through the bed, increasing storage in unconfined aquifers – for example, hard rock aquifers in India (Perrin et al., 2010). The Gajwel watershed (85 km2) in Andhra Pradesh uses more than 30 percolation tanks to retain the majority of surface runoff and has more than 600 ha under irrigation. • Infiltration ponds and channels: offstream storages to allow infiltration to an unconfined aquifer. Infiltration basins in the Burdekin Delta (Queensland) recharge up to 45 GL/year to support irrigated agriculture. This scheme provides a water resource, mitigates seawater intrusion and is Australia’s oldest MAR scheme – it has been in operation since 1965 (Dillon et al., 2009). Water is reliably supplied by the Burdekin Falls Dam. • Sand dams: built in ephemeral streams on low-permeability lithology to trap sediment and create a raised aquifer – for example, in Kenya (Ishida et al., 2011). These are typically referred to as sand dams overseas, but the term has a different connotation in Australia; in the Lower Burdekin, it refers to low mounds of sand partially spanning a river, behind which a pumping pool is created. • Recharge releases: dams on ephemeral streams that detain floodwater, which can be used for strategic slow releases downstream to match the infiltration capacity into underlying unconfined aquifers – for example, Little Para River in South Australia (NRMMC-EPHC-NHMRC, 2009; Dillon, 2016). Injection well techniques are higher cost, but may be suited to high-value water uses (e.g. mining, drinking water supply) or to mitigate impacts of groundwater extraction: • Aquifer storage and recovery: well injection and recovery using the same well, which can store water in aquifers of impaired quality (brackish) – for example, South Goulburn Island in the Northern Territory (Knapton, 2002; Page et al., 2010). • Aquifer storage transport and recovery: well injection and recovery from a different well to enhance water treatment – for example, Salisbury (South Australia) with urban stormwater (Page et al., 2010). For more information on this figure please contact CSIRO on enquiries@csiro.au Figure 9-1 Schematic diagram illustrating various techniques that can be used for managed aquifer recharge, depending on the hydrogeological conditions ASR = aquifer storage and recovery; ASTR = aquifer storage transport and recovery Source: NRMMC-EPHC-NHMRC (2009) 9.5.2 METHODS There have been a number of MAR studies in the vicinity of Darwin, primarily focused on augmented drinking water supply. Dames and More (1992a) recommended well injection techniques at Adelaide River to augment existing groundwater supply from highly sheared and fractured sections of the Burrell Creek Formation, overlain by a thick sequence of low-permeability sediments. They also reported that recharge via infiltration could be enhanced by modifying the streambed of Snake Creek. Just outside the Darwin catchments, aquifer storage and recovery was shown to be feasible on South Goulburn Island to store excess groundwater from a shallow laterite aquifer in a deeper sandstone aquifer, to expand dry-season water supply to the Warruwi community (Knapton, 2002; Pavelic et al., 2002). To the south of Darwin, a preliminary assessment of MAR to augment water supplies at Pine Creek (Dames & Moore, 1992b) also recommended well recharge of the fractured rock aquifer via gravity feed or low pressure as the most cost-effective method. Further afield in the Northern Territory, MAR has been implemented successfully using infiltration basins, and aquifer storage and recovery at Alice Springs. Here, infiltration basins are operated intermittently within a soil aquifer treatment scheme to recharge treated wastewater within a paleochannel, for planned use in irrigation (Page et al., 2010; Bekele et al., 2015). The opportunities for MAR in the Darwin catchments will be investigated in two stages. The first stage involves a regional-scale opportunity assessment, which will broadly outline the opportunity for MAR in the Darwin catchments. In the second step, a more detailed pre-feasibility analysis will be undertaken at two of the more promising locations for MAR. MAR region-scale opportunity assessment Before evaluating the feasibility of MAR, a conceptual understanding of the hydrogeology is an essential first step in identifying the likely opportunity and understanding which techniques may be best suited to the environment. Initially, an understanding of the hydrogeology and the opportunities for increased groundwater storage will be developed in conjunction with investigations being undertaken by the groundwater hydrology activity (Chapter 6). The potential opportunities for further storage resulting from increased groundwater extraction will be considered. Evaluation of the potential for injection well techniques in different parts of the Darwin catchments will be limited to a broad articulation of the likely opportunities, based on the conceptual understanding of the regional hydrogeology. In reality, the opportunity to use injection well techniques to store water will be highly site dependent, and assessment of this opportunity would require considerable site-based investigation and resources. It is beyond the scope of the Assessment to undertake investigations at this level of detail across the Assessment area. To assess the potential for infiltration-based MAR methods in the Darwin catchments, a region- scale suitability framework will be implemented within a GIS environment. This will include a combination of readily available spatial datasets and targeted field measurements. Table 9-4 outlines parameters that may be included in the region-scale assessment of the suitability of infiltration-based MAR in the Darwin catchments. Other factors that will be examined include proximity to existing infrastructure, maintenance requirements and potential demand, which could be from urban, industry or agriculture activities. Land suitable for irrigated agriculture or aquaculture (as evaluated by the land suitability activity) will be used as a surrogate for potential future demand for water. Table9-4Parameters likely to be included in aregion-scale assessment of the suitabilityof infiltration-basedmanaged aquifer recharge PARAMETERMETHODCONCEPTUAL RELATIONSHIP Geometryof riverSatellite imageryControls storage volumeSlopeSRTM-HControls storage volume,and potential for scouring and cloggingParticle size distributionOpportunistic field measurements,Controls storage volume (porosity),influences saturated of bedsandsrelationships with sourcegeology,hydraulic conductivity and waterretentiondistance of travel and othercatchment attributesDepth of bedsandsOpportunistic field measurements,Controls storage volumecatchment terrainattributes such asMulti-resolution Valley BottomFlatness (MrVBF) Groundwater levelsExisting bore dataAdditional storage volume that could be createdGroundwater salinityField measurements, existing boreDetermines whether water can be used and for whatdatapurposeDepth to regolithTerrestrial Ecosystem ResearchStorage volumeNetwork (TERN) depth to regolithmapping, bore logs, opportunisticfield measurements Soil textureDigital soil mapping of texture in top Informs likely infiltration rates intop 1.5m. Typically,clay1.5mcontent increases with depth, unless there is an underlying paleochannelHydraulic conductivityEstimation from particle sizeLeast permeable sediments limitinfiltrationratedistribution at depths greater than 1.5m, opportunistic field measurementsClogging particlesRemote sensing of turbidityToinform likelihood of clogging(Chapter7), upstream geology andslopeGeologyGeological mapping dataStructural controls on subsurfacestorage (e.g. likelihood ofleakage) Proximity to riverGIS spatial analysisSource of water is required,cost effectiveness of MARReliability of streamflowHydrological modelling (Chapter5)Determines the reliability with which MAR could be filled Rate of rise and fall ofHydrological modelling (Chapter5)Determines need for intermediate storage basinshydrographs Pre-feasibilityassessment of MAR in theDarwin catchments Based on theresults of the region-scale opportunityassessment, for a smallselection oflocationswhere MAR shows mostpromise, a pre-feasibility assessment will beundertaken. This will includea risk-based assessmentaddressing water qualityand environmental hazards (NRMMC-EPHC- NHMRC, 2009),and relative costs associated withdifferent infrastructure configurations. It would also address operationalconstraints, such as thepotential foraquifer clogging and the costsassociated withmanagement.Management strategies for clogging duetoparticles in the sourcewater may includetreatment prior to recharge andperiodic removalof accumulated particulatesfrom theMAR system.The pre-feasibility investigations will onlybegin when thegroundwateractivity has developed asufficient hydrogeological understanding. The pre-feasibility assessmentwill use a combination of existing data and data generated bythe groundwater investigations. It 98|Proposed project methods will identify the nature of investigations required for a feasibility assessment in accordance with the Australian MAR guidelines (NRMMC-EPHC-NHMRC, 2009). 9.6 References Bekele, E., Donn, M., Barry, K., Vanderzalm, J., Kaksonen, A., Puzon, G., Wylie, J., Miotlinski, K., Cahill, K., Walsh, T., Morgan, M., McFarlane, D. and Dillon, P. (2015) Managed Aquifer Recharge and Recycling Options (MARRO): Understanding clogging processes and water quality impacts, Australian Water Recycling Centre of Excellence. CSIRO (2009) Water in the Gulf of Carpentaria Drainage Division. A report to the Australian Government from the CSIRO Northern Australia Sustainable Yields Project. CSIRO Water for a Healthy Country Flagship, Australia. Dames & Moore (1992a) Preliminary assessment of the feasibility of artificial recharge at Adelaide River. A report for the Power and Water Authority, Darwin. Dames & Moore (1992b) Preliminary assessment of the feasibility of artificial recharge to augment water supplies at Pine Creek, Northern Territory. A report for the Power and Water Authority, Darwin. Dillon P (2016) 21st century water storage: conjunctive use of dams and aquifers. International Water Power & Dam Construction Water Power Magazine. Dillon P, Pavelic P, Page D, Beringen H and Ward J (2009) Managed aquifer recharge: an introduction. Waterlines Report Series No. 13, National Water Commission, Australia. Evans WR, Evans RS and Holland GF (2013) Thematic paper 2: conjunctive use and management of groundwater and surface water within existing irrigation commands: the need for a new focus on an old paradigm. Groundwater Governance: A Global Framework for Country Action. GEF ID 3726. GHD (1990) Marrakai dam site. Preliminary geotechnical investigation. Gutteridge Haskins and Davey Pty Ltd. Reference