Estimating soil salinity from remote sensing data in coastal districts of Nghe An province

Abstract

    Soil salinity is one of the serious threats to the environment, which has a negative impact on crop yieldsin agriculture. Estimation of soil salinity from remote-sensing data is a practical approach for longtermmonitoring quality of land, which assists the land management and environmental sustainability.This study presents the use of Landsat-8 OLI data which received on December 18 2018 to extractphysical indicators to estimate the spatial variation of salinity of land in coastal districts and towns ofNghe An province. Univariate and multivariate linear regression models were performed by usingelectrical conductivity (EC) of land from the field survey between December 25, 2018 and Januar 8,2018. The correlation between different indicators and field data on soil salinity is calculated to findhigh correlation indexes. The optimal regression model is selected when considering the maximum R2value and the smallest RMSE. The results show that the multivariate linear regression model has highaccuracy with the coefficient of determination (R2=0.67), and the root mean square error(RMSE=1.19). This suggests that remote-sensing data can be used effectively to model and map spatialvariations in soil salinity in coastal areas.