土壤湿度初始化对WRF-Hydro模式陆面水文过程模拟的影响

土壤湿度既影响大气低层和陆面之间的能量交换，也影响降水-径流过程的水分交换和分配。利用WRF-Hydro大气-陆面-水文全面双向耦合模式，研究土壤湿度初始化对流域陆面水文过程从天到次季节-季节尺度（S2S，0-60天）模拟的影响。首先，检验评估9种分析/再分析资料的表层土壤湿度产品在中国不同地区的时空适用性。其次，采用动态维度搜索（DDS）和多目标函数率定方法，评估WRF-Hydro模拟的水量和能量平衡，优化率定陆面和水文模式的敏感参数。利用动态贝叶斯平均方法，综合考虑多种土壤湿度产品的不确定性，生成融合的土壤湿度产品，并结合陆面同化系统的循环同化，优化设计土壤湿度初始化方案。采用WRF-Hydro耦合模式在淮河上游流域开展2年高分辨率数值模拟，流域尺度分辨率为200 m。深入分析土壤湿度初始条件对陆面水文过程模拟的影响及其水热循环变化的物理机制，以改进S2S尺度的降水预报和径流预报。

Soil moisture (SM) regulates the energy balance between the lower atmosphere and the land, and affects the exchange and partitioning of water fluxes at the land surface during the rainfall-runoff process. This study investigates the impact of SM initialization on the simulations of terrestrial hydrological processes at daily and subseasonal to seasonal (S2S, 0-60 days) scales over the upper Huaihe River basin, using the two-way fully coupled atmosphere-land surface-hydrology WRF-Hydro model system. First, the nine SM model products (0-10 cm), including one analysis, three reanalyses, and five Land Data Assimilation System (LDAS) products, are comprehensively evaluated using the dense (over 2400 stations) in situ observations at different timescales (annual to sub-daily) in eight climate regions over mainland China during 2010-2017. Next, the sensitive parameters in the land surface and hydrological processes in WRF-Hydro are calibrated using the dynamically dimensioned search (DDS) algorithm and multi-objective function, considering water and energy balance. Based on the nine SM products, the dynamic Bayesian Model Average (BMA) method is adopted to generate a merging SM product accounting for the SM uncertainties. In combination with the high-resolution land data assimilation system (HRLDAS), the merged SM is applied to the cycling data assimilation, providing the optimized soil initialization to WRF-Hydro. By running the two-way fully coupled WRF-Hydro, the 2-year simulations will be implemented and verified over the upper Huaihe River basin, with the finest basin-scale resolution of 200 m. This study will further explore the impact of initial SM conditions on the SM-precipitation feedback and rainfall-runoff generation, as well as the variation of water balance and energy balance, in order to facilitate improving S2S rainfall and streamflow prediction.