Relating soil electrical conductivity to remote sensing and other soil properties for assessing soil salinity in northeast Thailand

Mapping and monitoring of soil salinity is required to establish its areal extent and also to keep track of changes in salinity in order to formulate appropriate and timely management strategies for reclamation and rehabilitation of such soils. Remote sensing data have been increasingly used in soil-salinity studies as they are not only quicker but are also useful for making realistic predictions. A study was conducted in northeast Thailand to understand the relationship of spectral reflectance and physico-chemical soil proper-ties to electrical conductivity (EC) by using remote sensing data (Landsat (R) ETM+) and laboratory analysis of soil sample data. Multiple regression analysis was used to examine the relationships between EC and spectral/soil properties and to generate several models. In the case of spectral properties, mid-infrared band (Landsat (R) band 7) and near-infrared (band 4) were found to be most correlated with the observed EC values of the surface layer of the soil. For the soil properties, chloride (Cl), sodium (Na), phosphorus (P), and sodium adsorption ratio (SAR) were found to be most correlated with observed EC values. Similarly, for the subsurface layer, Cl and P were found to be significant predictors of observed EC values. Copyright (c) 2006 John Wiley & Sons, Ltd.