Spatial Soil Salinity Assessment by Using Principle Component Analysis and Geospatial Techniques in Central Punjab, Pakistan

Abstract

Both natural and man-made soil salinity is a chief geological disaster in semi-arid and arid parts. In cultivated land, it has a negative impact on plant development and harvests, while in semi-arid and arid non-agricultural zones, due to subsidence, corrosion and groundwater quality, it affects urban structures, leading to additional soil erosion and land deprivation. The study was conducted at central Punjab, Pakistan with the aim to develop a baseline and to show the precision and accuracy of Geographic Information System (GIS) technology for delineating soil salinity in no data region. The samples of soil were gathered at deepness of 0-15 and 15-30 cm, and three factors (pH, Electrical Conductivity, and Sodium Adsorption Ratio) were analyzed in the laboratory. Landsat 8 OLI imagery were used for salinity indices development. A statistical index association was found between soil salinity noted in soil samples of field and 13 GIS-based salinity indices. The effect importance and model parameters for various soil salinity indices were assessed using regression model fitting. The data were divided into 3 categories: i) ground or field data, ii) brightness/intensity indicators, and iii) salinity indicators. Data were analyzed using principal component analysis (PCA). The results indicated that salinity indices were favorably related with ground data sets, but brightness/intensity indices had no significant relationship with ground or field data.