no. 11090/00. Power and Water Authority Technical Report WRD90099. Power and Water Authority, Darwin. Hostetler S (2007) Water banking. Science for decision makers. Bureau of Rural Sciences, Australian Government Department of Agriculture, Fisheries and Forestry, Canberra. Ishida S, Tsuchihara T, Yoshimoto S and Imaizumi M (2011) Sustainable use of groundwater with underground dams. Japan Agricultrual Research Quarterly. DOI: 10.6090/jarq.45.51. Knapton A (2002) Aquifer storage and recovery project South Goulburn Island: hydrogeological investigations & trial storage and recovery. Northern Territory Government, Australia. Lennon L, Evans R, George R, Dean F and Parsons S (2014) The role of managed aquifer recharge in developing northern Australia. Ozwater’14, Proceedings of Australia’s International Water Conference and Exhibition. Australian Water Association, Australia. Lewis B (2002) Farm dams: planning, construction and maintenance. Landlinks, Collingwood, Victoria. Liu WT, Katsaros KB, and Businger JA (1979) Bulk parameterization of air-sea exchanges of heat and water vapor including the molecular constraints at the interface. Journal of the Atmospheric Sciences, 36(9), 1722-1735. Morton FI (1983) Operational estimates of lake evaporation. Journal of Hydrology 66(1–4), 77– 100. Ning NSP, Watkins SC, Gawne B and Nielsen DL (2012) Assessing the potential for using wetlands as intermediary storages to conjunctively maintain ecological values and support agricultural demands. Journal of Environmental Management 107, 19–27. NRMMC, EPHC and NHMRC (2009) Australian guidelines for water recycling: managing health and environmental risks (phase 2) – managed aquifer recharge). National Water Quality Management Strategy document 24. Natural Resource Management Ministerial Council, Environment Protection and Heritage Council, and National Health and Medical Research Council, Canberra. Viewed 22 February 2016, http://www.environment.gov.au/system/files/resources/d464c044-4c3b-48fa-ab8b- 108d56e3ea20/files/water-recycling-guidelines-mar-24.pdf. Page D, Dillon P, Vanderzalm J, Bekele E, Barry K, Miotlinski K and Levett K (2010) Managed aquifer recharge case study risk assessments. CSIRO, Australia. Paiva J (1990) Warrai dam flood studies. Report no. 77/1990. Water Resources Division, Power and Water Authority, Darwin. Paiva J (1991a) Marrakai dam yield study. Report no. 54/91. Water Resources Division, Power and Water Authority, Darwin. Paiva J (1991b) Warrai dam yield reappraisal. Report no. 53/91. Water Resources Division, Power and Water Authority, Darwin. Paiva J (1992) Mt Bennett dam yield reassessment. Report no. 32/92. Water Resources Division, Power and Water Authority, Darwin. Pavelic P, Dillon P, Barber C, Toze S, Yin Foo D, Jolly P and Knapton A (2002) Water banking trial at Warruwi, South Goulburn Island, NT. Stage 3 report to Department of Infrastructure Planning and Environment, NT. Centre for Groundwater Studies, Australia. Perrin J, Massuel S and Ahmed S (2010) Contribution of percolation tanks to total aquifer recharge: the example of Gajwel watershed, southern India. ISMAR7, Proceedings of the International Symposium on Managed Aquifer Recharge, Abu Dhabi. Petheram C, Rogers L, Eades G, Marvanek S, Gallant J, Read A, Sherman B, Yang A, Waltham N, McIntyreTamwoy S, Burrows D, Kim S, Podger S, Tomkins K, Poulton P, Holz L, Bird M, Atkinson F, Gallant S and Kehoe M (2013) Assessment of surface water storage options in the Flinders and Gilbert catchments. A technical report to the Australian Government from the CSIRO Flinders and Gilbert Agricultural Resource Assessment, part of the North Queensland Irrigated Agriculture Strategy. CSIRO Water for a Healthy Country and Sustainable Agriculture flagships, Australia. Read AM, Gallant JC and Petheram C (2012) DamSite: an automated method for the regional scale identification of dam wall locations. 34th Hydrology and Water Resources Symposium, Sydney, Australia, 19–22 November 2012. Engineers Australia, Canberra. Sammonds MJ, Vietz GJ and Costello JF (2012) Using water destined for irrigation to conserve wetland ecosystems: a basis for assessing feasibility. Water Resources Research 49, 4662– 4671. DOI: 10.1002/wrcr/20338. Schwartz D (2015) Andrew Forrest explores underground water reserved in bid to drought-proof Australia. ABC News, ABC. Viewed 29 March 2016, http://www.abc.net.au/news/2015-07- 24/andrew-forrest-explores-upside-down-weirs/6640994. SMEC (1979) Darwin water supply future source – appraisal study. Volume 2 – annexure. Technical Report WRD79022. Northern Territory Department of Transport and Works, Darwin. Stewart BJ and Baker A (1987) Marrakai dam safe yield studies. Internal file records. Water Resources Division, Power and Water Authority, Darwin. Ullman and Nolan (1990) Mount Bennet damsite. Report on geotechnical investigations. Technical Report WRD90100. Prepared by Ullman and Nolan (Geotechnic) Pty Ltd for Power and Water Authority, Darwin. UNEP (1997) Source book of alternative technologies for freshwater augmentation in Latin America and the Caribbean. Organization of American States, Washington. Viewed 29 March 2016, https://www.oas.org/dsd/publications/Unit/oea59e/begin.htm#Contents. Watkins SC, Ning NSP, Gawne B and Nielsen DL (2012) Managing wetlands as off-river storages: impacts on zooplankton communities. Hydrobiologia 701, 51–30. DOI: 10.1007/s10750-012- 1256-0. 10 Socio-economics 10.1 Overview The socio-economics activity will examine the social and economic viability of water resources development, with a particular focus on food and fibre production, and the opportunities to sustainably use water for multiple purposes. It will also examine how the values, perspectives and behaviours of individuals, groups and institutions may influence the nature and structure of any economic production. The key questions that this activity seeks to address in the Darwin catchments include: • What are the opportunities to profitably expand the existing array of agricultural activities (including aquaculture) through the development of local water resources for multiple uses? • What new agricultural activities (including aquaculture) might be developed that can viably cover the capital and operating costs of irrigation development within the constraints of transport and market conditions? • How has the economy of the Darwin catchments evolved in the past decade compared with other similar regions in Australia and how might it continue to evolve into the future? • What are the potential socio-economic impacts of different forms and scales of water resource development in the Darwin catchments? • What population and economic trends are likely to emerge under different development scenarios? • What are the likely domestic and export supply chain paths for the new agriculture options? • What are the current transport and logistical limitations that may impair development? • What are the best-bet infrastructure investment and regulatory strategies to reduce access issues? • What are the likely logistics costs (per tonne and total) to transport new and existing agricultural products? • What are the current limitations to consolidate produce on farm or at port, and what are the best-bet strategies to reduce these limitations? • What are the food perishability and shrinkage issues of transport, and what are the impacts for domestic and international market prices? • What current legislative, regulatory and policy requirements are relevant to the assessment and development of water resources (including water planning, environmental, tenure and native title, land management, and other relevant domains)? • Are the identified legislative, regulatory and policy requirements appropriate to facilitate the development of water resources while also protecting the catchments resource, environmental and Indigenous values? • What cross-jurisdictional legislative, regulatory and policy requirements affect the assessment and development of water resources? • What are key legislative and policy challenges to, and opportunities for, advancing the sustainable development of water resources? There are six major components: • multiscale agricultural viability assessment • aquaculture viability analysis • application of TraNSIT to identify bottlenecks and opportunities for logistics infrastructure • exploration of a range of stakeholder perspectives • characterisation of the policy and regulatory environment • exploration of regional socio-economic impacts. 10.1.1 LINKAGES TO OTHER ACTIVITIES The final design and scope of the economic and associated social analysis undertaken within this activity will necessarily be informed by the outputs of other activities in the Assessment.. These include, for example, inputs to the scheme scale (e.g. a development supporting multiple individually managed enterprises) and regional economic analysis from enterprise scale modelling undertaken for agriculture and aquaculture viability (Chapter 8), irrigation costs (Chapter 8), water storage costs (Chapter 9), and water availability and reliability estimated from surface water (Chapter 5) and groundwater (Chapter 6) modelling. Production responses under various development scenarios, where applied, will have also been informed by both present and projected climate data (Chapter 3) Assessment of the social licence-to-operate issues, and of the related policy, legislative and regulatory environment is supported by the Indigenous aspirations and water values activity (Chapter 11). The articulation of the social licence-to-operate issues, in an ecological context is also supported by the ecology activity (Chapter 12). It will also be indirectly supported by an additional CSIRO-funded project focusing on the multiple values of wetlands in the Darwin catchments (Table 1-2). 10.2 Multiscale economic assessment of water resources development for agriculture A set of economic analyses will address the question: Is irrigated agriculture developed on a significant scale (e.g. 30,000 hectares) economically viable? For the purposes of this chapter, these analyses will be conducted at scales ranging from the scheme scale to the regional scale. Paddock- and farm-scale economic analysis will be undertaken in conjunction with the agriculture and aquaculture viability activity (Chapter 8). Selection of appropriate baselines for potential irrigation development scenarios, while largely based on current land uses and market directions, will necessarily acknowledge the likely evolution of the agricultural sector within the catchments in the absence of further water infrastructure development. Otherwise, the projected returns on the investment in water development may be overvalued. Of particular focus will be the impacts of capital costs, water availability, crop type, irrigation system and commodity price on the viability of irrigation development, at both the farm scale and the scheme scale. Access to power for water extraction and delivery and the local presence of formed or sealed roads will impact on the cost of irrigation and inwards transport of inputs and outwards transport of produce. Consideration will also be given to the issue of how prospective user water charges might impact on the viability of potential irrigation developments, and the impact of infrastructure development costs being entirely user-funded or shared between private and public contributors. The analysis will be based on a number of development case studies and scenarios, including water availability and reliability, and projections of future climate used in other sections of the Assessment . In addition, the economic impact of various logistics infrastructure options will be evaluated (see below). 