基于SWAT模型多站点不确定性评价方法的比较  

Comparative Study of Multi-Site Uncertainty Evaluation Methods Based on SWAT Model

作　　者：李倩楠[1,2] 张静[1,2] 宫辉力[1,2] LI Qiannan ZHANG Jing GONG Huili(State Key Laboratory Incubation Base of Urban Environmental Processes and Digital Simulation, Capital Normal University Beijing 100048, China Key Laboratory of 3D Information Acquisition and Application of MOE, Capital Normal University, Beijing 100048, China)

机构地区：[1]首都师范大学城市环境过程与数字模拟国家重点实验室培育基地,北京100048 [2]首都师范大学三维信息获取与应用教育部重点实验室,北京100048

出　　处：《人民黄河》2017年第1期24-29,共6页Yellow River

基　　金：国家自然科学基金资助项目(41271004);北京市科技新星项目(2010B046)

摘　　要：分布式水文模型在拥有众多优点的同时,也面临着参数过多难以率定的问题。SWAT模型作为典型的分布式水文模型,同样存在着参数率定难的问题。基于SWAT模型,选取美国佛罗里达州中部Peace河流域为研究区,采用SUFI-2、GLUE、PARASOL和PSO共4种评价方法进行了SWAT模型参数的敏感性分析、校准、验证以及不确定性研究,通过对4种不确定性方法的模拟结果、难易程度、运行次数以及各方法的理论基础进行对比,总结了4种方法的适用情况。结果表明:4种方法具有各自的优缺点和适用性,SUFI-2方法是半自动的,可以结合分析者的主观和认知,对于较复杂的模型更具有优势;GLUE方法相对简单,要优于SUFI-2方法;PARASOL方法适用于需要找到全局最优纳什系数的模型;PSO方法既适用于较简单的模型,也适用于相对复杂的模型,与PARASOL方法相比,两者得到的纳什系数、相关系数等均差别不大,但PSO方法运行次数大大减少,故PSO方法的整体性能优于PARASOL方法的。Distributed hydrological model has many advantages, at the same time it also faces the challenge to calibrate over-do parameters. As a typical distributed hydrological model, SWAT parameter calibration also exists problems. Based on SWAT model, the Peace River basin of central Florida was chosen as study area, using four different evaluation methods ： Sequential Uncertainty Fitting algorithm （SUFI-2） , Generalized Likelihood Uncertainty Estimation （GLUE）, the Parameter Solution （PARASOL） and Particle Swarm Optimization （PSO） techniques, to carry out sensitivity analysis, calibration and uncertainty of SWAT model. Based on the statistical results of four uncertainty methods, difficulty level of each method, the number of runs and theoretical basis were compared. The results show that the four kinds of methods have their respective advantages and applicability, SUFI-2 method is semi-automatic, which can be combined with the cognitive of analyzer and this method has more advantage for more complex model; for a simple model, GLUE method is relatively simple, which is superior than the SUFI-2 methods; PARASOL method is suitable for the model to which needs find the global optimal Nash coefficient; PSO method applies to a simple model, and can also be applied to relatively complicated model. Compared with the PARASOL method, both the coefficient of Nash, correlation coefficient difference is not big, but PSO method run times is greatly reduced. In conclusion, the overall performance of the PSO method is better than that of the PARASOL method.

关 键 词：SUFI-2方法 GLUE方法 PARASOL方法 PSO方法 不确定性分析 SWAT模型 

分 类 号：P343[天文地球—水文科学]