水文循环是地球表层系统物质循环与能量交换的核心，决定着水资源的时空分布特征，深刻影响着区域自然生态系统和社会经济。流域水文循环过程是区域水资源形成、转化和演变的基础，其中的水分和能量交换机制至关重要。高精度、高时空分辨率的水文模型是研究流域水文循环过程的重要手段，可准确模拟和预测流域水文关键要素的时空分布和动态变化规律，深入理解各水文要素间的相互作用和反馈机制。我国西北部的黑河流域是典型的干旱-半干旱内陆河流域，具有以水为纽带的多元自然景观，且流域内寒区与旱区相伴而生、地形崎岖复杂、包气带-饱和带系统空间异质性较高，致使黑河流域呈现出总体水资源匮乏、水资源时空分布不均匀等特征。尤其，在气候变化和以农业灌溉为代表的人类活动的双重影响下，黑河流域地表水与地下水相互作用更加强烈，导致了日益严峻的流域生态环境问题和突出的水务矛盾。因此，急需开展黑河流域水文过程循环特征、演变规律及其影响机制的模拟与分析研究，厘清水文循环演变规律及其在区域水资源演化中的作用。然而，目前流域尺度水文模型受制于分辨率不足、建模过程复杂、参数方案经验化、运行效率低等因素，对水文循环过程的模拟能力有限，尚不能全面揭示水文循环中各要素的时空分布规律、相互作用机制和影响因素，严重制约了黑河流域水文过程的科学认知和水资源的有效管理。

本论文针对我国西北干旱-半干旱地区水资源问题，围绕黑河流域水文循环过程中关键水文要素相互作用机理的核心科学问题，基于地下水-地表水耦合模型ParFlow（PARallel FLOW）及其与陆面过程模型CoLM（Common Land Model）耦合之后的水文集成模型ParFlow-CLM，构建了高时空分辨率的黑河流域水文集成模型。采用河川径流、地下水位、地表蒸散发、地表温度等关键水文要素的野外观测数据作为验证，证明了该模型具有较高的模拟精度和计算效率。利用该模型，在黑河流域开展了一系列水文过程数值模拟实验，阐明了地下水位、压力水头、地表水热通量、表层土壤水分、地表蒸散发、地表温度等关键水文要素的空间分布规律、演变特征与相互作用机制，揭示了包气带-饱和带渗透系数异质性对关键水文要素的影响机制，为理解黑河流域各个关键水文要素相互作用机理提供了理论基础与模型工具。主要成果包括：

（1）基于一系列经典案例和黑河流域下游的典型剖面开展了模型基准测试，对地下水-地表水耦合模型ParFlow进行了严格的验证与评估。结果显示，ParFlow模拟的径流、地下水排泄量、土壤水分等结果与验证数据的一致性较好，证明ParFlow能够精准刻画出干旱区剖面上水分运移的动力学过程。

（2）在黑河流域中游的典型二维剖面上设置了三组不同分辨率的地形情景，基于ParFlow模拟了不同情景下压力水头、地下水位、地下水-地表水交换量、地下水响应时间和地下水位比率的特征差异，阐明了微地形对地下水流系统形成和演化的影响，揭示了微地形对地下水-地表水转化的影响机制。结果表明，若完全忽略微地形，地下水系统响应时间最高可增加115年，地下水流系统控制类型误判率最高可至37.2%。微地形显著影响水文变量的信号转化特征，同时影响了包气带和含水层浅部的阻尼作用。因此，微地形的精确刻画是准确模拟地下水-地表水交互作用的必要条件。

（3）基于包气带-饱和带渗透系数、地形与河网等公开数据集发展了适用于黑河流域的水文集成模拟参数化方案。采用校准后的黑河流域近地表大气驱动数据，成功构建了基于ParFlow-CLM的黑河流域水文集成模型，该模型具有较高的时空分辨率（空间分辨率为0.005°×0.005°，时间分辨率为1 h）。通过数值模拟实验，获取了黑河流域2010年压力水头、地下水位埋深、地表水热通量、地表蒸散发以及地表温度等关键水文循环要素的空间分布特征。与多个空间分布产品比较之后的结果表明，已构建的模型可较为准确地模拟水文要素的空间分布规律。

