延河流域作为典型的黄土高原丘陵沟壑区,水土流失是最主要的生态环境问题,其生态环境治理能为整个黄土高原生态恢复重建提供典型范例,因此评价延河流域景观格局变化对水文过程的影响,对流域景观格局与水文过程的研究具有重要的科学价值和现实意义。本文通过利用SWAT模型模拟的延河流域的41个子流域的水文响应单元,建立延河流域的“源—汇”水文响应单元景观格局指数,通过计算不同地形条件下水文响应单元随着坡度变化的面积累积百分比,构建了新的“地形—水文响应单元 (Slope-HRU) 综合景观指数”。并与SWAT水文模型所模拟获得的子流域径流量、产沙量进行相关性分析,评价“地形—水文响应单元 (Slope-HRU) 景观指数”对水文过程的表征能力和适用性。结果表明：在子流域尺度上Slope-HRU景观指数和产沙状况存在明显的区域分异规律,而且R值空间变异情况基本上与产沙量的空间变化规律相反,即R值呈现由南向北,由东向西递增,而产沙量由南向北,由东向西递减。R值与年产沙量在数值上呈显著相关关系,相关系数为-0.60;再将Slope-HRU景观指数与所选取的7个常用景观格局指数相结合,构成新的综合景观指数,发现它们与子流域的产沙量耦合关系明确,均显著负相关,相关系数绝对值范围为0.43~0.63。

As a typical experimental Soil and Water Conservation District, the Yanhe River basin has long been plagued by soil erosion due to severe human disturbances. The 'Grain-For-Green' project has been practiced to increase grain yield to feed the mass population in starvation and curbed the continually deteriorating ecological situation. It is importantly scientific to explore watershed landscape pattern and hydrological processes. Exerting remote sensing and geographic information technology, this paper firstly analyzed simulated ecological hydrological process in the Yanhe River basin based on SWAT model. On this basis, it applied landscape indices method, land use change on eco-hydrological processes was quantitatively described, particularly those soil erosion influences caused by the change of landscape pattern with the complicated topography and soil type status; then we defined the landscape unit including topography, soil and land use/land cover change (LUCC) information, constructed a comprehensive landscape indices which was closely related to soil erosion, and reflected the coupling relationship between regional landscape pattern change and soil erosion. The results are as follows: (1) this paper constituted the spatial heterogeneity of hydrological response units (HRUs) by several factors such as land use type, soil type, topography, etc. At the landscape level, seven landscape indices were selected with principal component factor analysis, which are DCAD, GYRATE_SD, COHESION, SHEI, TCA, PAFRAC, IJI, etc. Coupled analysis on relationships of landscape indices, annual runoff and annual sediment in each sub-basin, the correlation coefficient of seven selected landscape indices and runoff is very small and cannot pass all significant tests. But correlation among the indices except for TCA and IJI and sediment yield is significant; the absolute value of the correlation coefficient is between 0.3 and 0.5. (2) This paper built Slope-HRU landscape index: coupling relationship between Slope-HRU landscape indices and annual sediment in each sub-basin is very comprehensible, and correlation coefficient is -0.6, which is significantly negatively correlated. Based on 'source-sink' landscape theory of soil erosion, Slope-HRU landscape index was fabricated to reflect the relationship between landscape pattern and soil erosion process to a certain extent. According to the space distribution of R value of Slope-HRU landscape index in 41 sub-basins, the results showed that: in the sub-basin scale, Slope-HRU landscape indices and annual sediment have obvious regional differentiation regularity. And R value spatial variation of Slope-HRU landscape indices and the variation of annual sediment amount were on contrary. On the other hand, the R value increased from southeast to northwest, with annual sediment decreasing from southeast to northwest.