Includes the aims of the project and a description of the four climate and water resource development scenarios considered and modelled.


Overview of project methods. A report to the Australian Government from the CSIRO Murray-Darling Basin Sustainable Yields Project (CSIRO Research Publications Repository)


The project's aims were to:

  1. define major catchments for the Murray-Darling Basin (MDB) based on surface water catchments and existing river system models
  2. assess, catchment by catchment, the severity of drought for the last five, seven and ten years
  3. define four different scenarios of climate and water resource development for modelling current and future water availability
  4. model the surface runoff and groundwater recharge patterns for each defined catchment under each defined scenario
  5. assess the implications of the changes in runoff, recharge and water resource development, on water availability in each defined catchment under each defined scenario using the river system and groundwater models used by state agencies for water resource planning (and other models where necessary)
  6. assess the implications of potential changes in water availability on diversion volumes (and their reliability), environmental flows and on the likely ability to meet the stated aims of current water sharing plans, assuming current water sharing arrangements
  7. report results progressively – catchment by catchment, and finally for the entire MDB.

Scenarios considered

The four scenarios of climate and water resource development that were considered are broadly defined as follows:

  1. Current development and historic climate as used in current water sharing plans and associated models. This is generally the full available historic hydrologic record (typically 1895–2006) and the water resource development (land use, farm dams, infrastructure, diversion/abstraction levels) described by recent data sets. This is the baseline scenario against which other scenarios are compared.
  2. Current development and recent climate to assess the consequences of a continuation of the climate of the last ten years. Development conditions are as for baseline. The climate series are generated by using the average climate conditions for the last ten years to adjust historic daily climate records. Climate variability is considered by statistically generating 100 climate series and analysing three of these: a low-end, the middle and a high-end scenario from the full range.
  3. Current development and future climate to assess the consequences of climate change. This uses predictions for the climate in 2030 from global climate models from the 4th Assessment Report of the Intergovernmental Panel on Climate Change. Multiple model projections are considered for low, medium and high levels of global warming.
  4. Future development and future climate to assess the combined consequences of climate change and changes in land/water use. The future climate scenarios are as described for (3) above. The considerations of future development vary region by region, and are limited by current predictive ability and the time constraints of the project. Two aspects of catchment development that affect system inflows and can be reasonably modelled are the expansion of farm dams and plantation forestry. These aspects of development will be considered in collaboration with State agencies and with reference to current State policies. Detailed assessments of future increases in bushfire frequency and the resultant impacts on catchment water yield cannot be reliably made. However, a synoptic assessment of the likely magnitude and locations of bushfire impacts on catchment water yield will be made. Growth in groundwater extraction up to the currently imposed extraction limits will also be considered in the development scenario. The many other aspects that may affect future water availability and use, for example:
    • changed water use by native vegetation under altered atmospheric carbon dioxide concentrations
    • changes in on-farm and off-farm water use efficiency
    • water trading
    • new water infrastructure
    • are currently impossible to reliably predict and are thus beyond the scope of the project.


Each of the four scenarios were described by 112 year sequences of modelled daily climate data.

For the historic climate, historic data were used with gaps in the records filled using statistical modelling. For the recent climate scenario these data were statistically generated based on historic data.

For future climate results for high, medium and low emission runs of the global climate models used in the 4th Assessment Report of the Intergovernmental Panel on Climate Change (IPCC) were analysed and used to estimate future seasonal rainfalls, average temperatures and potential evapotranspiration.

A regional climate model was also run to estimate future statistical distributions of daily rainfall, season by season. The results from the global and regional climate models were then used to adjust the historic climate series to provide simulated future climate series that adhere to the historic sequencing of climate. Maintaining historic sequencing information enables sensible comparisons between scenarios.

The climate data were used to model daily surface runoff on 5 km x 5 km grids across each catchment, and to model daily groundwater recharge at point locations for separate aquifers. The climate and runoff data were used to adjust the inputs to river system models (that include models of agricultural and urban water demand) to determine:

  • daily river flows
  • reservoir levels
  • diversions to water users in each catchment for each scenario assuming current water sharing arrangements.

Recharge data were input to groundwater system models to determine water table levels for major aquifers for each scenario.

For minor aquifers simpler assessments were made of likely changes in water levels without undertaking numerical modelling.

Assessments were made of the exchanges between rivers and underlying aquifers. Where these exchanges were large, estimates of flow loss/gain were passed from the river model to the groundwater model, and from the groundwater model back to rerun the river model to adjust the modelled flow losses/gains.

A statistical assessment of the recent drought was made catchment by catchment and for the whole MDB based on rainfall and river flow observations for the previous five, seven and ten years compared to the full historical instrumental record.

An assessment of the uncertainty associated with the results was undertaken by considering the uncertainty of future climate and future water resource development and the uncertainty associated with the climate and hydrologic models used. Independent assessments of catchment water balances were made using remotely sensed estimates of evapotranspiration; typically, 90 per cent of rainfall is evaporated or transpired by plants (forests, grassland and crops), while only around 10 per cent becomes runoff. These estimates were compared to observed and modelled runoff and river flow.

Based on the range of modelling, water availability across each region and under each scenario is reported. This included reporting a range of measures of reservoir storage volumes, water diversion reliability and groundwater use. Measures were reported that indicate the potential for environmental impacts associated with current and future surface and groundwater availability.

Do business with us to help your organisation thrive

We partner with small and large companies, government and industry in Australia and around the world.

Contact us

Your contact details

First name must be filled in

We'll need to know what you want to contact us about so we can give you an answer.