（4）采用河川径流、地下水位、地表蒸散发、地表温度等关键水文要素的观测数据和多个已有空间分布产品（遥感、再分析和模型模拟产品）作为验证，证明了基于ParFlow-CLM的黑河流域水文集成模型具有较高的模拟精度与合理性。结果显示，河川径流模拟结果与观测的相关性较好（Spearman相关系数均高于0.6），地下水位埋深模拟结果与观测的概率分布曲线较为符合，地表蒸散发量和地表温度模拟结果与观测的一致性较高（Bias均小于0.5，Spearman相关系数均高于0.7），地下水位埋深、地表温度和地表蒸散发模拟结果的空间分布特征均与遥感、再分析或模型模拟产品表现出较好的一致性。以上定量分析结果证明了已构建的模型具有较高的模拟精度，能够准确模拟各关键水文要素的时空分布特征与变化规律，可作为模拟预测黑河流域水文循环特征的有效工具。

（5）采用基于ParFlow-CLM的黑河流域水文集成模型，在黑河流域中游农业区和黑河全域两个尺度上开展了模拟实验，模拟了四组渗透系数空间异质性情景下流域地下水位、地表水热通量、地表蒸散发和地表温度的空间分布特征，阐明了包气带-饱和带渗透系数异质性对地下水位和地表水热通量空间变异性的影响。结果表明，饱和带渗透系数异质性是模拟地下水位的关键参数，而包气带-饱和带渗透系数异质性共同决定了地表径流的预测精度。尤其，0.1 m ~ 5 m为黑河流域地下水位关键区（即水位与能量相关），若忽略渗透系数异质性，则地下水位关键区面积将缩减4.5%。在林地、灌丛和农田地表覆盖类型上，潜热通量、感热通量、土壤热通量和地表净辐射均与地下水位埋深呈现出明显的相关性。总体上，渗透系数空间异质性增加了地下水位空间变异性，减小了地表水热通量空间变异性。

The terrestrial water cycle is a vital component of the earth surface system, which controls the water and energy exchanges as well as the spatiotemporal distribution of water resources. Basins are the foundational units in the terrestrial water cycle, with interactions and feedback mechanisms between the water cycle and energy balance exerting significant influence on groundwater-surface water interactions and land-atmosphere interactions. Thus, it is important to precisely model and forecast the spatial and temporal patterns the water cycle and energy balance at the basin-scale. A comprehensive understanding of these dynamics is crucial for understanding interactions and feedback among different hydrologic components. The Heihe River Basin (HRB), located in the northwest of China, is a typical inland river basin in arid and semi-arid areas. The groundwater-land surface process in the HRB is unique and complex. The dynamic interaction between groundwater and land surface process is characterized by profound exchanges, illustrating the distinctive hydrologic features of the HRB. However, the diversity of hydrologic cycle processes, the complexity of groundwater-land surface processes and the lack of monitoring data contribute to a substantial knowledge gap of related scientific inquiries. This knowledge gap has become an obstacle in the fields of hydrology and water resources management, impeding effective management and sustainable development and utilization of water resources. In particular, the coupled water-energy balance in groundwater-land surface processes is a difficult issue in hydrologic studies. Consequently, there is an urgent need to clarify the precise characteristics of groundwater-land surface processes. However, the challenging topography of the basin, featuring multiple microtopographic units, coupled with the high spatial heterogeneity of hydraulic conductivity in the vadose zone-saturated zone system, has resulted in complicated water and energy exchange processes. These complexities present formidable challenges in unraveling the mechanisms of groundwater-land surface processes in arid and semi-arid regions.

