Spatial Modeling of Soil Salinity for Impact Assessment of Subsurface Drainage System

Abstract

Knowledge of spatial variability of the soil and water properties of the subsurtace drained field is essential to analyze and evaluate the spatial impact of a subsurface drainage system. The present study was carried out to develop the spatial variability maps of soil salinity using geostatistical approach and evaluate the impact of subsurface drainage system on soil salinity in an area ofabout 1000 ha in Western Yamuna Canal Command in Gohana sub-division of Sonepat District of Haryana Soil salinity datasets were subjected to exploratory data analysis using the GS+@ and Geostatistical module of ArcGIS"'. The semivariogram analysis revealed that the spherical semivariogram model fitted well to describe the spatial variability of soil salinity as evident from high coefficient of determination, R2, (0.957<R2<O.994) and least residual sum of squares, RSS, (0.003<RSS<0.006). The developed semivariogram models indicated strong spatial structure and low nugget values as an indicator of better model prediction. Moreover, the cross validation results and prediction error statistics showed that the mean error and the mean reduced error were close to zero, the mean reduced variances were observed to be within acceptable range of I ± 0.20, and the mean absolute error and the root mean square error were low for all the models. Subsequently, the kriged maps for soil salinity were generated using ordinary kriging. It was observed from spatial analysis of the kriged maps that the salt affected area decreased from 56% to 35% after introduction of subsurface drainage. Moreover, out of35% of salt affected area only 1.73% lands were strongly saline and the remaining land was moderately saline after drainage.