Chapter 18 Land-cover Classification from Landsat Imagery for Mapping Dynamic Wet and Saline Soils

Abstract

Wet and saline soils have been recognized as an important and complex component of wetland ecosystems in arid environments. Analysis and classification of remotely sensed spectral data is an effective method for discerning the spatial and temporal variability of soils. The East Shore Area (ESA) of the Great Salt Lake soil survey update is focused on updating soil map units containing wet and saline soils. The ESA provides a unique environment for the use of remotely sensed spectral data for map unit refinement because of low relief and a large extent of soils that are wet and saline to various degrees. Map units in the ESA containing wet and saline soils were updated and refined using Landsat 7 imagery. Five land-cover classes are related to dominant soil types that vary in soil wetness, salinity, calcium carbonate concentration and vegetation cover type. Supervised classification of the imagery was performed using the five land cover classes. The final classification resulted in 14 land cover classes, including nine additional classes that help describe the variability in the original five classes. The classification results were validated using visual inspection in the field, a priori knowledge of the area and an error matrix. The results of the classification were used to enhance original soil map units and calculate map unit composition in the final soil mapping process. This information was then incorporated into the updated soil map. Temporal variation in land cover classes has the potential to be considered in map-unit refinement to reflect the dynamic nature of the margins of the Great Salt Lake, Utah.