10.3
Quantifying the Impact of Land Surface Properties on Satellite-based Precipitation Estimates
Yudong Tian, UMBC/GEST, Greenbelt, MD; and C. Peters-Lidard, J. B. Eylander, M. Garcia, and J. Zeng
Precipitation estimates from passive microwave (PMW) sensors aboard low earth-orbiting satellites have become a critical component in hydrological modeling systems. For example, the operational Agricultural Meteorology modeling system (AGRMET) employed by the Air Force Weather Agency uses precipitation estimates from the SSM/I sensors as a major forcing data for its land surface model. In addition, several leading high-resolution, global precipitation products rely on these estimates as their primary source, such as Goddard Space Flight Center's TRMM Multi-satellite Precipitation Analysis (TMPA) and NOAA's Climate Prediction Center Morphing technique product (CMORPH).
We analyzed 3 years of precipitation data from TMPA, CMORPH and AGRMET, and quantified the leading contribution of errors over the contiguous United States (CONUS). Using surface gauge data as reference, we found the leading source of errors common to the three products is missed detection of rain during winter seasons. The accumulated missed rain in each winter season measured as much as 150 mm over the northwestern and northeastern CONUS. Such misses are coincident with the presence of snow, based on AMSR-E snow water equivalent retrievals. For comparison, during summer, the total missed rain over these two regions, as well as over whole CONUS, is much smaller.
Our study also shows there are false alarms in the satellite-based estimates, but their overall contribution to the errors is very small over CONUS. However, over inland water bodies, there are systematic anomalies in false alarm rate in TMPA and CMORPH data. The satellite estimates over inland water bodies produce twice as many raining days with light rain intensity (< 2 mm/day), causing different rainfall statistics between water-body pixels and their adjacent land pixels.
Recorded presentationSession 10, Advances in Remote Sensing and Data Assimilation in Hydrology, Part II
Thursday, 24 January 2008, 11:00 AM-12:15 PM, 223
Previous paper Next paper