Salinity - Risk and Hazard - Western Australia
Western Australia
Where is there a dryland salinity risk or hazard and how was this determined?
The assessment was restricted to the south-west of Western Australia where dryland salinity is widespread. All analysis was based on groundwater depth and trend and the risk of shallow watertables was derived from these two attributes. As dryland salinity is caused by shallow watertables, the risk of salinity is inferred from the risk of shallow watertables. Not all shallow watertables will be saline. Estimates and projected risk areas are based on analysis of existing groundwater levels and trends at a scale of 1:250 000 based on soil systems mapping. There are limitations in the resulting assessment due to gaps in data.
The Natural Resource Assessment Group (NRAG) of Agriculture WA has mapped the south-west areas of WA using a hierarchical landscape mapping system. The mapping process is based on geology and geomorphology. In WA most groundwater movement and process is controlled by the geology and regolith properties. Therefore, this system was seen as appropriate for attributing hydrologic properties. The maps were examined by members of the Catchment Hydrology Group (CHG), to ensure that there were no major conflicts between the map units and the current understanding of groundwater processes.
This mapping provided the spatial units to attribute groundwater depth and trend and also provided a basis for spatial analysis. The largest scale for which the mapping occurred ranged between 1:50,000 and 1:100,000. This represents a regional scale, and is not suitable for sufficient to extrapolating data to the catchment or farm level, ie to scales of 1:10,000 to 1:25,000.
Risk of shallow watertable maps were constructed using groundwater depth and trend data. The risk categories applied were:
| High Risk | water levels <2m |
| water levels 2-5m and rising | |
| Moderate Risk | water levels <2m and falling |
| water levels 2-5m and flat or falling | |
| water levels 5-10m and rising | |
| water levels >10m and rising | |
| Low Risk | water levels 5-10m and flat or falling |
| water levels >10m and flat | |
| No Risk | water levels >10m and falling. |
Risk of shallow watertables in year 2000 was based on both trend and depth analyses. Predicted groundwater depths for the years 2020 and 2050 were determined by projecting trends at current rates of rise. An assumption is that land use and rainfall do not change.
Constraints and issues raised from applying the assessment method
The following constraints and issues were identified:
- Groundwater data is not evenly distributed through the State. Therefore whilst 90% of zones area have some bore data, there are a significant 'gaps' in some areas, particularly the northern and central-eastern areas.
- Due to the lack of data some systems have no allocated trend or risk. In some cases risk may have been allocated based on the experience of regional hydrologists.
- For reporting at a national, state or regional scale there can be poor data distribution both within and between catchments and regions. There may be bias in the data as some bores will have been sited for specific purposes such as looking at the response of certain treatments. However, getting the initial data quality controls done by Agriculture WA hydrologists with good local knowledge has reduced this bias.
- The scale of reporting necessarily required some generalisations to be made. Whilst systems are mapped at between 1:50,000 and 1:100,000 scale, there is still significant variation within a unit, particularly with respect to landforms.
- For existing bores on data bases there can be some uncertainty regarding well construction. It may not be clear that recorded water levels represent the watertable level or piezometric surface. Bore construction details and drill logs are required to definitively assign the data to a watertable.
- Analysis of trend data has an average length of record of 5.5 years, with a substantial skew to less than five years. Therefore some data may not represent long-term trends.
- There can be temporal differences within the entire data set. Length of record is a significant determinant of the reliability of calculated watertable elevation trends. In many areas there will be few records longer than ten years. Longer records become crucial as rainfall decreases and watertable accessions become increasingly episodic.
- Current status of salinity has been based on areas mapped as wet and waterlogged, meaning the profile is saturated for longer than six months, (ie. it is not purely a winter waterlogging situation). This does not necessarily imply that the site is saline. However, due to high levels of salt stored in soil profiles, it is highly likely that if these are not presently saline, they are at risk of becoming so.
- Estimates of hydrological risk for the years 2020 and 2050 have been based both on linear extrapolation of measured rates of rise and hydrological opinion where data are lacking.
- Determining trends in hydrograph data depends on the frequency of observation. Automatically logged wells can show daily response. On a catchment and larger scale, annual response can generally be determined from four observations at three monthly intervals. This however would require between 10 and 20 years of data to verify a robust trend in some highly responsive catchments.
How can dryland salinity risk change over time?
