长江科学院院报 ›› 2023, Vol. 40 ›› Issue (3): 30-36.DOI: 10.11988/ckyyb.20211115

• 水资源 • 上一篇    下一篇

SWAT模型参数修正及应用——以德泽水库径流区为例

辛京达, 李亚强, 陈成, 杨栋淏, 余倩, 王建雄   

  1. 云南农业大学 水利学院,昆明 650000
  • 收稿日期:2021-10-23 修回日期:2022-01-02 出版日期:2023-03-01 发布日期:2023-03-28
  • 通讯作者: 王建雄(1976-),男,云南昆明人,教授,博士,从事资源与环境遥感工作。E-mail:jianxiongw@126.com
  • 作者简介:辛京达(1997-),男,吉林长春人,硕士研究生,从事资源与环境遥感工作。E-mail:819045742@qq.com
  • 基金资助:
    云南省教育厅基金项目(2020Y0177,2020Y0186,2021J0126)

Modification of SWAT Model Parameters and Application: Case Study on the Runoff Area of Deze Reservoir

XIN Jing-da, LI Ya-qiang, CHEN Cheng, YANG Dong-hao, YU Qian, WANG Jian-xiong   

  1. School of Water Resources, Yunnan Agricultural University, Kunming 650000, China
  • Received:2021-10-23 Revised:2022-01-02 Online:2023-03-01 Published:2023-03-28

摘要: 受人类活动和重大水利项目的影响,SWAT模型在对河道参数和平均坡长的计算中存在一些缺陷。通过遥感技术和经验公式对不合理的参数进行了修正。在对参数进行修正后,选取对水文、水质过程较为敏感的参数运用SWAT-CUP软件对模型进行优化,其率定和验证精度均达到了较为满意的效果(径流量R2在0.764 3~0.967 4之间、NS在0.754 8~0.899 8之间,总氮负荷R2在0.546 2~0.659 9之间、NS在0.526 8~0.638 8之间),结果表明在其他参数不变的情况下,这种修正方式不仅可以更好反应流域内真实的地形地貌及河道信息,同时可提高模型月尺度径流的模拟精度(R2从0.834 9提升到0.871 9,NS从0.698 7提升到0.771 8)。可实际应用于德泽水库径流区。

关键词: SWAT, 参数修正, 河道参数, 平均坡长, 遥感技术, 经验公式, 德泽水库

Abstract: Affected by human activities and major water conservancy projects, the SWAT model has some shortcomings in the calculation of river parameters and average slope length. We use remote sensing technology and empirical formulas to modify the unreasonable parameters. After modifying the parameters, we select parameters that are more sensitive to hydrology and water quality processes and employ SWAT-CUP software to optimize the model. The calibration and verification accuracy are satisfactory. The R2 and NS of runoff is 0.764 3-0.967 4 and 0.754 8-0.899 8, respectively, and the R2 and NS of total nitrogen load is 0.546 2-0.659 9 and 0.526 8-0.638 8, respectively. The results indicate that given the other parameters unchanged,the present method better reflects the real topography and river channel information in the basin, and also improves the simulation accuracy of the model's monthly runoff. R2 increased from 0.834 9 to 0.871 9, and NS increased from 0.698 7 to 0.771 8.

Key words: SWAT, parameter modification, channel parameters, average slope length, remote sensing technology, empirical formulas, Deze reservoir

中图分类号: