|
|
|
|
|
|
|
|
|
| CSIRO | SOLVE | Issue 10 | FEB 07 | |
|
| ||||
|
ARTICLE
MATHEMATICAL MODELLING/ ENVIRONMENTAL MANAGEMENT:
Error Factor Hurdle for Reef Management By Jason Major
Resource managers use computer models to estimate how much sediment enters the waters around the Great Barrier Reef, but there is no standard error factor attached.Along the Great Barrier Reef (GBR) land-use changes have increased the sediment loads of water running into the near-shore lagoons, seriously affecting the reef ecosystem. Compounding this is a poor understanding of the ecological processes that produce the sediment, compromising efforts by resource managers responsible for reef health. Resource managers use computer models to estimate trends in sediment load, but the number provided by the model that corresponds to the amount of sediment has no standard error attached. That is, no one knows how right or wrong that number may be – a critical deficiency for resource managers making investment and environmental management decisions that affect reef health, and for adjacent industries such as agriculture, mining, fishing and tourism. To fix this modelling flaw Dr Petra Kuhnert, a research statistician with CSIRO Mathematical and Information Sciences, is developing tools to quantify the standard error associated with sediment load calculations. These tools will be used to improve the accuracy of existing models and associated environmental management decisions. Dr Kuhnert’s research is part of CSIRO’s Reef Catchment Futures program in the Water for a Healthy Country Flagship. She is using the Upper Burdekin sub-catchment in Queensland as part of a pilot study. A 2003 Productivity Commission report into land use and water quality in the GBR region says that agriculture, which makes up 80 per cent of land use in the GBR catchments, is one of the largest contributors of sediment and other pollutants into the inner reef; but mining, tourism, fishing and urban development are also, to varying extents, responsible for or affected by the threat from increasing sediment, which in excessive amounts can affect coral, seagrass and other marine ecosystems. But to take economically acceptable action the affected industries need reliable information.
Much of their information on sediment transportation comes from a model called SEDNET, which uses hill-slope, gully and stream-bank erosion inputs at each link in a river network. However, Dr Kuhnert says the model is missing an allowance for error or uncertainty in these inputs, which means sediment output calculations could be questionable. For scientists, Dr Kuhnert’s research should identify the most critical areas – those where the degree of error is greatest – and intensify monitoring that will feed back into the model and reduce uncertainty in the results. But a second significant outcome will be within a catchment’s communities and industries on whose shoulders the responsibility to reduce sediment loads falls. Although it sounds paradoxical, it is thought that if statistics can identify, quantify and acknowledge the uncertainties in the model, it will be easier to gain community and industry support for any management action. Rachel Eberhard, a consultant managing the Reef Water Quality Partnership, says one of the biggest difficulties with SEDNET is communicating to local people how the model works. Because they know intuitively that much of the data is potentially unreliable, they are justified in being sceptical of the results. She says Dr Kuhnert’s research will allow resource managers, when engaging with the community, to be transparent about the uncertainty in the modelling, helping to build trust between communities, managers and authorities. “And because we will have identified the levels of uncertainty in the data, it will actually provide a more solid basis from which to develop a community-accepted management plan to reduce sediment load.” For further information contact: |
 IN THIS ISSUE
|
| Home | About Us | eSubscribe | Links | |
|
|
||
|
|
|
Back to main