The assessment was restricted to the south-west of Western Australia where dryland salinity is widespread. All analysis was based on groundwater depth and trend and the risk of shallow watertables was derived from these two attributes. As dryland salinity is caused by shallow watertables, the risk of salinity is inferred from the risk of shallow watertables. Not all shallow watertables will be saline. Estimates and projected risk areas are based on analysis of existing groundwater levels and trends at a scale of 1:250 000 based on soil systems mapping. There are limitations in the resulting assessment due to gaps in data.
The Year 2000 map of dryland salinity risks represents both depth and trend of groundwater at year 2000 within broad regional systems. Trend data were then used to predict future watertable depths and to allocate a risk rating in 2020 and 2050. The future risk of a shallow watertable is based on depth to groundwater. Predicted groundwater depths for the years 2020 and 2050 were determined by projecting trends at current rates of rise. An assumption is that land use and rainfall do not change.
Rainfall analysis by the Bureau of Meteorology, Western Australia, has shown that over much of the agricultural regions total winter rainfall (May-October) has decreased, whilst trends over the warmer months show a smaller increase. Annual totals have generally been declining between 5 to 30mm per decade over the agricultural regions between Geraldton and Albany in the 1910 to 1997 period, (Bureau of Meteorology, 2000).
Rainfall analysis suggests that the gradient of the rising groundwater trend represents the minimum gradient expected based on rainfall alone and disregarding any change in land use. Should the average rainfall increase, the gradient might also increase and accelerate the impacts of salinity.
Risk has been allocated to entire systems based on the most frequently occurring depth and trend time series data for bores. In each system there will be variability that will change the risk at the local scale. The current extent map gives some indication of the proportion of each system at risk in 2000. However there are no estimates of the actual extent predicted for 2020 and 2050 beyond the system scale.
In 2000 the risk is predominantly in the eastern wheatbelt in valley floors and adjacent areas (the northern and southern zones of ancient drainage respectively). Eastern sections of the northern wheatbelt also exhibit high risk.
There are some coastal areas at high risk around Bunbury and the southern systems (Donnybrook Sunkland) and within Scott River system. The latter has shallow watertables and is therefore categorised as high risk however groundwaters are relatively fresh (200mS/m) in this region.
Most areas at low risk are adjacent to the coast in the Northern Perth Basin and coastal systems around Albany and on the south western edge the Leeuwin Zone. Further inland, low risk areas include the Donnybrook Sunkland, the Stirling Ranges and at Salmon Gums north of Esperance. Low risk areas are related to deep watertables and/or watertables with no rising trend.
The remaining areas of the State have been classified as having a moderate risk (excluding those areas with insufficient data). There will be local variability in the extent and impacts of shallow groundwater in these regions.
In total, approximately 16% of the south west areas of WA have the potential for salinity in 2000 due to shallow watertables. Of the area at risk, 81% is occupied by agricultural land and the remainder by perennial vegetation.
The changes in areas at risk in 2020 are reasonably subtle in the eastern wheatbelt where high risk areas expand from the valley floors to secondary tributaries and lower slopes. Other soil-landscape systems and zones that move into a high risk category occur in the higher rainfall such as the Zone of Rejuvenated Drainage and areas west and to the north-east of Esperance.
In total, approximately 20% of the south west area of WA has the potential for salinity in 2020 due to shallow watertables. Of the area at risk, 80% is occupied by agricultural land and the remainder by perennial vegetation.
In 2050 high risk has expanded in the eastern wheatbelt to include all valleys and lower slopes. In these zones this will represent approximately 30% of the landscape.
High risk is also apparent in the majority of zones located east of the Darling Scarp. An exception is the vegetated Western Darling Range zone which has a moderate risk.
Risk has also substantially increased in the region around Esperance and west of Albany.
In total, approximately 33% of the agricultural areas has the potential for salinity in 2050 due to shallow watertables. Of the area at risk, 73% is occupied by agricultural land.
Further information
- Western Australia Dryland Salinity Assessment 2000 report
- Australian Dryland Salinity Assessment 2000 report
- National Technical Overview Report of the State-based dryland salinity assessments
- Australian Groundwater Flow Systems Report
- Agriculture WA
- National Dryland Salinity Program
- National Action Plan for Salinity and Water Quality
Western Australian Dryland Salinity Assessment 2000
Link to Map maker to make a map using this information.
PDF files
Some documents on this website are available as PDF files. Adobe Acrobat Reader is required to view PDF files.
Key
Links to an another web site
Opens a pop-up window