摘要
以黄河源区为研究区,运用GLUE法分析了HBV模型和新安江模型模拟水文极值的不确定性。目标似然函数阈值分别选用0.7、0.6、0.5,将得到的洪水和枯水水文极值以及选出的模拟结果按丰水年、平水年和枯水年过程分别进行分析。结果表明:两个模型均擅长于模拟洪水,而且对洪水模拟的不确定性较对枯水模拟的不确定性低;从两个水文模型模拟日过程估计区间的差别看,HBV模型估计区间与实测相比,洪水年总体偏低,枯水年总体偏高,平水年不存在明显趋势,新安江模型则不存在这个特点;在相同的参数采样方法和策略下,HBV模型的不确定性比新安江模型更显著;两个水文模型对枯水指标Q_(90)和Q_(75)模拟的不确定性都较大;新安江模型对洪水指标Q_(25)和Q_(10)的估计区间小于HBV模型的,不确定性水平较低。
The source area of the Yellow River was chosen as a study region. This study used GLUE method to analyze two typical watershed hydrological models'( HBV and Xinanjiang model) uncertainty to hydrological extremum. This paper selected 0.7,0.6 and 0.5 as thresholds of objective likelihood function,analyzed hydrological thresholds of floods and droughts( insurance rates was 10%,25%,75% and 90%,confidence was 95%) and selected simulation results during wet year( 1981),normal year( 1968) and dry year( 1997). The results show that HBV and Xinanjiang model do well in the simulation of floods,and the uncertainty of flood simulation is lower than that of in dry seasons; from the difference of two hydrological models' simulations' estimation interval,we can find that,to the measured values,the estimation interval of HBV model is lower in flood year and higher in dry year,and there isn't obvious trend in normal year,however,Xinanjiang model doesn't have this characteristic; under the same parameters' sampling method,HBV model's uncertainty is more obvious than Xinanjiang model; HBV and Xinanjiang model's uncertainty are very high in the simulation of dry season's indexes( Q_(90) and Q_(75)); Xinanjiang model's uncertainty and estimation interval of flood year's index( Q_(25) and Q_(10)) is lower than that of HBV.
出处
《人民黄河》
CAS
北大核心
2018年第2期1-9,共9页
Yellow River
基金
国家自然科学基金面上项目(41371051)
"十二五"国家科技支撑计划项目(2013BAC10B01)
关键词
HBV模型
新安江模型
GLUE法
不确定性分析
水文极值
黄河源区
HBV model
Xinanjiang model
GLUE method
uncertainty analysis
hydrological extremum
source area of Yellow River