10.3 Economic assessment of water resources development for aquaculture The aquaculture sub-activity (Chapter 8), which will identify the potential production models (e.g. intensive, semi-intensive and extensive) and locations of aquaculture developments, economic analyses will be undertaken to explore the economic viability of aquaculture developments under a range of scenarios. The analysis will take into account: • species selection • markets • price movements • infrastructure and plant requirements, and associated capital and operating costs • regulatory, establishment and commissioning time frames for establishing an aquaculture enterprise • labour and skills required • accessibility of supporting infrastructure (e.g. distance to grid power, transport logistics) • environmental risks to the business • key environmental issues that need to be considered at development sites and associated costs • projections of future climate (Chapter 2, Chapter 3). 10.4 Application of TraNSIT to identify logistics infrastructure bottlenecks and opportunities Beyond sourcing the necessary capital investment to support the relatively high cost of new water- supported infrastructure, expanding agricultural development in the Darwin catchments presents at two further challenges. These are establishing new and viable markets for the produce generated by the development, either internationally or in distant mainland centres, cost- effectively accessing those markets via efficient supply chains (Ash and Gleeson 2014). Addressing the latter challenge requires an understanding of critical supply chain and infrastructure investment issues that can reduce barriers to market access and help to foster development opportunities. The CSIRO Transport Network Strategic Investment Tool (TraNSIT) performs a mass optimal routing of vehicle movements between thousands of enterprises, and scales up to provide industry-, domain- or locality-wide logistics costs (Higgins et al., 2013). This provides the ability to test infrastructure scenarios that can reduce logistics costs for thousands of enterprises. TraNSIT accounts for all the features and costs associated with the transport of a commodity across the road and rail network. TraNSIT has been set up for agriculture transport across Australia, with a high level of granularity for industries across northern Australia. It has been used to test the benefits of road upgrades and regulatory changes, and to calculate transport benefits of strategically (or optimally) located processing facilities. TraNSIT has been further developed and applied through initiatives in both the Northern Australia and Agriculture Competitiveness white papers. As part of the implementation of the Northern Australia White Paper, CSIRO is applying TraNSIT to inform the $100 million Northern Australia Beef Roads Fund (CSIRO, 2016) and maximise transport cost savings in beef supply chains across northern Australia. CSIRO is also extending TraNSIT to broader Australian agriculture transport as part of the Agricultural Competitiveness White Paper. It will compromise about 25 commodities – more than 95% of Australia’s agriculture transport volume – including grains, cotton, dairy, rice, sheep, poultry, pigs, sugar and horticulture crops. TraNSIT and related expertise will be used to inform the Assessment (agriculture and aquaculture activity, Chapter 8) with regard to the logistics implications and opportunities for the Darwin catchments case study, in the event of new irrigated agriculture. This will include: • producing scenarios of logistics costs and freight flows across the road network under different crop and crop–forage options, and likely resulting changes to market supply chains • comparing the baseline of existing freight flows and costs to identify bottlenecks across the supply chain networks; these may include high logistics costs from long transport distances, poor roads, storage issues, road flooding, lack of suitable processing and port limitations • identifying a range of infrastructure upgrades or new infrastructure that would reduce the costs of logistics for crop and crop–forage options in the Darwin catchments. This will include consultation with the territory departments of transport (or main roads), agriculture and industry. 10.5 Exploring stakeholder perspectives 10.5.1 POTENTIAL INVESTORS AND THEIR PERSPECTIVES ON IRRIGATION AND AQUACULTURE DEVELOPMENT A range of factors can affect a potential investor’s readiness to use water resources for irrigation and/or aquaculture, and these factors are not restricted to issues of economic viability. Issues affecting the transition to irrigated agriculture and aquaculture include appropriate technical knowledge and skills, compatibility with existing businesses, risk and uncertainty, access to markets, and access to labour and contractors. Transition to irrigation or aquaculture typically requires investors to make a significant up-front investment, to integrate new knowledge, and to operate (at least initially) under higher levels of uncertainty. Part of the new knowledge and uncertainty is likely to stem from the need to persevere through a period adaptive research and learning during which the crop types, varieties and agronomic practices required to maximise the likelihood of success are trialled; or existing technologies are adapted to regional circumstances. The Assessment will seek to increase understanding of the transition issues by exploring other potential investor perspectives on opportunities and barriers to investing in irrigation and aquaculture in the Darwin catchments. Potential investors may include: • pastoral companies • family-owned enterprises • smallholder Asian vegetable farmers • large agri-business investors • Indigenous corporations and enterprises • government as a potentially significant investor in infrastructure, regulation and oversight. The activity will conduct semi-structured interviews with representatives of potential investor categories in the catchments to explore issues affecting uptake of irrigation and aquaculture. Interviews may also be undertaken with territory government staff, representatives of regional economic development organisations, local government representatives and other industry stakeholders who are well placed to comment on these issues. 10.5.2 SCOPING THE SOCIAL LICENCE-TO-OPERATE AND THE WIDER AUDIENCE FOR INVESTMENT IN IRRIGATED FARMINGOR AQUACULTURE Potential investors in irrigation and/or aquaculture are only one set of stakeholders with interests in the management of natural resources in the Darwin catchments. The focus of this activity concerns the opportunity for the production of food and fibre through the provision of water resources, but other stakeholders derive other goods and services from the natural resource base and are therefore an important additional audience for Assessment findings and irrigation or aquaculture investment discussions. These stakeholders often place high value on the functional integrity of the water ecosystem – for example, to provide recreation, tourism, fishing, aesthetic appreciation, cultural and religious experiences, and aquatic habitat. One set of values (Indigenous values) will be examined in the Indigenous aspirations and water values activity (Chapter 11), but these values also need to be placed in the context of a wider set of social values. To provide this context and to refine understanding of the wider social licence-to- operate for investment in irrigation or aquaculture, this activity will identify the key stakeholder groups that have an interest in the natural resource management of the catchments – for example, industry groups, non-government organisations and government. Through desktop study and a set of interviews, the Assessment will document and map the interests, issues and concerns of each group that relate to water resources development, and highlight any potential development synergies or conflicts that may be relevant in each catchment and to the broader social licence-to-operate. 10.6 Understanding the policy and regulatory environment for the assessment of water resources, and development of water- dependent industries This task will involve: • documentation and analysis of current legislative and policy requirements, and regulations relevant to the assessment of water resources in the Darwin catchments (including water planning, water pricing, environmental, tenure and native title, land management, and other relevant domains) • documentation and analysis of current legislative and policy requirements relevant to the development of water resources in the catchment to support key water-dependent industries such as irrigated agriculture and aquaculture • highlighting of any cross-jurisdictional (national–state) legislative and policy requirements affecting the assessment and development of water resources • identification of key legislative and policy challenges to, and opportunities for, advancing the sustainable development of water resources. Additional legislative and policy analysis may be undertaken in response to client and stakeholder interests in particular jurisdictions, and key findings from other activities in the Assessment. Drawing on consultations undertaken by the wider Assessment team, these will be iteratively evaluated through the course of the Assessment. The research will be complemented by supporting analysis in the Indigenous aspirations and water values activity (Chapter 11), undertaken by the same researcher, which will focus on the specific interests of Indigenous people in water resource assessment and development. Research methods are: • desktop analysis of primary documents relating to case law, legislation, regulations and policy • a literature review and synthesis of scholarly interpretations of law and policy. 