Soil salinity detection and monitoring using Landsat data: a case study from Kot Addu, Pakistan

Abstract

Despite decades of research in soil mapping, characterizing the spatial variability of soil salinity across certain area remains a crucial challenge. This research illustrates the potential use of Landsat 8 satellite reflectance data (30 m resolution) to facilitate soil salinity mapping. For this purpose, soil samples from surface soils of Kot Addu, Muzaffargarh District in Pakistan, were collected in the 3rd week of January 2017 (January 20, 2017, to January 27, 2017). Multi-temporal data (atmospherically corrected Landsat 8 images from the United States Geological Survey (USGS) website; January 19, 2017) and soil sample electrical conductivity (ECe) were obtained and processed for various indices’ calculation and supervised classification. Soil sample ECe and pixel index values were compared for correlation using linear regression. Normalized Differential Vegetation Index (NDVI) showed a coefficient of determination (R2) of 54.98%, Soil Adjusted Vegetation Index (SAVI) has R2 = 54.98%, Mosaic Simple Ratio (MSR) has R2 = 47.59%, and Moisture Stress Index (MSI) indicated R2 of 51.48%. These indices were stacked, and supervised classification of stacked images was performed to produce 3 salinity classes ( 10 dS m−1). The maximum area of Kot Addu falls at < 4 dS m−1 class, while in other two classes small patches of soil salinity were observed which were distributed across the map. It was concluded that remote sensing technology can be used for mapping the spatial distribution of the soil salinity. The results suggest that Landsat 8 image data with information of soil ECs can be used to identify spatial variations of soil salinity in arid areas.