A Multi-Scale Analysis of Namibian Rainfall: Comparing TRMM Satellite Data and Ground Observations

Friday, 19 December 2014
Xuefei Lu1, Lixin Wang1, Ming Pan2 and Kudzai Farai Kaseke1, (1)Indiana University Purdue University Indianapolis, Indianapolis, IN, United States, (2)Princeton University, Princeton, NJ, United States
Rainfall is critically important in dryland regions, as it is the major source of water for natural vegetation as well as agriculture and livestock production. However, the lack of ground observations has long been a major obstacle to the study of rainfall patterning in drylands. In this study, a continuous 6-year record of ground observations collected at Weltevrede Guest Farm Namibia was used to evaluate the Tropical Rainfall Measuring Mission (TRMM) 0.25-degree (~25 km) 3-hourly satellite rainfall estimates for the period of 2008-2013 for two locations. The agreement between ground and satellite rainfall data was generally good at annual scales but a large variation was observed at the hourly scale. A trend analysis was carried out using bias-corrected annual satellite data (1998-2013) to examine the long-term patterns in rainfall amount, intensity, frequency and seasonal variations. Our results suggest that satellite rainfall estimates offer reasonable performance at annual scale. The preliminary trend analyses showed significant changes in frequency, but not in intensity or total amount in one of the two locations during the rainy season (November – March), but not in the other, emphasizing the spatial variability of the dryland rainfall.