考虑地貌空间分异和植被响应特征的黄河中上游陆面水文模型发展及应用

The upper-middle reaches of the Yellow River are famous for a wide variety of landforms and vegetation types, with large spatial variabilities in underlying surface among different regions. It then requires detailed surface information to accurately simulate eco-hydrological processes such as vegetation dynamics. However, most of current land surface hydrological models fail to meet the needs of precise vegetation growth simulation, because of a lack of a strict, clear and localized geomorphological parameterization scheme which leads to a simplification of eco-hydrological processes. In this study, we aim at analyzing the spatial variability in geomorphology and vegetation responses, and developing a land surface hydrological model with improved geomorphological parameterization and surface-groundwater interaction in the upper-middle reaches of the Yellow River. Based on multi-source multi-scale dataset (e.g., remotely sensed, gauge observed and reanalysis data), we propose to extract necessary geomorphological and vegetation factors. With study area divided into several regions according to land use and vegetation types, we explore the differences of vegetation response between these regions, and identify the dominant eco-hydrological processes with corresponding geomorphological factors. Considering the topological relationships between grids and features within certain grid, an improved geomorphological parameterization scheme is designed and coupled with a typical land surface model called CLM. The newly developed land surface hydrological model is then evaluated based on verification data such as soil moisture content, water table depth and water and heat flux observations, and applied for future vegetation growth simulations under different global warming levels. This study can provide theoretical and model support for underlying modification suggestions and climate change adaptations in different regions of the upper-middle reaches of the Yellow River.

黄河中上游地形地貌复杂，植被类型众多，不同区域下垫面差异明显，因此准确模拟植被动态等过程需要精细的地表信息。目前，大部分陆面水文模式缺乏严格明晰、因地制宜的地貌参数化方案，生态水文过程表述过于简化，不能满足不同区域植被生长情势模拟的需求。因此，本项目拟针对黄河中上游地貌空间分异和植被响应特征，发展改进地貌因子参数化与地表-地下水交互过程的陆面水文模式。项目基于多源、多尺度资料，提取下垫面特征地貌和植被因子；结合土地利用与植被类型资料，探究不同区域植被响应的特征，识别对应主导生态水文过程及相应特征地貌因子；依据网格间拓扑关系与网格内微地貌特征，设计适应模式尺度、植被特征的地貌因子参数化方案；基于典型陆面模式CLM框架，开发合理描述植被-地表-地下水关键交互过程的陆面水文模式，模拟未来不同升温幅度下当地植被生长情势。本研究可为黄河中上游下垫面改造、气候适应措施提供理论与模式支撑。

陆面水文模型是表述全球变暖情景下不同区域关键水文过程演变的重要工具，而如何在陆面水文模拟时空尺度下准确表述复杂下垫面空间分异特征，是提升模型模拟性能的重要问题之一。本项目围绕设计适应于陆面水文模型尺度的下垫面地貌因子参数化方案这一关键问题，开展了黄河流域陆面水文模型研发和应用研究。项目搜集了黄河流域卫星遥感、站点观测、再分析数据等多源下垫面和水文气象数据，构建了黄河流域基础数据集，分析了区域下垫面变化特征；设计了适应于陆面水文模型尺度的嵌套化网格，建立了对应河道汇流、作物灌溉、水库调节等关键过程的地貌因子参数化方案；构建了黄河流域改进陆面水文模型，进行了模型适用性验证；应用模型模拟了全球变暖背景下未来百年的径流变化，分析了未来不同升温情景下气候和植被变化等因素对径流变化的影响；进一步探索了耦合人工智能算法在模型发展中的贡献和潜在前景。本研究所设计的参数化方案提升了陆面水文模型的模拟性能，可为水文气象部门的水资源管理等决策提供支撑。