Abstract
In October 2010, the water level upstream of the Three Gorges Dam (TGD) reached the designated 175-m level. The associated inundation and land use–land cover changes have important implications for water resource management, agriculture, ecosystems, and the hydroclimate. Ultimately, it is important to quantify whether the dam-related changes have altered precipitation patterns. Since rain gauges are limited in the region, satellite-based methods are viable. This study is the first to validate NASA Tropical Rainfall Measuring Mission (TRMM) Multisatellite Precipitation Analysis (TMPA) data from 1998 to 2009 using 34 national meteorological rain gauges in the Three Gorges region. Areal average satellite estimates are first verified with areal average rain gauge data both annually and seasonally. Then based on empirical orthogonal functions, the study area is divided into two subregions, and similar validation procedures are performed for both subregions. TMPA data are found to have high correlations with rain gauge data for the whole study area, and correlations for the subregions are only slightly lower. The seasonal analysis yields the lowest correlations for winter. Compared with the gauge data, rainfall is slightly overestimated by about 3 mm month−1. At daily scale, satellite data show good agreement with gauge data for all rain intensity categories except light rain (<1 mm day−1). Spatially, the point-source gauge data are gridded using Thiessen polygons for comparison with satellite data, and the results suggest the satellite-based product may overestimate rainfall in mountainous areas near the reservoir, especially in spring and summer. Overall, the validation results yield strong statistical support for applying satellite rainfall data for hydroclimate studies in this region.