10.7 Regional socio-economic impacts Water resources development has socio-economic implications which will be shaped by the history and economic-demographic context of the Darwin catchments economies. Given this basic consideration, an exploratory and comprehensive economic and demographic baseline assessment for the Assessment area is key to better understand how the regional economy has evolved over the last decade, especially in comparison to what has occurred in other regions of northern Australia. Econometric modelling using counterfactual regions will be employed to analyse how the Darwin catchments have distinctively developed and how potential future water infrastructure development projects are likely to affect the catchments development pathway compared to other regions of Australia. The empirical methods involved in this regional economic development assessment will include: • Secondary data collection at different levels of aggregation: Statistical Area1 (SA1), Statistical Area 2 (SA2), Local Government Area (LGA) and Statistical Districts (ABS, 2013) • Assessment of trends and current levels of key economic, social and demographic variable • Empirical design of counterfactual regions using statistical methods (e.g. propensity score matching) or quantitative data • Comparison of levels and trends of key variable in the Darwin catchments to the counterfactual group • Econometric modelling designed to capture the effect of water infrastructure development in the Darwin catchments and comparison regions • Projections of potential economic and demographic impacts of future water infrastructure development using the econometric model outputs, numerical simulations and quantitative scenario analyses. With these empirical methods and steps the key question that will be explored is: What are the likely effects of water resources development, on different socio-economic indicators, such as direct employment, indirect employment and income, in the Darwin catchments? A core part of the analyses will be based on econometric models designed to analyse a set of key regional development indicators, which will be analysed considering scenarios projecting changes in a 30- year time frame. The analysis will control and examine how regional socio-economic projections could depend on a range of internal and external factors. Finally, all of the analyses to be conducted will be supported by software codes for replication and sensitivity analysis checks of a range of variables identified as having significant influence on model behaviour and results. 10.8 References ABS (2013) 1217.0.55.001 - Glossary of Statistical Geography Terminology, 2013, Australian Bureau of Statistics, Canberra http://www.abs.gov.au/AUSSTATS/abs@.nsf/Lookup/1217.0.55.001Main+Features12013?OpenDocument. Ash, A and Gleeson, T (2014) Northern Australia: Food and Fibre Supply Chain Synthesis Study. CSIRO/ABARES Australia. Higgins, A, Watson, I, Chilcott, C, Zhou, M, GarcÍa-Flores, R, Eady, S, McFallan, S, Prestwidge, D, and Laredo, L (2013) A framework for optimising capital investment and operations in livestock logistics. The Rangeland Journal 35, 181-91. 11 Indigenous aspirations and water values 11.1 Introduction The Indigenous aspirations and water values activity is a pre-feasibility analysis that will provide an overview of key Indigenous values, rights, interests and aspirations with respect to water and irrigated agricultural development in the Darwin catchments. This analysis is intended to assist, inform and underpin future discussions between developers and Indigenous people about particular developments, and their potential positive and negative effects on Indigenous populations. It is supported by and augments previous work by members of the Indigenous activity research team that has addressed Indigenous water issues from a range of perspectives. This includes: Indigenous water issues in irrigated agricultural development in Queensland as part of the Flinders and Gilbert Agricultural Resource Assessment (Barber 2013); Indigenous water values and water requirements in Northern Territory catchments to inform government water planning needs (Barber and Jackson 2011b, Jackson et al., 2011, Woodward et al., 2008); Indigenous hydrological and seasonal knowledge in the Fitzroy (WA) and Daly (NT) catchments (Woodward et al 2012) and in the Mitchell (Qld) catchment (Barber et al 2012); and Indigenous water issues with respect to mining in Western Australia (Barber and Jackson 2011a). Some of this work was undertaken by the Tropical Rivers and Coastal Knowledge (TRaCK) program, and a range of research from that program will inform the current study. The key questions that this activity will seek to address in the Darwin catchments include: • What is the existing documented information pertaining to Indigenous people in the Assessment area, and to Indigenous water and development issues more generally? This will emphasise: - the historical and contemporary context for Indigenous people - local Indigenous residence and tenure regimes - key issues in Indigenous water values, rights, interests and aspirations - key issues for Indigenous people regarding water and irrigated agricultural development. • How do current Indigenous Traditional Owners of, and residents in, the Darwin catchments perceive water resource assessment and development? • What potential issues regarding cultural heritage arise from water resource assessment and development? • What are the key legal and policy issues with respect to Indigenous people, water and irrigated agricultural development? • What are the barriers to, and enablers of, Indigenous people participating in water resource development? • What are the barriers to, and enablers of, Indigenous people deriving social and economic benefits from water resource development? 11.2 Linkages to other Assessment activities Components of the Indigenous aspirations and water values activity will have close connections with components of the socio-economics activity (Chapter 10) and the ecology activity (Chapter 12). The Indigenous aspirations and water values activity will support the socio- economics activity by providing specific Indigenous-related content to: • the broadscale characterisation of stakeholder values, as part of scoping the wider social licence- to-operate • identification of Indigenous aspirations relating to water resources and agricultural development, as part of the analysis of stakeholder perspectives • consideration of the enablers of, and barriers to, participation in water resource and agricultural development by potential small to medium-sized Indigenous investors, particularly local Indigenous corporations and prescribed bodies corporate. The Indigenous aspirations and water values activity will support the ecology activity through the collaborative identification of key natural and cultural assets of significance to Indigenous people. Indigenous knowledge will be crucial to understanding how particular assets have human ecological significance, and how they function as nodes in interconnected networks supporting biocultural diversity and sustainability. Connections with other activities will be identified as the study progresses. 11.3 Linkages to other research projects in the Darwin catchments A range of other research initiatives being undertaken in the focal catchments provide important context for the research undertaken through the Assessment. A key research initiative relevant to the Indigenous aspirations and water values activity is a CSIRO-funded project that focuses on the knowledge gaps in wetland ecology and the multiple values of water-dependent ecosystems in the Darwin catchments (Table 1-2). the Assessment will support greater understanding of ecological water requirements at locations that are important for Indigenous people, particularly for hunting, fishing and other cultural practices. The focus on multiple values will enable further exploration of Indigenous values, particularly as they relate to human ecology and cultural practices. This will complement and augment the orientation of the Indigenous aspirations and water values activity, which will encompass issues such as native title and cultural heritage; Indigenous principles for sustainable development; barriers to, and enablers of, participation; the distribution of benefits; and whole-of-catchment and intergenerational impacts. 11.4 Context and consultation The three regions being studied in the Assessment have substantial variations in conditions that are relevant to the Indigenous aspirations and water values activity, including variations in: • governance and tenure regimes • population size and demographics • levels of pre-existing development • existence of previous research • ongoing current and proposed future research. This variation in research context makes ongoing consultation with key stakeholders crucial to determining the exact scope of the Indigenous activity in the respective jurisdictions. Consultation is expected to continue throughout the conduct of the research and is likely to include stakeholders from: • the Australian Government • state and territory governments • local government • regional land councils • local Indigenous landholders and prescribed bodies corporate • sacred site and cultural heritage bodies • catchment management agencies • Indigenous development agencies • research agencies and researchers. 11.5 Scope Previous experience of the Assessment team with the Flinders and Gilbert Agricultural Resource Assessment suggests that consultations about Assessment scope and methods are likely to be iterative, and that maintaining some flexibility in Assessment scope is important in the initial planning stages. At the outset, it is intended that the research will not seek to directly enable or facilitate consensus with traditional owner groups about water and irrigation development in general, or specific development scenarios considered by the Assessment. Rather, the Assessment will focus on generating a representative set of Indigenous issues, perspectives and aspirations regarding water development that can be used as a guide and foundation for subsequent discussions between public and private developers, and Indigenous interests. The evaluation and refinement of development options undertaken by the other components of the Assessment provide further shared foundations for this process. Further refinements in Assessment scope, including jurisdiction-specific refinements, are expected to be made following further consultation. 