基于SWAT模型的开都河上游未来土地利用变化对径流的影响

【目的】揭示土地利用变化下的流域径流变化。【方法】以开都河上游区域为研究对象,利用分布式水文模型SWAT,采用元胞自动机-马尔科夫模型(CA-Markov)预测法、当地土地利用规划要求和极端土地利用情景法共设置了4种2025年土地利用情景,包括自然发展情景、规划发展情景、极端草地情景、极端林地情景。【结果】①构建的SWAT模型在率定期(R2=0.74,NSE=0.73)与验证期(R2=0.58,NSE=0.57)的模拟结果可以接受。②采用CA-Markov模型模拟的2010年土地利用类型图Kappa系数为0.83,可用于预测2025年土地利用类型。1980—2010年土地利用类型变化主要表现为草地转换为水体和未利用地,其转换面积分别为106.09 km2和0.5 km2。③相较于基准年2010年,各土地利用情景下的年均径流量均减少,自然发展情景减少最快,极端草地情景减少最慢;从丰水期与枯水期来看,2种时期下各情景月均径流变化完全相反;从其季节分配来看,各土地利用类型在夏季对径流变化影响最大,冬季影响较小。【结论】该地区草地是影响径流最主要的调控因子,原有草地的减少会引起夏季径流较大幅度减少,对于未利用地,可考虑开发为草地。

Influence of Future Land Use Change on Runoff in the Upper Reaches of Kaidu River Based on SWAT Model

【Objective】Revealing the change of watershed runoff under the change of land use has important practical significance for the planning and rational utilization of water resources in the basin.【Method】Taking the upper reaches of the Kaidu River as the research object, this paper used the distributed hydrological model SWAT as a tool, and used the CA-Markov model prediction method, the local land use planning requirements and the extreme land use scenario method to set up 2025 land use scenarios. These include natural development scenarios,planned development scenarios, extreme grassland scenarios and extreme forestland scenarios.【Result】① The SWAT model constructed was statistically acceptable calibration period(R2=0.74, NSE=0.73) and validation period(R2=0.58, NSE=0.57). ② The Kappa coefficient of the 2010 land use type map simulated by the CA-Markov model was 0.83, which was highly accurate and can be used to predict the land use type in 2025. The change of land use type from 1980 to 2010 was mainly reflected in the conversion of grassland area into water body and unused land. The conversion area was 106.09 km2 and 0.5 km2 respectively. ③Compared with the base year(2010),the annual average runoff under each land use scenario is reduced, the natural development scenario was the fastest, and the extreme grassland scenario was the slowest; From the perspective of the wet season and the dry season, the monthly average runoff changes in each of the two periods are completely opposite. In terms of seasonal distribution, the land use types have the greatest impact on runoff changes in summer and less in winter.【Conclusion】Grassland in this area was the most important regulation factor for runoff. The reduction of original grassland will cause a large decrease in summer runoff. For unused land, it can be considered as grassland.