Soil wetness as an indicator of stream salinity: a landscape position index approach

Abstract

Dryland salinity is a major environmental issue in much of Australia’s agricultural lands and is expressed as salt affected land or degraded stream water quality. Maps showing areas at risk of land and stream degradation are needed by regional, state and national management and planning authorities, as well as farmers. Part of the management involves establishing end-of-valley targets for water quality. Developing maps of salinity risk is limited by the availability of appropriate spatial data. Elevation data at appropriate scales are available for all Australia. This paper explores the potential to develop catchment and regional scale soil wetness maps, based only on elevation data, as a surrogate for stream salinity risk. Soil wetness indices were derived through the Fuzzy Landscape Analysis GIS (FLAG) model. While FLAG avoids the direct use of process models it uses process understanding. It does this through an index-based approach that requires a training set of areas of wetness, salinity or other attribute associated with position in the landscape. We test whether only one of the FLAG landscape position indices (UPNESS), that has been shown to characterise depositional zones, is correlated to baseflow stream salinity. UPNESS is the uphill area monotonically above each point in the landscape, and is a specialised form of contributing area, a measure of surface and sub-surface water accumulation. This measure characterises catchments as the ‘connectedness’ of drainage or prevalence of runoff sinks. It distinguishes, for example, between freely drained catchments and those with more extensive depositional zones that allow the accumulation and storage of salt and formation of preferential pathways in the system. FLAG analysis was applied over an area of ~12 000 km 2 in southeastern Australia where salinity research was being conducted by state agencies. Stream electrical conductivity measurements were obtained to compare with the UPNESS index. The results suggest that the model is useful for targeting further investigations in regional scale salinity management planning and research. FLAG is suggested as a first step for obtaining a highly visual rapid assessment of potential wetness, discharge and salinisation at catchment scales.