基于WEP模型和TRMM_PR的唐家山堰塞湖入湖径流预报

Predicting inflows to Tangjiashan barrier lake based on WEP model and TRMM precipitation data

为解决缺资料地区的堰塞湖入湖径流模拟预报问题,本文以2008年5月12汶川大地震形成的唐家山堰塞湖为研究对象;采用分布式水文模型与热带降雨测量卫星的降雨数据(TRMM_PR)相结合的方法,设置了3种降水输入模式,对堰塞湖形成期内(5月12日—6月9日)的日均入湖流量进行了模拟预报,并利用监测期内(5月29日—6月9日)入湖径流实测资料与模拟结果进行了对比。结果表明:根据气象站实测降水数据对TRMM_PR数据进行修正后输入模型(模式3),比将气象站实测数据直接输入模型(模式1)和将TRMM_PR观测数据直接输入模型(模式2)的模拟结果要好。本文的模拟结果表明,分布式水文模型可以和星载雷达测雨数据相结合可有效弥补因气象站点过于稀疏、单站控制面积过大造成的资料缺乏问题,而利用气象站点观测数据对TRMM_PR数据进行修正可以有效改善模型的降水输入,从而可以提高入湖径流预报精度。

In order to solve the problem of simulating and predicting inflows to barrier lakes in ungauged region,this study took Tangjiashan barrier lake as the research object which caused by Wenchuan earthquake on May 12th,2008.The distributed hydrological Model WEP and the TRMM-based precipitation estimates were combined together and 3 precipitation input scenarios were considered.The average daily inflow to the lake during the forming period of barrier（May 12th-June 9th） was simulated and predicted.The result was compared with the observed inflow during the monitoring period（May 29th-June 9th）.It shows that the modified precipitation data（i.e.modified TRMM_PR data by the gauge data,mode 3） can generate more accurate results than the other two modes（i.e.gauge data and TRMM_PR data that directly input to the model）.It may conclude that by connecting the distributed model with TRMM_PR data the problems caused by the scarcity of gauge stations can be solved.The modification of TRMM_PR by gauge data can improve the precipitation input of the model and generate more precise simulated results.