高级检索+

GSSHA模型在黄家河流域的次洪模拟研究

Application of GSSHA Model to Flood Simulation in Huangjiahe Watershed

  • 摘要: 分布式水文物理模型GSSHA(Gridded Surface Subsurface Hydrologic Analysis)模型增加了描述非饱和土壤水分运动的一维理查德下渗方程进行下渗强度和累积下渗量的计算。理查德下渗方程对土壤水分运动的描述更为准确,但是其参数量较多,在实际应用中存在一定难度。对GSSHA模型具有明确物理意义且对产汇流计算影响较大的8个参数进行了敏感性分析,采用控制变量法对单一参数在范围内选择10个左右的参数值进行模拟和对比,洪峰流量和峰现时间模拟最大值与最小值的比值或差值作为判断敏感参数的标准,峰筛选出对洪峰流量和峰现时间有重要控制作用的敏感参数并进行敏感性排序。选择宁夏南部地区的黄家河流域为研究流域,对2000-2015年间18场洪水进行次洪径流模拟,其中前13场为率定期,后5场为验证期,参数率定过程按照先径流量后洪峰流量,研究GSSHA模型在半干旱地区的应用效果。参数敏感性分析结果显示对洪峰流量敏感性最高的参数依次是饱和水力传导度、泡点压力和初始土壤含水量,同时这三个参数也是峰现时间敏感性最高的参数。GSSHA模型对黄家河流域次洪过程在率定期和验证期内均具有较高的模拟精度,其中率定期内径流深相对误差为3.51%,洪峰流量相对误差为10.84%,峰现时差为-0.77 h;确定性系数达到0.80;验证期内径流深相对误差为2.18%,洪峰流量相对误差为-6.98%,峰现时差为0 h,确定性系数为0.79。

     

    Abstract: The physically-based distributed hydrologic model( GSSHA model)develops a one-dimensional Richards’ infiltration equation based on saturated or unsaturated zone to describe the soil water infiltration process. Richards’ equation is more accurate in describing soil water movement. However, a large number of parameters lead to certain difficulties in practical applications. In this paper, the sensitivity of the eight parameters, which have clear physical meaning and significant influence on runoff and concentration calculation in GSSHA model, is analyzed. The control variable method is used to simulate and compare with about ten different values of single parameter. The ratio or difference between simulated maximum values and minimum values of peak discharge and peak time is used as a criterion to judge sensitive parameters. Then the sensitive parameters of peak discharge and peak time are selected and sorted. Huangjiahe watershed in the southern part of Ningxia is selected as the research area. 18 flood events from 2000 to 2015 are simulated, of which the first 13 events are calibration period and the rest of 5 events are validation period. The application effect of GSSHA model in semi-arid areas is studied with the parameters calibrated based on the principle of runoff depth first and peak discharge then. The result of parameter sensitivity analysis show that the saturated hydraulic conductivity, bubbling pressure, and initial soil moisture are the highest sensitivity to peak discharge, and these three parameters are also the parameters with the highest sensitivity to peak time. GSSHA model has high simulation accuracy for the flood events in Huangjiahe watershed in both calibration period and validation period. The relative error for runoff depth, the relative error for peak discharge, the relative error for peak time, and Nash efficiency coefficient in the calibration period are 3.51%, 10.84%,-0.77 and 0.80, the values in the validation period are 2.18%,-6.98, 0 and 0.79.

     

/

返回文章
返回