In this thesis, a hydrologic integrated model with high spatio-temporal resolution was constructed based on the coupled groundwater-surface water model ParFlow (PARallel FLOW) and its extended version, the integrated hydrologic model ParFlow-CLM (ParFlow-Common Land Model), with a primary focus on unraveling the intricate groundwater-land surface process mechanism in the HRB. The observation data of key hydrologic variables, such as streamflow, water table depth and evapotranspiration, were used for validation. The results showed that the accuracy of model simulations was high. Then, the model was used to simulate the groundwater-land surface processes at different scales in the HRB, the evolving patterns of key hydrologic variables such as groundwater table, pressure head, surface heat fluxes, surface soil moisture, evapotranspiration, surface temperature, etc. Furthermore, it revealed the influence mechanisms of heterogeneity of microtopography and infiltration coefficients on groundwater and surface water interactions. These findings provided a theoretical basis for understanding the interaction mechanism of groundwater and surface water in arid and semi-arid areas, thereby offering a modeling tool for this field of study. The main results include:

(1) Based on a series of benchmarking case, the  ParFlow model was evaluated by analytical solutions, laboratory-based experimental data, well-validated model simulation results and field observation data. The results demonstrated a high correlation between ParFlow and validation data. This indicated that ParFlow was able to characterize the dynamics of water transport in arid regions with high spatial and temporal accuracy.

(2) Based on the integrated groundwater model ParFlow, a typical 2D transect was constructed in the middle reaches of the HRB using multi-source driven data. The differences in the characteristics of pressure head, groundwater table, and coupled water-energy exchange fluxes were simulated using three scenarios representing various degrees of microtopography. Then, the influence of microtopography on the fractal behaviors as well as groundwater response time and water table ratio patterns of the groundwater flow system were explored. The results showed that the classification of the groundwater flow system might exhibit a misclassification rate as high as 37.2% when microtopography was disregarded, and the response time of the groundwater system might be extended by up to 115 years. Furthermore, microtopography played a pivotal role in governing the transformation of precipitation, irrigation, and streamflow into infiltration, as well as modulating the damping effect observed within the vadose zone and the shallow portions of the aquifer. Additionally, microtopography acted as a critical factor for accurately modeling the temporal phase, magnitude, and spatial orientation of groundwater-river exchanges and groundwater-surface water exchanges, represent fundamental prerequisites for a precise modeling of groundwater-surface water interactions.

(3) Based on the integrated hydrologic model, ParFlow-CLM, a parameterization scheme for the HRB was established using calibrated near-surface atmospheric driving data, and a high spatio-temporal resolution (0.005°×0.005°, 1 hour) groundwater-land surface process simulation platform for the HRB was constructed. The results showed that this high spatio-temporal resolution model precisely simulate the spatial and temporal distribution patterns of key elements of the hydrologic cycle, such as pressure head, groundwater level, surface water heat flux and surface temperature.

(4) In-situ observations and spatial distributed datasets of key hydrologic elements such as streamflow, water table depth, evapotranspiration, surface temperature were used to validate the integrated groundwater-land surface process simulation platform in the HRB. Firstly, the streamflow simulations exhibited a strong correlation with observations (R-Spearman > 0.6). Moreover, the water table depth simulations closely aligned with the probability distribution curve derived from observations. Additionally, the evapotranspiration and surface temperature simulations were in good agreement with observations (Bias < 0.5 and R-Spearman > 0.7). Meanwhile, the spatial distributed characteristics of ParFlow-CLM and datasets for intercomparison were similar with varying degrees of agreements. This collective evidence attested to the high precision of the ParFlow-CLM model. It proved that the ParFlow-CLM model can characterize the accuracy of key hydrologic elements and their seasonal variation.

(5) Based on the groundwater-land surface process simulation platform ParFlow-CLM, two 3D simulation experiments were constructed in the middle reaches of the HRB and the entire HRB, to simulate the differences in the characteristics of the groundwater table, surface heat fluxes, evapotranspiration and surface temperature. Four hydraulic conductivity field scenarios with different spatial heterogeneity were set. The impacts of the hydraulic conductivity heterogeneities within both vadose zone and saturated zone were simulated and analysed. The results showed that the hydraulic conductivity heterogeneity in the saturated zone had a controlling effect on the groundwater table, and the hydraulic conductivity heterogeneity in the vadose zone and saturated zone jointly controlled the streamflow. Furthermore, the critical zone for groundwater table in the HRB, spanning from 0.1 m to 5 m, was characterized by a correlation with energy fluxes. Neglecting hydraulic conductivity heterogeneity resulted in a 4.5% reduction in the extent of this critical zone.