基于地面雨情信息的长江三峡区间洪水预报研究  

Flood forecast for Three Gorges region of the Yangtze based on ground-observed rainfall

作　　者：李哲[1] 杨大文[1] 田富强[1] 

机构地区：[1]清华大学水利水电工程系水沙科学与水利水电工程国家重点实验室,北京100084

出　　处：《水力发电学报》2013年第1期44-49,62,共7页Journal of Hydroelectric Engineering

基　　金：水利部创业公益项目(201001004;201101004);水沙科学与水利水电工程国家重点实验室与灾害天气国家重点实验室联合资助项目

摘　　要：长江三峡区间的暴雨洪水对三峡水利枢纽的影响迅速而且直接,是进行三峡坝址洪水预报的重要边界条件。本文基于分布式物理模型GBHM与概念性新安江模型开展了三峡区间洪水预报研究。利用2004-2007年地面观测的雨量及流量信息,首先在日尺度上对两个预报模型进行了参数率定及模型检验,然后在小时尺度上对典型洪水过程进行了预报试验。结果表明,两个模型对逐日洪水过程的模拟均有较高精度。考虑到逐时流量数据的限制,本研究直接利用在日尺度上拟合的模型参数进行逐时的洪水预报。结果显示新安江模型的预报精度大幅下降,GBHM模型对洪水过程及洪水总量的预报仍能够保持一定精度,表明分布式物理性水文模型在三峡区间洪水预报中具有应用优势和潜力。Flood generated in the Three Gorges region （TGR） of the Yangtze impacts fast and significantly on the operation of Three Gorges reservoir, and it offers critical boundary conditions for flood forecast at the dam site. We studied the short-term forecast of this flood with a physically-based distributed model （GBHM） and a conceptual model （Xin＇ anjiang model） separately. These models were calibrated and validated using the daily runoff and ground-observed rainfall from 2004 to 2007, and the forecasts were also on a daily time scale. The results show that both models capture the daily flood process with good accuracy, and the GBHM model is slightly better. To overcome the limitation of hourly runoff data available, calibration of the model parameters was maintained on a daily time scale, while an hourly time scale was adopted for the forecasts. These hourly forecasts reveal that the Xin＇ anjiang model shows a dramatic decrease in accuracy, while the GBHM model can preserve certain accuracy in calculations of total volume and time history of runoff. Thus, the potential of physically-based distributed model in application to flood forecast for the Three Gorges region are demonstrated.

关 键 词：水文学 短期洪水预报 地面观测雨量 GBHM模型 新安江模型 

分 类 号：P338.2[天文地球—水文科学]