VIC模型在东江流域城市化水文响应研究中的应用

Abstract

VIC作为基于单元网格的半分布式水文模型，是目前评价流域环境变化下水文响应的重要工具。本文以东江流域为研究对象，利用Maryland大学研制的1 km植被数据、HWSD土壤数据集以及流域内气象和雨量站资料，构建了东江流域空间分辨率0.25˚ × 0.25˚的VIC模型。利用博罗水文站1961~1970年月流量实测数据并结合东江流域的汇流特点，采用Dag Lohmann汇流模型进行参数率定和验证。结果表明：多年平均径流相对误差分别为6.2%和2.0%，效率系数分别为90.1%和84.3%，模型对流域水文模拟具有较好适用性。针对VIC城镇区域产流的不足，利用模型讨论了不同植被和Ksat对网格径流的影响，结果表明植被文件中没有城镇面积产流关键的参数，而调整土壤文件中的Ksat值则能很好模拟城镇面积的产流计算。 As a grid based semi-distributed hydrological model, VIC is the important tool in evaluating the basin hydrological response to environmental changes. The Dongjiang VIC model is established at the spatial resolution of 0.25˚ by using the University of Maryland Global Land Cover Facility 1 km dataset, Harmonized World Soil Database and the Meteorological Data from the meteorological or precipitation stations located in the Dongjiang basin. Based on the characteristics of the flow confluence in DongjiangBasin, the routing model developed by Dag Lohmann is coupled with the VIC to calculate the stream-flow. Monthly values of simulated stream-flow are compared with observations collected from Boluo hydrologic station (1961-1970) to calibrate and verify the VIC model. The results show that the relative errors of average annual runoff are respectively 6.2% and 2.0%, and the Nash-Sutcliffe coefficients are respectively 90.1% and 84.3%, which means that VIC model has good adaptability to the hydrological simulation of Dongjiang Basin. Aiming at the parameterization deficiency of VIC model for stream-flow generation from urbanized areas, the influence of different types of land use and Ksat values to stream-flow generation for grid cell is discussed. The results show that there are no key parameters in the vegetation file for stream-flow generation from urbanized areas, while to adjust the value of Ksat in the soil file has a positive effect on simulating the base-flow and runoff generation from urbanized areas.