Published by Australian Academy of Science
Monitoring the white death – soil salinity Box 2 | Mapping salinity

Testing remote-sensing techniques

Scientists may peer at satellite images or process them using high-powered computers, but the only way to assess their accuracy is to go out into the field and measure the salinity at ground level.

A recent study by scientists at CSIRO Mathematical and Information Sciences tested a remote sensing technique in three study areas in Western Australia. They analysed a series of Landsat images, which they combined with information on contours, the location of roads and farm boundaries, and farm management histories. They then compared the results of these analyses with the locations of known salt-affected and changing sites, as supplied by farmers, field officers from Agriculture Western Australia and from previous salinity mapping exercises.

Results were very encouraging. At one study site, salt-affected land was mapped remotely at an accuracy of almost 100 per cent. Accuracy was lower at other sites, but refinement of the techniques will continue to improve results.

Predicting where salinisation will occur next

Scientists have shown that a number of factors determine the vulnerability of sites to salinisation. These include:

• the position of a site within a landscape – generally the lower it is, the more likely it is that the water table will reach the surface and cause salinisation;
• soil type;
• management – such as the extent of clearing;
• rainfall.

Combining information on these and other factors could allow the prediction of sites vulnerable to the saline menace. This is where a geographic information system (GIS) can play a role. GIS is a computer application that involves the storage, analysis, retrieval and display of data that are described in terms of their geographic location. The most familiar type of spatial data is a map – GIS is really a way of storing map information electronically.

A GIS has a number of advantages over old-style maps, though. One is that because the data are stored electronically they can be analysed readily by computer. In the case of salinity, scientists can use data on rainfall, topography, soil type – indeed, any spatial information that is available electronically – to first determine the combinations most susceptible to salinisation, and then to predict similar regions that may be at risk.

Much information is already in a form that can be used in a GIS, and more is being added continually – including that produced by Landsat. As the databases and prediction techniques improve, farmers and land management agencies will be better placed to wage an assault on salt.

Related sites

• Airborne geophysics – a tool for salinity assessment control (Agriculture Western Australia)

• Predicting areas at risk from salinity (CSIRO Mathematical and Information Sciences)

• Australian developments in airborne electromagnetics – from minerals to dryland salinity (Australian Academy of Technological Sciences and Engineering)

Other boxes

Box 1. Salinisation – causes and prevention