|Ryu, D -|
Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: December 1, 2011
Publication Date: December 5, 2011
Citation: Ryu, D., Crow, W.T. 2011. Correcting errors in catchment-scale satellite rainfall accumulation using microwave satellite soil moisture products [abstract]. Meeting Abstract. 2012 CDROM. Technical Abstract: Streamflow forecasting in poorly gauged or ungauged catchments is very difficult mainly due to the absence of the input forcing data for forecasting models. This challenge poses a threat to human safety and industry in the areas where a proper warning system is not provided. Currently, a number of studies are in progress to calibrate streamflow models without relying on ground observations as an effort to construct a streamflow forecasting systems in the ungauged catchments. Also, recent advances in satellite altimetry and innovative application of the optical has enabled mapping streamflow rate and flood extent in remote areas. In addition, remotely sensed hydrological variables such as the real-time satellite precipitation data, microwave soil moisture retrievals, and surface thermal infrared observations have the great potential to be used as a direct input or signature information to run the forecasting models. In this work, we evaluate a real-time satellite precipitation product, TRMM 3B42RT, and correct errors of the product using the microwave satellite soil moisture products over 240 catchments in Australia. The error correction is made by analyzing the difference between output soil moisture of a simple model forced by the TRMM product and the satellite retrievals of soil moisture. The real-time satellite precipitation products before and after the error correction are compared with the daily gauge-interpolated precipitation data produced by the Australian Bureau of Meteorology. The error correction improves overall accuracy of the catchment-scale satellite precipitation, especially the root mean squared error (RMSE), correlation, and the false alarm ratio (FAR), however, only a marginal improvement is observed in the probability of detection (POD). It is shown that the efficiency of the error correction is affected by the surface vegetation density and the annual precipitation of the catchments.