Comparison of spectrum indices for mapping soil salinity in saline lands of Chezan plain (Markazi province)

Abstract

Soil salinity phenomena are one of the main problems of arid and semi-arid lands. Saline soils constitute a huge part of Iran, and also threaten its neighboring lands. Therefore, in order to optimum exploitation of such soils, qualitative monitoring is necessary. Recently, remote sensing techniques have been increasingly applied in monitoring soil characteristics. The present study was carried out between 2014 and 2015, with the purpose of using remote sensing for mapping soil salinity in the saline rangelands of Chezan plain (Markazi province). In the first step, 50 soil samples were taken from the topsoil (30 cm depth) and their Electrical Conductivity (EC) was measured by EC- meter. To use the soil salinity map using remote sensing, we first used Indian Remote sensing Satellites (IRS) satellite imagery and the satellite's LISS sensor (LISS III, 2008). After geometric and radiometric correction, this image was been classified using the Maximum Likelihood method. Then in the next step, Comparison of Spectrum indices were done by extracting maps of soil salinity. For this purpose, Four indices including: Brightness Index (BI), Salinity Index (SI1), SI2, and Normalized Difference Salinity Index (NDSI) were used. Among these indices, SI2 had the most correlation with ground control points (0.63 in 1% level) and is introduced as a more suitable index than others for zoning soil salinity. Regarding to the salinity map, the results showed that a sizeable portion of the study area was classified class 2 with a salinity of between 4-8 dS/m (55% of whole land).