11.6 Research ethics Before the fieldwork component of the Assessment begins, the research aims and proposed methods will be reviewed by the CSIRO Social Science Human Research Ethics Committee (CSSHREC). Assessment information sheets and a free, prior and informed consent form will be submitted for approval to the CSSHREC as part of the application. CSSHREC oversight will continue throughout the Assessment. It is expected that further consultation will be undertaken with Indigenous organisations in the catchment regarding the conduct of the research – in particular, with the Northern Land Council. This may result in a formal research agreement being negotiated between the Assessment team and local and/or regional Indigenous representative organisations. The agreement would be expected to reflect key elements of past research agreements between the Northern Land Council and the CSIRO researchers involved, updated and augmented to suit the circumstances of the Assessment and its intended outcomes. Participation in the Assessment will be entirely voluntary. Potential research participants will be provided with clear explanations of the research process and outcomes, through a combination of telephone, face-to-face and written contact, before they make any decision to participate. Wherever practicable, research participants will be afforded an extended period (of 1 month or more) after first contact by research staff to allow time for further consideration and consultation before making a decision to participate. During initial contact, the Assessment information sheets and the written consent form will be supplied. Verbal consent will then be sought, and confirmed through the participant signing the consent form. The forms are to be retained by CSIRO staff in a secure location. Based on experience of past projects, it is expected that comments that appear in any report will be identified using a general group identifier, rather than identifying individual participants. This retains anonymity but also provides a level of geographic specificity. 11.7 Methods 11.7.1 REVIEW OF EXISTING INFORMATION The review of existing documented information will encompass: • the historical and contemporary context for Indigenous people living in the Darwin catchments • local Indigenous residence and tenure regimes in the Darwin catchments • key issues in Indigenous water values, rights, interests and aspirations across northern Australia • key issues for Indigenous people regarding water and irrigated agricultural development across northern Australia. Relevant supporting data generated by other activities (e.g. ecological data with biocultural implications) will be integrated with the review. 11.7.2 FIELDWORK AND DIRECT CONSULTATION The fieldwork will emphasise direct consultation with Indigenous Traditional Owners of, and residents in, the focal catchments, through a combination of: • telephone and face-to-face interviews • group meetings and workshops • trips to key locations • other research methods developed in consultation with local and regional stakeholders. Common approaches and methods to the field consultations will be adopted wherever possible, but it is expected that local Indigenous organisations and individuals in each jurisdiction will influence the degree to which particular methods (e.g. individual interviews) are positioned with respect to other options (large workshops and groups). Participants will be identified through a ‘snowball’ method of iterative consultation with key local and regional Indigenous organisations – land councils, local group-based Indigenous corporations, and catchment management agencies. The objectives and intended methods of the research will be explained, copies of Assessment information and consent forms will be provided, and further direction will be taken about people and organisations who should be contacted in the preliminary scoping and identification stage of the Assessment. Key organisations and individuals nominated by these Indigenous organisations will then be approached for further consultation during planned field trips. 11.7.3 CULTURAL HERITAGE ASSESSMENT Cultural heritage assessments will contain both general and catchment-specific elements. The general elements will encompass: • broadscale cultural heritage values and issues, with an emphasis on archaeology and material culture • methods required to adequately assess cultural heritage values and issues in any future water resource development • categories and types of impacts on cultural heritage associated with water and irrigated agricultural resource development The jurisdictional and catchment-specific elements will include: • cultural heritage obligations on water resource developers • preliminary desktop analysis of known and potential cultural heritage issues relating to key water storage options identified through the Assessment’s research activities. Further cultural heritage matters are expected to be identified through additional consultation with local Indigenous organisations and cultural heritage specialists. 11.7.4 LEGAL AND POLICY ANALYSIS This component of the activity will involve: • documentation and analysis of current legislative and policy requirements relevant to the inclusion of Indigenous interests in the assessment of water resources in the Darwin catchments • documentation and analysis of current legislative and policy requirements relevant to the inclusion of Indigenous interests in the development of water resources in the Darwin catchments (including water rights, cultural heritage and native title) • highlighting of any cross-jurisdictional (national–state or state–state) legislative and policy requirements affecting Indigenous interests in the assessment and development of water resources • identification of key legislative and policy challenges to, and opportunities for, recognising and valuing Indigenous interests associated with water resources. The analysis will focus on rights and interests recognised in the Australian legal system, while acknowledging that Indigenous customary laws and cultural understandings of water that are not currently recognised by the Australian legal system are important considerations, which will be addressed in this activity and/or key findings from other activities in the Assessment. Drawing on consultations undertaken by the wider Assessment team, these issues will be iteratively evaluated through the course of the Assessment. The research will be complemented by supporting analysis in the socio-economics activity (Chapter 10), focusing on the broader legislative and policy environment relevant to water resource assessment and development. The two primary research methods adopted are: • desktop analysis of primary documents relating to legislation, regulations and policy • a literature review and synthesis of scholarly interpretations of law and policy. 11.8 Data analysis and preliminary dissemination The data from the literature review and interviews will be iteratively analysed using NVivo qualitative analytical software (QSR International, 2016) to identify major themes and key findings. Key information and research participant comments from the interviews will be identified, extracted, and then formally checked with the respective research participants to ensure that they are an accurate reflection of their views, and can be used in further analysis and public presentation. The resulting information and analysis will then be combined into a draft research report. The draft report will be disseminated to local Indigenous research participants and key Indigenous stakeholders for further comment, correction and confirmation. It will be augmented by further presentations to group meetings. The resulting feedback will be incorporated into a revised draft report, which will then be subjected to scientific peer review and further community comment before it is finalised. It is expected that undertaking concurrent field research across the three jurisdictions will yield a range of contrasts and comparative insights. Where such an approach will be beneficial, additional data, findings, issues and analysis from other jurisdictions may be referred to in the analysis of the local jurisdiction. A further component of final reporting will identify a range of matters (values, rights, interests, aspirations, themes, issues, principles, strategies, pathways, etc.) that are common across jurisdictions. Such cross-jurisdictional analyses will be further augmented by comparison with findings from the National Environmental Science Programme and other CSIRO research activities in the same jurisdictions. 11.9 References Barber M (2013) Indigenous water values, rights and interests in the Flinders and Gilbert catchments. A technical report to the Australian Government from the CSIRO Flinders and Gilbert Agricultural Resource Assessment, part of the North Queensland Irrigated Agriculture Strategy CSIRO Water for a Healthy Country and Sustainable Agriculture flagships, Australia, Canberra. Barber M and Jackson S (2011a) Indigenous water values and water planning in the upper Roper River, Northern Territory. CSIRO, Darwin. Barber M and Jackson S (2011b) Water and Indigenous people in the Pilbara, Western Australia: a preliminary study. CSIRO, Darwin. Barber M and Jackson S (2012) Indigenous water management and water planning in the upper Roper River, Northern Territory: history and implications for contemporary water planning. CSIRO, Darwin. Barber M, Shellberg J, Jackson S, and Sinnamon V (2012) Working knowledge: local ecological and hydrological knowledge about the flooded forest country of Oriners Station, Cape York. CSIRO, Darwin. Jackson S, Finn M, Woodward E, Featherston P (2011) Indigenous socio-economic values and river flows. CSIRO and TRaCK, Darwin. QSR International (2016) QSR InternationalTM NVivoTM software. Viewed 14 June 2016, http://www.qsrinternational.com/product. Woodward E, Jackson S, Finn M, and Marfurra McTaggart P (2012) Utilising Indigenous seasonal knowledge to understand indigenous aquatic resource use and inform water resource management. Ecological Management and Restoration 13. Woodward E, Jackson S, and Straton A, (2008) Water resources of the Howard River region, Northern Territory: A report on the social and cultural values and a stakeholder assessment of water use scenarios CSIRO Sustainable Ecosystems, Darwin. Part IV Trade-offs: How can development opportunities be maximised by understanding and quantifying trade-offs with existing industries and ecosystems? 