Evaluation of a multitemporal ERS-1 dataset for hydrologic applications

ERS-1 scenes covering the Ottawa region were acquired every 35 days for the period May 1992 through November 1993. The data were calibrated and used to evaluate the effects of hydrologically significant environmental conditions on backscatter from a variety of land covers targets. A land cover classification derived from Landsat TM data was used to facilitate extraction of backscatter statistics for classes such as wetlands, coniferous and deciduous forest, pasture, bare soil, and a variety of crop targets. Associated meteorological data and ground information were used in the interpretation of backscatter response. Each class exhibited a similar pattern of seasonal variation in backscatter, however, the magnitudes differed. The smallest seasonal range of 5 dB was associated with forest types while greater variation was evident for crop targets. The lowest backscatter values for all classes were found during the dry snowcover period. Forest and pasture classes exhibited the least within scene variance. The analysis has shown that multitemporal satellite SAR holds promise for monitoring hydrologic/environmental parameters