Deriving antecedent moisture conditions from airborne SAR for input into a flood forecasting model

Recent research into active microwave remote sensing has demonstrated the feasibility of obtaining spatial and temporal soil moisture maps. C-band SAR data is used to determine the distribution of soil moisture for a study region within a medium size watershed in southern Ontario in October of 1991. A multiple regression model for assessing average field volumetric soil moisture for bare fields within the basin is applied to a data set. The results were then extrapolated to the entire basin using bare index fields derived from a LANDSAT TM image. The resulting basin moisture map is compared to a similar map derived using a distributed Antecedent Precipitation Index value. The two maps are both used as input initial conditions into a distributed flood forecasting model. The resulting hydrographs are compared with the actual hydrograph obtained for the simulation period