北方半干旱区典型流域土地利用变化的水文效应

Hydrological response to land use and land cover change (LUCC) is a hot spot in related cross-disciplinary global change researches. The integration and application of physically based, distributed hydrological models and GIS, RS, and other modern technology is an important direction in this field. However, there is a lack of spatiotemporal surface information and weak coupling with 3S technology in spatially distributed hydrological modeling in China. Aiming at the above deficiencies and focusing on the integration of SWAT hydrological model with GIS and RS techniques, this research selects a typical semi-arid northern catchment (the Laohahe River basin) as a case study to quantitatively investigate the impacts of LUCC on surface hydrology in semi-arid North China. Field measurements and experiments of vegetation and soil are applied to build a suitbale model database. The input of precipitation in SWAT model is improved using GIS spatial analysis and programming techniques. The spatiotemporal actual evapotranspiration distribution is estimated from MODIS images and used to calibrate SWAT model in addition to observed discharge data. Highly precise SWAT simulation will be achieved based on the above efforts. Based on LUCC maps during 1970s-2000s interpreted from multi-source remote sensing information and GIS spatial analysis techniques, hydrological modeling, and time series of observed hydro-meteorological data, quantitative response analyses of various factors of hydrologic processes to LUCC from multiple aspects are implemented. The efforts of this research help to advance related hydrological response study. The results of this research will reveal the mechanism of land use and land cover changes impacts to water cycle in semiarid region in China. Thus the research will provides scientific support for water resources distribution and ecological environment restoration in semi-arid North China.

土地利用变化的水文效应是全球变化交叉研究的热点，基于物理机制的分布式水文模型与GIS、RS等现代技术的结合及应用是该领域的重要方向。针对目前我国分布式水文模型缺乏流域下垫面时空变异信息以及与3S技术结合较薄弱等问题，本项目着眼于RS和GIS技术与SWAT模型的集成，选择北方半干旱区典型流域，通过野外观测和试验构建适宜的模型数据库，采用GIS技术改进SWAT降雨输入，利用MODIS反演的流域实际蒸散发及流量观测数据进行参数优化，从而实现高精度的SWAT模拟；在多源遥感信息融合及GIS空间分析技术实现自上世纪70年代至今的LUCC动态监测与制图的基础上，基于遥感监测、水文建模和水文气象观测时间序列数据，从多个层面分析水文过程诸因子对LUCC的定量响应。该研究有助于改进LUCC的水文效应研究，揭示LUCC对北方半干旱区水循环的影响机理，为我国北方半干旱区水资源调配和生态环境修复提供科学支持。

土地利用变化的水文效应是全球变化交叉研究的热点，基于物理机制的分布式水文模型与GIS、RS 等现代技术的结合及应用是该领域的重要方向。目前我国的该方面研究中存在分布式水文模型缺乏流域下垫面时空变异信息，且与3S 技术结合较薄弱等问题。本项目着眼于RS 和GIS 技术与分布式的SWAT 模型的集成，选择北方半干旱区典型流域——老哈河流域作为研究区，通过收集现有数据和野外观测，构建了适宜于老哈河流域的模型数据库，采用GIS空间分析技术结合基于薄盘光滑样条函数（Thin Plate Smoothing Spline, TPS）的专用气候数据空间插值程序包ANUSPLIN对流域内52个雨量站的长时间序列降雨数据进行空间插值和处理，作为SWAT 模型的降雨输入，从而改进了SWAT模型降雨输入的改进。利用2000年至2008年近9年的老哈河流域长时间序列的MODIS Level1B级数据产品和MOD03数据，构建基于地表能量平衡方程的SEBS模型，实现了老哈河流域蒸散发的时空分布序列估算。在上世纪70 年代至21世纪00年代的LUCC 动态监测与制图的基础上，基于遥感监测、水文建模和水文气象观测时间序列数据，定量分析了土地利用变化对流域产水特性（主要选择了径流深和实际蒸散发量）的影响以及对其年内分配的影响，探讨了多种土地利用类型的变化及其组合与径流深和实际蒸散发量的不同函数关系。本研究从多个层面分析水文过程诸因子对LUCC 的定量响应，研究结果有助于改进LUCC 的水文效应研究，揭示LUCC 对北方半干旱区水循环的影响机理，为我国北方半干旱区水资源调配和生态环境修复提供科学支持。