12 Ecology The ecology activity seeks to assess the potential for impacts on freshwater and marine ecosystems from possible changes to flow regimes due to new infrastructure across the Assessment area. The focus is on water-related ecosystems because water developments, particularly irrigation, can result in substantial changes to streamflow and water quality, yet they typically occupy only a small proportion of the landscape (<1% of a catchment). Impacts on terrestrial systems will be dealt with in an issues paper, which will be prepared to identify future assessment priorities within the Assessment areas. The ecology activity analysis is limited to understanding the impact of changes in flow on the ecological system, with potential changes in land use and nutrient enrichment characterised qualitatively. The key questions that this activity seeks to address in the Darwin catchments include: • What terrestrial, freshwater and marine assets are found within the Darwin catchments? • What freshwater and marine assets are of key significance? • In freshwater and marine systems, what is the current understanding of the dependencies of assets on flow? • In freshwater and marine systems, what is the current understanding of the other system drivers that can affect assets? • In freshwater and marine systems, what are the potential impacts of infrastructure development on assets? • Using qualitative and quantitative analysis methods, what are the predicted impacts on freshwater and marine systems? • What is the certainty of the predicted outcomes? Northern Australia has diverse and highly valued natural ecosystems, which provide commercial, recreational and cultural value, as well as maintaining ecological functions and habitats for plants and animals (Abel and Rolfe, 2009). The ecology of aquatic ecosystems is fundamentally linked to the seasonality of the wet–dry tropical climate that regulates the flow regime, as well as the landscape it drains (Warfe et al., 2011; Aquatic Ecosystems Task Group, 2012). The natural flow regime and connectivity between aquatic ecosystems are critical for sustaining freshwater and marine biodiversity, and natural ecological processes. Many of northern Australia’s aquatic systems are in a near-natural condition; historical and current use of, and change in, surface and ground waters has typically been localised. The freshwater ecosystems in northern Australia have a rich biodiversity, supporting at least 170 fish species, 150 waterbird species, 30 aquatic to semi-aquatic reptiles, more than 60 amphibian species and more than 100 macroinvertebrate families (Van Dam et al., 2008). The estuaries of northern Australia also support a rich biodiversity, and are critical in supporting productive fisheries, where the greater freshwater inflow to estuaries results in increased fisheries production (Aquatic Ecosystems Task Group, 2012). Catchment flows also support high-value commercial and recreational marine fisheries, such as the prawns, barramundi, mudcrab and a suite of other species. These systems also provide species of high economic and cultural significance to Indigenous Australians. Several species of conservation significance, such as dugongs, sea turtles and a variety of sharks and rays, as well as habitats such as mangrove forests and seagrass beds, are also dependent on coastal and estuarine systems, and the support of catchment flows and nutrient inputs. 12.1 Region overview The four river basins in the Darwin catchments have numerous and extensive seasonal wetlands and floodplain systems. There are five wetlands of national significance: the Finniss Floodplain and Fog Bay Systems (Finniss River), the Port of Darwin (Finniss River), the Adelaide River Floodplain System (Adelaide River), the Mary Floodplain System (Mary River) and Kakadu National Park (Wildman section) (Environment Australia, 2001). Only Kakadu National Park is Ramsar listed, and the greater park area extends into the neighbouring Alligator River catchments. A large proportion of the Darwin catchments is protected within national parks and reserve systems (DLRM, 2015). In addition to Kakadu National Park, protected areas within the Darwin catchments include Litchfield National Park (Finniss River), Djukbinj National Park (Adelaide River) and Mary River National Park (Mary River). The inland aquatic ecosystems within these catchments form a complex mosaic of habitats, supporting rich biodiversity. The richness of the biodiversity is attributed to the integrity, extent and heterogeneity of the wetland habitats throughout the catchments (Environment Australia, 2001; DLRM, 2015). The Adelaide River Floodplain and Mary Floodplain systems are recognised as important breeding areas for waterbirds and crocodiles, and are among the most important breeding sites for magpie geese in Australia (SKM, 2009). The Darwin catchments also have a high dependence on groundwater. Spring-fed monsoon vine forests of high conservation value in the Wildman catchment are vulnerable to increased use of groundwater resources. Many permanent waterholes through the river systems are also in part replenished by groundwater, with the waterholes creating refugia in the dry season. Riparian zones, which support high biodiversity and productivity, are sensitive to changes in both surface water and groundwater regimes (Pusey and Kennard, 2009). Significant fauna in the catchments include saltwater and freshwater crocodiles, the dwarf sawfish (listed as vulnerable), the freshwater sawfish (vulnerable) and the northern river shark (endangered) (Cth Environment Protection and Biodiversity Conservation Act 1999). The ecology of many of these species is highly dependent on the quality and quantity of water resources, and maintenance of habitat heterogeneity. Many species in the catchments (such as the northern river and speartooth sharks and sawfish) have a coastal component to their life cycle; other marine species depend on the productivity engendered by the annual wet season. Some of these, such as banana prawns, barramundi and mudcrabs, have life histories that span freshwater through estuarine to marine environments, and are valuable across commercial, recreational and Indigenous fishery sectors (Bayliss et al., 2014). 12.2 Ecology activity breakdown The ecology activity has three major tasks: 1) identification and prioritisation of assets; 2) conceptual model development to characterise flow–ecology relationships and the potential impacts of infrastructure development; and 3) analysis of potential impacts using quantitative and qualitative methods, and development of an assessment workflow. The asset and conceptual model methods are described jointly for marine and freshwater systems (activities 1 and 2), whereas the analysis methods (activity 3) are described separately. The activity will draw on stakeholders from Australian, state and territory government agencies; universities and other research providers; and industry bodies. Stakeholders will provide input and review to each of the steps, and be engaged through workshops and one-on-one meetings. This activity is highly reliant on the surface and groundwater hydrology activities for flow inputs, the Earth observation activity for remote sensing of waterholes and water-dependent vegetation communities, the Indigenous water values activity for cultural assets and the land suitability activity for areas of potential development. 12.2.1 IDENTIFICATION AND PRIORITISATION OF ASSETS A review and prioritisation of assets will be undertaken for the Assessment area. For the purposes of the Assessment, assets are defined as: • listed threatened, vulnerable or endangered species or communities • wetlands, species or communities that are formally recognised in international agreements • providing vital habitat for water-dependent flora and fauna • near-natural, rare or unique habitats for water-dependent flora and fauna • supporting significant biodiversity for water-dependent flora and fauna • recreational, commercial and cultural value. Defined assets must be water-dependent – that is, they must have some level of dependency on groundwater or surface water flows, resulting in either periodic or sustained inundation. Species and community assets will be identified based on their conservation status in Commonwealth, state and territory listings. Formally recognised Ramsar sites will also be included as assets. Reviews of existing asset datasets such as state and territory data, commercial and recreational fisheries data, spatial data and species occurrence data will be used. For Indigenous assets, this activity will rely on the Indigenous aspirations and water values activity (Chapter 11). Habitat-based assets can include wetlands, rivers, karst and other groundwater-dependent ecosystems, saltmarshes, estuaries and coastal areas (BMT WBM, 2010). Key habitats such as mangroves, near-shore coral reefs and seagrass beds, and salt flats may also be included, and used as target or indicator species of marine ecosystem health. To determine the significance of an asset that is not formally recognised in an international or national agreement, the High Ecological Value Aquatic Ecosystems Criteria will be used (BMT WBM, 2010): • Diversity – the aquatic ecosystem exhibits exceptional diversity of species, habitats, or geomorphological features or processes • Distinctiveness – the aquatic ecosystem is rare or threatened; the habitat supports rare or threatened species or communities • Vital habitat – an aquatic ecosystem provides vital habitat for flora and fauna species • Naturalness – the aquatic ecosystem is near-natural and not adversely impacted by existing threats • Representativeness – the aquatic ecosystem is an outstanding example of an aquatic ecosystem type. A threat and value matrix will be used to prioritise assets. Information for prioritising assets will be obtained from existing spatial databases, literature information and expert input. The outcome will be the development of an asset map, identifying places of significance and ranges of species of significance. 12.2.2 CONCEPTUAL MODELS Conceptual models will be developed based on the best available information. They provide a representation of the biophysical drivers of function within any given system. Conceptual models will also guide the subsequent analysis step. Conceptual models can take many forms, including maps and diagrams, such as causal box and arrow diagrams; this activity will use causal box and arrow diagrams. These conceptual models will synthesise the best available knowledge of flow–ecology relationships for freshwater and marine assets, and identify potential risks from other sources, including water resource development and agricultural intensification. Other risks include physical changes, (e.g. increases in sedimentation) and water quality changes (e.g. increased nutrients), as well as invasive species (spread of pests and weeds). For the subsequent analyses, the ecology activity will be limited to linking an ecological response to a flow, with some considerations of water quality as a modifying factor. The conceptual models will underpin qualitative, semi-quantitative and quantitative analyses of the impacts of water extraction and flow regime change on selected marine and coastal assets, such as habitats, species and natural resource production values (e.g. coastal and offshore fisheries). The conceptual models will highlight systems’ lateral and longitudinal connectivity. They will incorporate knowledge about the life history of critical species and habitats, and ecological flow relationships such as threshold effects, where such information exists. To understand the potential for impacts of infrastructure on assets, a synthesis of existing knowledge will be undertaken, documenting the potential effects that infrastructure can have on physical systems (e.g. connectivity and migration; flow triggers and thresholds; sediment, nutrient fluxes and productivity; salinity; exacerbation of pests and weeds), and how this affects assets. The development of conceptual models involves engagement with Commonwealth, state and territory government agencies; the National Environmental Science Programme (NESP) Northern Hub and NESP Water Quality Hub; industry stakeholder groups; and other research providers. 