| Summary: | Remotely sensed data from satellites has the potential to provide spatially and temporally relevant hydrologic information. This data has been used in the development of the PERSIANN (Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks) system. At a global scale, the results of this system can be used at coarser 1 0 x 1 0 resolution, which allows for greater accuracy in daily precipitation values. However, at regional and watershed levels, which are customary to most hydrologic applications, higher spatial resolutions are required (4 km x 4 km). The accuracy at this spatial resolution is investigated at the 0.25° x 0.25° grid scale and is accomplished by comparing precipitation gauges and ground-based radar (NEXRAD) to the PERSIANN output at both scales. More importantly, watershed average precipitation, obtained from NEXRAD and Thiessen polygon interpolation of gauges is, for the first time, compared against satellite precipitation estimates. A robust methodology for both of these types of estimation is presented along with other factors influencing the data assimilation process including the relative performance measures of the corresponding data and seasonal variability in data platform implementation.
|
|---|
