Abstract:

Although the precipitation data from weather stations are widely applied in weather forecasting，climatic prediction and climate change research，etc，it is quite hard to reflect effectively the spatial variation of precipitation． As a precipitation radar satellite，TRMM has been collecting plentiful temporal and spatial precipitation information，so it becomes effective data to study hydrology and climate． In this study，based on the TRMM 3B43 precipitation data with 0．25° × 0．25° spatial resolution from 2001 to 2010 and digital elevation model (DEM)in Linxia of Gansu Province，the precipitation from TRMM 3B43 were downscaled to 1 km × 1 km by using regression model and residuals method，firstly． Combined the observation value of 6 rain gauge stations，the downscaled precipitations were subsequently validated． And on this basis the temporal and spatial variation of precipitation in Linxia were quantitatively studied in recent 10 years． The results showed that the downscaling processing of TRMM 3B43 improved effectively the spatial resolution and quality of precipitation data，and it was certain credibility on the whole，but the values of precipitation were smaller than the ground observation．The average annual precipitation decreased from southwest to northeast in Linxia，and gradually increased with elevation rising，their correlation coefficient was 0．82．The rainfall in Linxia mainly concentrated from May to September，which was more than 70% of the annual precipitation，and precipitation was maximum in June and minimum in December．

Key words: precipitation characteristics, TRMM 3B43, downscaling, Linxia

摘要：

基于2001～2010年TRMM3B43降水资料和数字高程模型(DEM)数据，采用回归模型+残差的方法，对甘肃临夏回族自治州近10a的TRMM 3B43降水数据进行降尺度运算，并结合研究区6个雨量站的观测值，对TRMM 3B43降尺度结果进行精度检验，在此基础上定量研究了临夏回族自治州近10a的降水量时空变化特征。结果表明：TRMM 3B43降尺度降水量数据整体上具有一定的可信度，但比地面台站的观测值偏小；甘肃临夏州年降水量呈现出由西南向东北递减的趋势，且降水量随着海拔高度的升高而逐渐增加，两者相关系数为0．82;年内降水主要集中在5～9月，基本占全年降水量的70%以上，其中6月降水量最大，12月降水量最小。

关键词: 降水特征, TRMM 3B43, 降尺度, 临夏