自然资源学报

自然资源学报 ›› 2014, Vol. 29 ›› Issue (10): 1746-1757.doi: 10.11849/zrzyxb.2014.10.010

• 资源评价 • 上一篇    下一篇

基于SWAT模型的东江流域径流模拟

吕乐婷1, 彭秋志2, 郭媛媛1, 刘颖慧3, 江源1   

  1. 1. 北京师范大学 地表过程与资源生态国家重点实验室, 北京100875;
    2. 昆明理工大学国土资源工程学院, 昆明650093;
    3. 北京师范大学 资源学院, 北京100875
  • 收稿日期:2013-09-24 修回日期:2014-01-20 出版日期:2014-10-20 发布日期:2014-10-20
  • 通讯作者: 江源
  • 作者简介:吕乐婷(1984-), 女, 辽宁大连人, 博士研究生, 主要从事水文水生态系统保护与管理研究。E-mail: lvleting@qq.com
  • 基金资助:

    国家自然科学基金(41271104);国家水体污染控制与治理科技重大专项(2012ZX07501002)。

Runoff Simulation of Dongjiang River Basin Based on the Soil and Water Assessment Tool

Lü Le-ting1, PENG Qiu-zhi2, GUO Yuan-yuan1, LIU Ying-hui3, JIANG Yuan1   

  1. 1. State Key Laboratory of Earth Surface Processes and Resource Ecology, Beijing Normal University, Beijing 100875, China;
    2. Faculty of Land Resource Engineering of Kunming University of Science and Technology, Kunming 650093, China;
    3. College of Resources Science and Technology, Beijing Normal University, Beijing 100875, China
  • Received:2013-09-24 Revised:2014-01-20 Online:2014-10-20 Published:2014-10-20

PDF (PC)

赞

被引次数

13 | 17

摘要：

基于分布式水文模型SWAT以及SUFI-2 程序对东江流域进行了51 a(1960—2010 年)月径流模拟,并在此基础上讨论了东江流域水文过程的空间差异。模拟结果表明,河源站受新丰江水库影响,模拟结果不确定性较大,确定性系数R2和Nash-Sutcliffe 效率系数相对其他站点较低,龙川站、岭下站、博罗站R2和Nash-Sutcliffe 效率系数皆满足模型适用性要求。流域出口博罗站率定期R2、Nash-Sutcliffe 效率系数分别为0.83、0.83,验证期为0.84、0.84,模型模拟精度较高。东江流域径流深主要受降水空间不均影响,形成由北向南先增后减的趋势。流域下垫面差异对产流过程也有一定影响。其中土壤含水量受土壤性质、人类活动等影响,由北向南差异明显;蒸散发量受植被覆盖影响亦较为明显,北部蒸散量占降雨量的比例大于中部及南部。

关键词: SUFI-2 算法, 径流模拟, SWAT模型, 东江流域

Abstract:

The Dongjiang River (East River) is one of the three major tributaries of the Pearl River in southern China. It is also the vital water source for lots of important cities, such as Hong Kong, Shenzhen and Dongguan. Changes of runoff will have a significant impact on the environmental security and economic development of this region. A distributed hydrological cycle model of Dongjiang River Basin was developed based on the SWAT model (Soil and Water Assessment Tool) and the hydrological cycle of 51 years (1960-2010) was simulated. The“Sequential Uncertainty Fitting Algorithm”(SUFI- 2) was used to calibrate/validate and analyze the uncertainty of the model. Firstly, the monthly flow data from 1960 to 1985 of the four main hydrometric stations was used to calibrate. Then the model was validated with the monthly flow data from 1986 to 2010. And prediction uncertainties were quantified by P-factor and Rfactor. Model performance varied from weak at Heyuan to satisfactory at Longchuan, Lingxia and Boluo (outlet of the basin). This is because Heyuan hydrologic station was located closely to Xinfengjiang Reservoir, the largest reservoir in Guangdong Province. The results (coefficient of determination, Nash-Sutcliffe efficiency) of calibration and validation at Boluo station were 0.83, 0.83 and 0.84, 0.84 respectively. The P-factor and R-factor at Boluo were 0.76, 0.54 and 0.74, 0.59 respectively. On the whole, the model performed well, and then the spatial difference of the hydrological processes was discussed. The runoff depth of Dongjiang River Basin was affected mainly by precipitation, ascending from the north to the central and descending from the central to the south. Hydrological characteristic of basin underlayer (slope, vegetation, soil, etc.) also affects the processes of runoff generation. For example, soil moisture differs significantly from the north to the south. In the north more rain turned into soil water because of the high forest coverage and the soil properties. This is conducive to water conservation and soil conservation. And also, the evapotranspiration accounted for the largest proportion of rainfall in the north attributed to the impact of forest cover.

Key words: SWAT, Dongjiang River Basin, SUFI-2, runoff simulation

中图分类号: 

  • P333.9