基于模糊聚类法的无资料流域水文过程模拟

Hydrological Process Simulation in Ungauged Areas Based on Fuzzy Clustering Method

以流域面积、平均高程、平均坡度、长度、平均宽度、形状系数及平均曲率作为相似分析指标,采用模糊聚类法对新安江流域内12个有观测资料的子流域进行相似分析。在相似程度λ取0.837时将子流域分为两组,即岩前、渔梁、万安坝、屯溪、源口、临溪、月潭及良干流域为一组,屯溪作为参证流域;百罗畈、中洲、流口及上邹畈流域为一组,流口作为参证流域。通过新安江模型进行参数率定与移植,验证流域模拟结果的相对径流误差在±20%范围以内,即在规定的许可误差范围内,表明采用模糊聚类法的参数移植模拟计算能较好地反映径流的变化过程。

The watershed area,the average elevation of watershed,the average gradient of watershed,the length of watershed,the average width of watershed,the shape coefficient of river basin and the average curvature of river basin are used as similarity index.Fuzzy clustering method is chosen to classify 12 sub basins with observational data in Xin＇anjiang Basin.The sub basins are divided into two groups in the similarity degreeλtakes 0.837：Yanqian,Yuliang,Wan＇anba,Tunxi,Yuankou,Linxi,Yuetan and Lianggan as a group,Tunxi is chosen as the reference basin;Bailuofan,Zhongzhou,Liukou and Shangzoufan as another group,the reference basin is Liukou.Through the Xin＇anjiang model for parameter calibration and transplantation,the relative runoff error of the calibration result is within ±20%.Within the specified permissible error range,the simulation results of parameter transplantation based Fuzzy Clustering method can reflect the variation trend of runoff.