基于全球降水数据估计值的地表径流模拟--以长江上游地区为例

为了评价在相对较短时间内针对长江上游地区地表水径流所建立模型的模拟效果,以及检验以GCM模型和其他卫星数据所估算的降水数据作为输入数据的可行性,选择分布式水文模型HSPF以及1987年和1988年的ISLSCP降水数据作为输入数据。模型模拟结果表明:从整个长江上游地区看,在校正期内,5天平均流量的Nash–Sutcliffe相关系数(R2) 为0.94;在验证期内,Nash-Sutcliffe相关系数(R2) 为0.95。此外,该模型对长江上游主要支流的5天平均流量的模拟效果也很好,R2的值在0.46到0.96之间。例外的情况主要发生在沱江和嘉陵江,模型对2年洪水期的峰值流速的估计值偏低,沱江只有实际值的71%,嘉陵江只有实际值的61%。ISLSCP估计的降水比实际测量的降水频繁且程度要弱,这可能是HSPF不能在所有时间和所有区域都具有较好模拟效果的一个主要原因。

Simulation of Water Runoff Using Estimated Global Precipitation Data: Taking the Upper Reaches of the Changjiang River as an Example

To evaluate the performance of a computer model simulating runoff in the upper reaches of the Changjiang (Yangtze) River Basin over a relatively short time interval, including examining the applicability of the input precipitation data generated from global circulation models and satellite data, we used a spatially distributed model, HSPF with the ISLSCP precipitation data for 1987 and 1988 as input data. The Nash-Sutcliffe coefficient (R 2 ) for 5-day average stream flow was 0.94 in the calibration period and 0.95 in the verification period for the whole upper region. Moreover, the model simulated the 5-day average stream flow well in each main tributary, as shown by R 2 values of 0.46 to 0.96, except that it underestimated the peak flow rates during the flood season over two years by up to 71% in the Tuojiang River and 61% in the Jialingjiang River. The ISLSCP precipitation tended to be more frequent and less intense than the measured precipitation. This was probably the main reason why the HSPF did not perform well in all regions at all times.