Acid sulphate soil mapping with hyperspectral imagery at South Yunderup, Western Australia

Abstract

One of Australia?s National Research Priorities for achieving an ?Environmentally Sustainable Australia? is tackling the problem of acid sulfate soils (ASS) which affect many of our most desirable lands for urban and agricultural development. Fitzpatrick et al. (1998) stated that currently ?There is no consistent standard for mapping their [sic ASS] extent or severity.? To help address this problem a national atlas on ASS has been established which has been incorporated into the ?Australian Soil Resource Information System? website. However, these current GIS map products are based on interpolation of isolated ground sample points and the interpretation of remote sensing data that is not optimum for detection of ASS related mineralogy. In addition, these map products, especially for the large area assessments such as the Murray-Darling Basin, are not suitable for monitoring purposes except for those (few) sites where ground data is being routinely collected. Hyperspectral sensing technologies from both aircraft and satellites have the potential to provide accurate mapping and monitoring of ASS through its ability to measure often subtle but diagnostic compositional information of surface materials (Clark & Roush, 1984; Clark et al., 1990a; 1990b). This includes the measurement of the abundances, compositions and crystallinity of the specific minerals, including those associated with ASS such as jarosite, goethite, hematite, kaolinite and gypsum. Typical ASS minerals include iron oxyhydroxides (goethite), oxyhydroxyl-sulphates (jarosite) and sulphates minerals (barite and gypsum) (Fanning et al., 2002). Fitzpatrick & Self (1997) found that minerals found in acid sulphate soil environments are similar to those analogous to acid mine drainage. These environments consisted of ferrihydrite, schwertmannite, goethite, lepidocrocite and jarosite. Although there have been a number of studies of the application of hyperspectral for mapping and monitoring acid conditions associated with mined environments (e.g. Swayze et al., 2000; Crowley et al., 2003 and Ong et al., 2003) there has been little work on its use for mapping acid sulphate soils (e.g. Lau et al., 2006) despite its potential value for large area (catchment-scale) mapping and monitoring of ASS. The main aim of this study is to assess whether airborne hyperspectral imagery can be a useful tool for mapping surface mineralogy potentially associated with ASS in coastal areas undergoing urban development. A collaborative project was established in 2007 between the Department of Environment and Conservation (DEC) and CSIRO to investigate the ability of the airborne hyperspectral imagery to help identify potential ?hot-spot? sites of ASS contamination in the South Yunderup area of the Peel region of WA.