ABSTRACT
Salinity is one of the major constraints for successful crop production. However, there are few studies in which the effect of climate variables on soil salinity dynamics has been studied, using geospatial techniques. The salt-affected areas in parts of south-western Punjab (Bathinda, Faridkot, Mansa and Muktsar districts) were classified using Landsat satellite data during the pre-monsoon, monsoon and post-monsoon seasons of the years 2003, 2008 and 2013. Normalized Difference Salinity Index (NDSI) was used for automatic extraction of salinity data from the satellite image. In general, salinity was higher during pre-monsoon seasons than in monsoon and post-monsoon seasons. The salt-affected area was highest during the pre-monsoon season of 2003. The overall accuracy of salt-affected areas was 85%, with a Kappa coefficient of 0.35. The values of NDSI, rainfall and temperature, corresponding to ground truth points, were extracted from the raster images and the relationships among these variables indicated that NDSI was significantly negatively correlated with rainfall but significantly positively correlated with temperature. The integration of maximum temperature and rainfall maps with salinity during May 2003 indicated higher salinity in the areas with elevated temperatures and lower precipitation. The results of this study suggested that integration of salinity with climate variables and other factors (like slope, depth to ground water level, etc.) in geographic information system (GIS) may help to develop a decision support system, which may be useful for policy makers and scientists in planning strategies to manage salt-affected areas and soil resources.
