土地利用/覆被数据精度对流域水文过程模拟的影响研究

Effect of Different LULC Datasets on Hydrologic Modeling

土地利用/覆被是水文过程模拟的关键参数,影响截留、入渗、蒸散发和地下水补给等过程,最终影响河道径流量。【目的】更好地了解不同精度的土地利用/覆被数据对水文过程模拟的影响。【方法】将黑河上游祁连山区3种具有不同分辨率、不同分类精度的土地利用/覆被数据(NLULC,NLULC-1km以及MICLULC)分别输入SWAT水文模型中模拟黑河上游出山径流量,分析了模型率定前和率定后的模拟效果。【结果】将空间分辨率更高、土地利用/覆被分类较少的土地利用/覆被数据输入SWAT模型,能够减少模型校准的复杂度和工作量;模型校准后,分类程度越详细的土地利用/覆被数据对应的SWAT模型径流模拟效果越差;水文模型的校准过程会导致土地利用/覆被数据的空间分辨率对径流模拟的影响变小。【结论】在利用SWAT模型进行径流模拟时,输入分辨率高且分类较少的土地利用/覆被数据能够使模型模拟效果更好。

Land-use/land-cover is an important input into the hydrologic model. It can affect the rates of interception, infiltration, evapotranspiration, and groundwater recharge that result in influences to the timing and amounts of river runoff.【Objective】Better understand the impacts of the different accuracy of land-use/land-cover data on the hydrologic modeling.【Method】We evaluated the different accuracy of land-use/land-cover data by comparing three land-use/land-cover datasets（NLULC, NLULC-1 km and MICLULC） based on different spatial resolution and interpretation accuracy. Subsequently, the three land-use/land-cover datasets were input into SWAT model and simulated the river runoff from mountainous area in Upper Reach of the Heihe River, respectively.【Result】Inputting the land-use/land-cover dataset with the higher spatial resolution and less land-use/land-cover classification was important to reduce the complexity and works during the model calibration process; After model calibration, the land-use/land-cover dataset with more detailed land-use/land-cover classification corresponded to a worse model efficiency. Moreover, the impacts of spatial resolution of the land-use/land-cover dataset would be weakened with model calibration.【Conclusion】Inputting land-use/land-cover data with higher spatial resolution and less classification to SWAT can improve the model efficiency when simulating the stream flow.