Seasonal Variability in Soil Moisture Retrieval from Risat-1 Satellite Images by Using Emperical Model

The present study was undertaken to retrieve surface soil moisture from RISAT-1 satellite data and comparing it with actual ground truth observations by using Topps model. RISAT-1 is India's first indigenously developed C band SAR satellite. The groundtruth observations were carried out synchronized with satellite passes for two different location viz. Muthur and Nambiyur and for both Wet and Dry seasons. Two different locations are chosen as representative for Western Agro Climatic zone of Tamil Nadu. There is good correlation obtained between predicted and observed estimates. The soil moisture correlated with raster value obtained from Topps model for dry season showing less correlation coefficient (0.48 to 0.49) compared wet season (0.54 to 0.56). Similarly when bulk density correlated with same raster values for Dry season showing less correlation coefficient (0.36 to 0.44) compared to wet season (0.35 to 0.57) respectively. When we compared the estimated soil moisture values obtained from Topps model for Muthur location (0.49 to 0.55) and Nambiyur location (0.48 to 0.56) locations, there is no significant different is observed. But when same raster values compared with the bulk density there is significant different in between Muthur (0.37 to 0.57) and Nambiyur (0.35 to 0.49) locations. It is found that the raster value from Topps model during wet season having more nearness to actual soil moisture observations compared to dry season. While raster value from Topps model during dry season having more nearness to bulk density compared to wet season. There is no significant different is found in accuracy in estimation of soil moisture for two different locations. These results suggest that more research is required to suggest the suitability of Topps model for RISAT-1 data.