The Jiadhal basin in northeast India suffers from serious soil erosion and landslides. During the monsoon, the eroded soil travels with the flood waters and gets dumped on the fertile agricultural lands downstream. This sandy soil has very little productive value, and the farmers have to wait an average of 6–7 years before they can grow paddy on it. The study uses satellite images and a predictive model ‘E30’ to find out the severity of erosion and the loss of agricultural lands from the period of 2000 to 2014. Findings have shown that for 4 years of analysis, ‘severe’ and ‘very severe’ erosions were taking place from 17 and 5 % of the total basin area. Landcover analysis showed that agricultural lands have been lost during the flood years. Part of these lands have been converted to sand-filled areas. The study proposes the idea that these sand-filled areas can be adequately utilized for growing a grass ‘Chrysopogon zizanioides’ native to India, which can be used for checking the erosion at the first place. In this way, the hazard can be mitigated by identifying erosion-prone areas from the model output and planting the grass. By utilizing only 17.5 % of the total basin area, checking of sand deposition by this method can be possible in a period of 1 year. The method has the potential to be replicated in all the basins in the region suffering a similar problem.