|December 2006||National Research Flagship||Water For A Healthy Country|
The Ring of Confidence - Incorporating
Uncertainty into Sediment Transportation Models
Damage to the near shore regions of the GBR due to declining water quality from terrestrial run-off has become an issue of national concern. Estimates of sediment loads entering the GBR from river catchments are used to inform government bodies about the health of the GBR and trigger management actions when the sediment load is too high.
The estimates generated by modelling often give little or no indication of how reliable they are, which makes it difficult to establish the likelihood of particular loads occurring at particular locations. Without incorporating uncertainty measures to address this issue, it is difficult for resource managers and policy makers to determine what actions, if any, need to be taken.
Upper Burdekin sediment loads
Dr Petra Kuhnert of CSIRO Mathematical and Information Sciences has been taking part in a pilot study on the Upper Burdekin River subcatchment, near Townsville. The pilot is part of the Reef Catchment Futures program in the Water for a Healthy Country Flagship. If the pilot is successful, her techniques will be applied across the whole of the Burdekin catchment.
The Upper Burdekin subcatchment is one of four that comprise the Burdekin River catchment, Queensland's second largest. The subcatchment covers about 128,000km2, much of which is cattle grazing country. Peak river flows are between December and April with monsoonal rains.
Using data on sediment loads in the Upper Burdekin, Petra is investigating better ways of dealing with uncertainty in the load estimates produced using sediment transportation modelling software.
SedNet, for example, is a commonly-used software package for modelling sediment transportation through a river network. It makes use of hillslope, gully and stream bank erosion deposits at each link in the river network. Petra's research aims to understand how uncertainty measures could be incorporated into these three types of erosion and propagated throughout the entire catchment from river source to mouth.
More reliable predictions
A key feature considered for implementation with these types of models is the standard errors on the end-of-catchment load estimates and those estimates calculated further up stream. Standard errors indicate the size of the variation and, therefore, the degree of uncertainty in the load estimate. The bigger the standard error, the less certain the estimate is.
Standard errors are important in determining the accuracy of the load estimate derived from the model as they can identify specific links in the network where the uncertainty is appreciably large and therefore cannot be regarded as reliable. Significantly, this can tell where in the network extra data may be required to reduce the level of uncertainty.
Preliminary research is looking at ways of incorporating uncertainty into one source of input, hillslope erosion and methods for incorporating uncertainty into the RUSLE equation. This equation multiplies six factors together to derive a value for hillslope erosion in terms of tonnes of soil lost per hectare of land per year. It is based on a statistical approximation of the uncertainty around the derived measure at each watershed.
This information will be used to assess error through the SedNet model and so provide resource managers and policy makers with more informative sediment load estimates for confident decision-making.
This research was presented at the 'Australian Statistical Conference/ New Zealand Statistical Association Conference 2006' held in Auckland in July 2006. A report will be written for the Water for a Healthy Country Flagship early 2007.
IN THIS EDITION:
the Flagship's new Science Director
one of our
The Water for a Healthy Country Flagship is a CSIRO initiative and part of the National Research Flagships program that aims to deliver scientific solutions to advance Australia's most important national objectives. One of the largest scientific initiatives ever mounted in Australia, it aligns closely with the Federal Government's National Research Priorities. The initiative brings together our national research resources to deliver breakthroughs in fields ranging from healthcare to light metals and the environment.
About this email
This E-Newsletter uses an HTML-rich media format to provide an easy-to-follow, visually attractive layout. If for any reason, your computer does not support HTML format email, please let us know by emailing email@example.com with your full name and email address, and we will ensure you receive our E-Newsletter in a plain-text format.
Please feel free to forward it to those in your organisation who would be interested in receiving future newsletters.
If you do not wish to receive this electronic newsletter from the Water for a Healthy Country Flagship, please reply to this message with the word "unsubscribe" in the subject line. Alternatively, please contact firstname.lastname@example.org or telephone us on +61 2 6246 4565 to request the removal of your email address.
This E-Newsletter is a publication of CSIRO. It does not purport to be comprehensive or to render professional advice. All material in this publication is subject to copyright. For permission to reproduce any part or all of an article, please contact the editor.
a Healthy Country
Editor: Leane Regan
CSIRO Black Mountain Laboratories,
Canberra, ACT, 2601
Phone: +61 2 6246 4565
Fax: +61 2 6246 4564