论文部分内容阅读
为利用卫星降水数据开展流域水文模拟预报,基于瓯江流域内94个雨量站1998~2013年逐日降雨量数据,分析了多源卫星降水数据(TMPA、RT、CMORPH和PERSIANN)在该地区的适用性。结果表明,PERSIANN数据精度最低,其次为CMORPH和RT,TMPA数据精度最高,其中PERSIANN和CMORPH数据低估了实际降水量,而RT和TMPA略高估了降水量;梅汛期和非汛期的精度高于台汛期,月时间尺度降水量的精度高于日时间尺度。各卫星降水数据中,TMPA和RT数据在中雨级别下精度最高,对小雨和大雨级别的降水分别存在低估和高估现象,而PERSIANN和CMORPH数据对所有量级下的降水均存在低估现象;随着雨量的增加,卫星降水的误报率均有不同程度的增加。
In order to carry out hydrological forecasting of the watershed using satellite precipitation data, based on the daily rainfall data of 94 rainfall stations in the Oujiang River Basin from 1998 to 2013, the application of multi-source satellite precipitation data (TMPA, RT, CMORPH and PERSIANN) in this area was analyzed Sex. The results showed that the accuracy of PERSIANN data was lowest, followed by CMORPH and RT, and the accuracy of TMPA data was the highest. PERSIANN and CMORPH data underestimated the actual precipitation, while RT and TMPA slightly underestimated the precipitation. The accuracy of plum and non-flood season was higher than Taiwan flood season, the monthly time scale precipitation accuracy higher than the time scale. Among the satellite precipitation data, TMPA and RT data have the highest accuracy at mid-rain level, underestimated and overestimated rainfall at light rains and heavy rains respectively, and PERSIANN and CMORPH data underestimate precipitation at all magnitudes. With the increase of rainfall, the false alarm rate of satellite precipitation has increased to some extent.