12.2.3 ANALYSIS OF POTENTIAL IMPACTS ON ASSETS For marine and freshwater assets that are of high priority, the relationship between flows and assets will be analysed. A multiple-lines-of-evidence approach will be used for the analysis. Where information is poor, the environmental assessment will be qualitative. Where information is rich, the assessment will be semi-quantitative or quantitative. Analysis methods are described separately below for marine and freshwater assets. Marine systems The analysis of potential impacts on marine assets will focus on species and habitats of significance, and natural resource production values (e.g. coastal and offshore fisheries). Water requirements will be defined for fisheries, conservation assets, and other marine–coastal assets in estuarine and marine habitats. These will encompass end-of-system flow volume and timing as they affect key ecological processes and habitats, and life histories of selected species. Where relevant, water quality (e.g. nutrient and sediment load) requirements will also be defined. Water requirements will be defined through a literature review, from expert elicitation in workshops, and from existing ecological models For selected assets, the analysis will seek to develop ecological flow thresholds and corresponding preference curves and response indicators. The focus will vary between case study areas, depending on available ecosystem knowledge and data, need and time. The analysis will use qualitative methods developed for the Assessment, methods from Bayliss et al. (2014), and the semi-quantitative Productivity Sustainability Assessments method (Thorburn and Morgan, 2005a), which has been applied to the entire Northern Prawn Fishery (Griffiths et al., 2006). For selected species (e.g. banana prawns, barramundi, threadfin salmon, black bream, mudcrabs) in commercial, recreational and Indigenous fisheries that have sufficient time series data, catch– effort–flow empirical models will be developed to assess case study experiments. These models use statistical methods developed in Bayliss et al. (2014) for the Gulf assessment of ecological and fisheries values. Freshwater systems The analysis of potential impacts on freshwater assets will focus on species and habitats of significance, and potential for changes across the landscape. The analysis will entail collation of existing data and knowledge, and a synthesis of literature that describes the condition, drivers of change and flow requirements of assets. The synthesis will seek to identify baseline asset data; key surface water and groundwater requirements of species; existing flow–ecology relationships, including environmental water requirements; water quality as modifiers of outcomes; and information on other non–flow related system drivers. Scientific studies, existing management plans and expert knowledge will be used as inputs to the analysis. An uncertainty analysis will be conducted to explore the sensitivity of models to input information, metric selection, combination method and flow scenarios. Species of significance To evaluate the potential for change in significant species, a population condition approach is being used, which relates the life history requirements of a species to a flow regime. Population condition reflects a range of aspects of relevance to the species, such as the population size, the age structure of the population and the health of individuals. This approach is similar to the Ecological Elements method (IUCN, 2011) applied in the Murray–Darling Basin. The method uses environmental water requirements to define the relationship between specific flow attributes – such as waterhole depth; and flow magnitude, duration and timing – and species population condition. Environmental water requirements will be defined as thresholds. Preference curves will be used to represent trajectories of population condition change, to reflect how the sequencing of flow events in a flow regime can impact on a species’ life history. The approach can consider species requirements across river reaches, floodplains, waterholes or wetlands as habitats, provided that the habitat is represented in the hydrologic and inundation models. The outcome is a set of environmental water requirements that enable an assessment of change in a species’ population condition based on a modelled flow, which is represented as an assessment workflow (a set of automated steps) that can be used for any assessment scenario. Habitats of significance To evaluate the potential for change in significant habitats, a habitat condition approach is being used. This uses a similar approach to that used for species of significance, drawing on the Ecological Elements method (Overton et al., 2014). Habitat condition can be applied to a single habitat type (a waterhole) or to a mosaic of habitats (floodplain wetland mosaic), where the habitat quality and extent support a diversity of species and processes, such as foraging or breeding. The method uses environmental water requirements to define the relationship between specific flow attributes – such as waterhole depth; and flow magnitude, duration and timing – and habitat condition. Preference curves are used to represent trajectories of habitat condition change, to reflect how the sequencing of flow events in a flow regime can impact on habitat extent and quality. Water quality will be considered in the analysis as a potential modifier of outcomes. The approach can consider water requirements for river reaches, floodplains, waterholes or wetlands as habitats, provided that the habitat is represented in the hydrologic and inundation models. The outcome is a set of environmental water requirements that enable an assessment of change in habitat condition based on a modelled flow, which is represented as an assessment workflow (a set of automated steps) that can be used for any assessment scenario. Landscape diversity To evaluate the potential for change in landscapes, a landscape diversity approach is being used. This approach will assess the potential for change across assets of different types (e.g. wetlands, floodplains, river channels, waterholes). The approach will be at a landscape scale, to explore how changes in flow may impact on the diversity of habitats and processes across the landscape. The method will focus on the use of flow metrics that have ecological significance, which will be evaluated at each gauging point in the river system models. The flow metrics will include commence to flow; length of inter-wet and dry periods; frequency, seasonality and duration of inundation; and depth. Changes in flow metrics from the modelled scenarios will be compared with the ‘baseline’ hydrologic model. This approach enables an assessment of potential hydrologic change across the landscape. The outcome is a set of flow metrics and flow thresholds that enable an assessment of landscape diversity based on modelled flows. This will be represented as part of an assessment workflow (a set of automated steps) that can be used for any assessment scenario. 12.3 References Abel N and Rolfe J (2009) 12 Public and private conservation of aquatic systems in northern Australia: threats and opportunities. Northern Australia Land and Water Science Review. Northern Australia Land and Water Taskforce, Canberra. Aquatic Ecosystems Task Group (2012) Aquatic ecosystems toolkit. Module 3: Guidelines for identifying high ecological value aquatic ecosystems (HEVAE). Australian Government Department of Sustainability, Environment, Water, Population and Communities, Canberra. Bayliss P, Buckworth R and Dichmont C (2014) Assessing the water needs of fisheries and ecological values in the Gulf of Carpentaria. Final report prepared for the Queensland Department of Natural Resources and Mines. CSIRO, Australia. BMT WBM (2010) Ecological character description for Kakadu National Park Ramsar site. A report prepared for the Australian Government Department of Sustainability, Environment, Water, Population and Communities, Canberra. Department of Land Resource Management (2015) Darwin coastal – bioregional description. Northern Territory Government. Viewed December 2015, http://lrm.nt.gov.au/plants-and- animals/herbarium/nature/bioregional/darwincoastal. Environment Australia (2001) A Directory of Important Wetlands in Australia, 3rd edn, Environment Australia, Canberra. Griffiths S, Kenyon R, Bulman C, Dowdney J, Williams A, Sporcic M and Fuller M (2006) Ecological risk assessment for the effects of fishing: report for the Northern Prawn Fishery. A report for the Australian Fisheries Management Authority, Canberra. IUCN (2011) The IUCN Red List of Threatened Species, version 2011.2. IUCN Species Survival Commission, Gland, Switzerland, and Cambridge, United Kingdom. Viewed April 2016, http://www.iucnredlist.org. Overton, I. C., C. Pollino, J. Roberts, J. R. W. Reid, N. R. Bond, H. M. McGinness, B. Gawne, Stratford D.S., L. E. Merrin, D. Barma, S. M. Cuddy, D. L. Nielsen, T. Smith, B. L. Henderson, D. S. Baldwin, G. S. Chiu and T. M. Doody (2014) Development of the Murray-Darling Basin Plan SDL Adjustment Ecological Elements Method. Report prepared for the Murray-Darling Basin Authority. CSIRO, Canberra. Pusey, B. J. and M. J. Kennard (2009) 03 Aquatic ecosystems in northern Australia. Northern Australia Land and Water Science Review. Canberra, Northern Australia Land and Water Taskforce. SKM (2009) Ecological assets of northern Australia study. Sinclair Knight Merz, Canberra. Thorburn DC and Morgan DL (2005a) Threatened fishes of the world: Glyphis sp. C (Carcharhinidae). Environmental Biology of Fishes 73, 140. Van Dam RA, Bartolo R and Bayliss P (2008) Chapter 2 – Identification of ecological assets, pressures and threats. In: Bartolo R, Bayliss P and Van Dam (eds) Ecological risk assessment for Australia’s northern tropical rivers. Land and Water Australia, Canberra. Warfe DM, Pettit NE, Davies PM, Pusey BJ, Hamilton SK, Kennard MJ, Townsend SA, Bayliss P, Ward DP, Douglas MM, Burford MA, Finn M, Bunn SE and Halliday IA (2011) The ‘wet–dry’ in the wet–dry tropics drives river ecosystem structure and processes in northern Australia. Freshwater Biology 56, 2169–2195. Part V Case study experiments, reports, key protocols and standards 13 Case study experiments 13.1 Rationale By its nature, the Assessment will produce information from a very broad range of disciplines. This is as it should be because the development of northern Australia will require the integration of knowledge from a similarly broad range of disciplines. The purpose of the case study experiments is to help the reader: • understand how to ‘put the Assessment information together’ to answer their own questions about water resource development in the Darwin catchments • understand the type and likely scale of opportunity of different types of water resource development in selected geographic parts of the Assessment area • explore some of the nuances associated with ‘greenfield’ developments in the Darwin catchments, which are often difficult to capture in discipline-based information. 