SWAT模型在大尺度流域径流模拟影响因素分析——以巴拉那河上游流域为例

Influencing Factors Analysis in Discharge Simulation of Large⁃scale Watershed Using SWAT Model--Taking the Upper ParanáRiver Basin in Brazil as an Example

为了探讨SWAT模型在大尺度流域水文分析模拟的效果,以巴西巴拉那河上游流域为研究区,利用构建的流域大尺度SWAT模型,定量分析美国国家环境预报中心再分析数据(CFSR)降水再分析数据质量、子流域划分数量及模型参数对径流模拟结果的影响。结果表明,CFSR再分析数据对流域降水的空间分布估计偏差较大,导致径流模拟误差较大,模型参数经1次迭代后,确定系数(R^(2))值均小于0.2。随着子流域划分数量的增加,实测站点降水数据可以更准确地估计子流域面雨量,进而提高径流模拟拟合度。在子流域划分数量分别为66、781和4111的情况下,比较模型参数经过1次迭代后的径流模拟效果发现,子流域个数为781个时,R^(2)值相对于两种情况有明显提高。在涉及SWAT模型径流模拟的20个参数中,土壤蒸发补偿系数(ESCO)和SCS径流曲线数(CN2)是最敏感的两个参数。由分析可知,SWAT模型可以在大尺度流域径流模拟中应用,研究结果可以为大尺度流域水文模型的构建与应用提供参考。

In order to analyze the effect of SWAT model in hydrological analysis and simulation of largescale watershed,a SWAT model of the upper Paranáriver basin in Brazil was constructed.The effects of CFSR reanalysis rainfall data,subbasin delineations and parameters of SWAT model were analyzed based on the simulation results.The results are as follows.The CFSR reanalysis data has a large deviation in the estimation of the spatial distribution of precipitation in the study area,which leads to a large error in the discharge simulation.After one iteration of the model parameter calibration,the R^(2)values were less than 0.2.With the increase in the number of subbasin delineations,the measured precipitation data could more accurately estimate subbasins'areal precipitation,thereby improving the accuracy of the discharge simulation.However,the number of subbasins has a reasonable range.The discharge simulations of 66,781 and 4111 subbasins after one iteration of model parameters calibration were compared,and it is found that when the number of subbasin delineations was 781,the R^(2)value was significantly improved.SWAT model can be used in runoff simulation of largescale watershed,the research results can provide reference for the construction and application of largescale watershed hydrological model.