13.1.1 WHAT THE CASE STUDIES ARE DESIGNED TO DO AND NOT DO Although the case study experiments are designed to be realistic representations – that is, they will be ’located‘ in specific parts of the Assessment area, and use specific water and land resources, and realistic intensification options – they are illustrative only. They are not designed to demonstrate, recommend or promote particular development opportunities being proposed by individual development proponents, nor are they CSIRO’s recommendations on how development in the Darwin catchments should unfold. 13.2 Proposed case study framings The specific case study experiments will be developed over time as the Assessment proceeds, and the Assessment team begins to assemble information from a range of sources and disciplines. These sources will include ideas generated by stakeholders within the Assessment area and will be guided by current enterprise types found in the Assessment area. Proposed case study experiments will then be tested with the Program Steering Committee before being finalised. Although the case study experiments are yet to be developed, the following framings can be used as a guide: 1. large schemes, privately funded (greater than about $500 million) 2. large schemes, publicly funded (greater than about $500 million) 3. medium-size schemes, such as might be developed by pastoral corporates or large family businesses (from tens of millions up of dollars to about $500 million) 4. small-scale schemes, such as an individual or family business might develop (from about $1 million to about $10 million). Coupled to this framing, various methods of water capture and supply will be considered. These will include surface water capture (water harvesting, instream dams), groundwater and managed aquifer recharge. Various agricultural and aquaculture systems will be included in the case study experiments. Some of these systems will be relatively straightforward, whereas others will include integrated crop–livestock systems, (possibly) agriculture and aquaculture systems, industrial-scale production of commodities and energy, and industrial-scale by-product systems. The case study experiments will consider the costs and challenges of supplying water using various configurations of capture and distribution, the timing of water supply and demand, assumptions concerning annual water yield (at varying reliabilities), and conveyance and field application losses. Soil and landscape attributes will be considered in terms of their proximity to the water supply, risks such as secondary salinity, and suitability for various crop and aquaculture types. Interactions and links with other industries (e.g. the beef industry), processing facilities, transport logistics, and hard and community infrastructure requirements will be factored in. The case studies will be grounded within the social and cultural context of the Darwin catchments, including infrastructure availability and constraints. Economic analyses will consider gross margins, as well as the ability of the enterprise to service capital costs. Ecological changes will be assessed using Ecologic Elements modelling (Chapter 12), and a range of trade-offs will be articulated. 13.2.1 EXAMPLE CASE STUDY EXPERIMENTS One way to envisage what these case studies will look like is to consider those used in previous land and water resource studies in the Flinders catchment (Petheram et al., 2013a) and the Gilbert catchment (Petheram et al., 2013b). The analysis undertaken by Petheram et al. (2013a, b) will be further refined as part of the Darwin Water Resource Assessment. In the Flinders catchment, three case studies were developed: 1. An instream dam of approximately 250 GL storage with an annual yield of 40 GL at 85% reliability. This water would be used to grow sorghum grain to supply a newly established local feedlot and abattoir. 2. An offstream storage of approximately 130 GL with an annual yield of 34 GL at 85% reliability. This water would be used to irrigate rice. 3. Small-scale offstream water harvesting at a number of sites throughout the catchment. These were combined to review the opportunity and impact from five different catchment entitlements, from 80 to 560 GL. Crop areas of about 500 ha each were used to grow a range of high-value crops, such as mungbean, cotton and rice. In the Gilbert catchment, three case studies were similarly developed: 1. An instream dam of approximately 230 GL storage with an annual yield of 172 GL at 85% reliability. This water would be used to grow a rotation of cotton, peanuts and irrigated fodder, with a new cotton gin located in Georgetown. 2. Two new instream dams with a combined capacity of 725 GL and an annual yield of 498 GL at 85% reliability. This water would be used to grow sugarcane, which would be supplied to a newly established sugar mill. 3. Raising of the dam wall of a small existing dam by 2 m to provide storage of 20 GL and an annual yield of 17 GL at 85% reliability. This water would be used to grow irrigated fodder, to provide a supplementary feed source to the local beef industry and hay for use in drought years. 14 Reports, products, protocols and standards 14.1 Reports, products and protocols The Assessment management team will provide quality assurance for all data and reports produced from the Northern Australia Water Resource Assessment. To meet this objective, the team will: • provide templates, standards, processes and workflows for reporting • provide collaborative working spaces (including SharePoint, Google Drive) • review all technical material • ensure that sensitive and important modelling is undertaken within a best modelling practice framework – that is, a three-stage independent review process of: i) conceptual model, ii) calibration model, and iii) simulation model • edit and produce catchment and summary reports • develop processes, and provide information sheets and training to Assessment members on data management protocols, the CSIRO metadata catalogue, and the CSIRO data access portal and data audit trails. SharePoint is a website that provides a central storage for the Assessment team to share documents that require version control. The Assessment team will store all versions of the catchment and summary report documents on the CSIRO SharePoint. All final versions of the technical reports will be stored on the SharePoint https://teams.csiro.au/units/NAWRA/_layouts/15/start.aspx#/SitePages/Home.aspx. A Google Drive site has been established for the Assessment team. This will be the team’s primary collaboration space to share non-sensitive documents, calendars, photos, meeting minutes, guideline documents and videos, stakeholder contact information and other similar material. Table 1-1 details key deliverables for the Darwin Water Resource Assessment. These will be complemented by a minimum of 15 technical reports, which will provide the technical underpinning for the summary material. Alternative (non-report) products for the delivery of information to key stakeholders will be investigated during the course of the Assessment and in consutation with stakeholders. For example, the CSIRO data access portal was used as the final repository for the land suitability grids in the Flinders and Gilbert Agricultural Resource Assessment. 14.2 Standards The Assessment management team will define editorial standards to guide authors in reporting the findings of the Assessment. These standards will include map and figure conventions, and will be available in a document titled Reporting standards, which will be available to the Assessment team on Google Drive at https://drive.google.com/drive/folders/0BxAzeqgsYoOrdXc1VVFnaWZwYkE. These standards are based on: • the Australian Government Style manual for authors, editors and printers • CSIRO brand identity guidelines • standards used in the Flinders and Gilbert Agricultural Resource Assessment, and the Sustainable Yield projects • the Australian Oxford dictionary. Since many specialist terms are not found in these resources, additional conventions specific to the Assessment will be developed in consultation with the Assessment team. Reporting standards will be a ‘living document’ that changes as the Assessment progresses, to document decisions on language and formatting. Conventions specified in early drafts, however, will not be changed unless absolutely necessary; the aim is to add conventions, not to backtrack on earlier decisions. Reporting standards will be published as a report at the end of the Assessment. CONTACT US t 1300 363 400 +61 3 9545 2176 e csiroenquiries@csiro.au w www.csiro.au AT CSIRO, WE DO THE EXTRAORDINARY EVERY DAY We innovate for tomorrow and help improve today – for our customers, all Australians and the world. Our innovations contribute billions of dollars to the Australian economy every year. As the largest patent holder in the nation, our vast wealth of intellectual property has led to more than 150 spin-off companies. With more than 5,000 experts and a burning desire to get things done, we are Australia’s catalyst for innovation. CSIRO. WE IMAGINE. WE COLLABORATE. WE INNOVATE. FOR FURTHER INFORMATION Land and Water Dr Peter Stone t +61 7 3833 5659 e peter.stone@csiro.au w www.csiro.au/en/research/LWF Land and Water Dr Cuan Petheram t +61 2 6246 5987 e cuan.petheram@csiro.au w www.csiro.au/en/research/LWF Agriculture Dr Ian Watson t +61 7 4753 8606 e ian.watson@csiro.au w www.csiro.au/en/research/AF Land and Water Dr Chris Chilcott t +61 8 8944 8422 e chris.chilcott@csiro.au w www.csiro.au/en/research/LWF