中文摘要：

针对海河流域不同的下垫面类型，选取密云（果园林地）、大兴（城郊农田）、馆陶（平原农田）3个观测站，建立垂直方向上以含水率臼为因变量、含根系吸水项的非饱和土壤水分运动数值计算模型。该模型以一维Richards方程为基础（以下简称RE模型），采用实测的降水和蒸散数据作为模型的上边界条件，运用全隐式有限差分法，分别对不同生长期内的土壤水分进行数值模拟，得到时间序列的土壤水分廓线，并分别采用成熟软件HYDRUS-1D的模拟结果和各观测站实测土壤水分对RE模型进行交叉验证和直接验证。结果表明RE模型能够很好地模拟海河流域不同下垫面土壤水分动态变化过程，3个站模拟结果与实测土壤水分数据的均方根误差（RMSE）分别为0．03127，0．0359和0．0409cm3／cm3。与HYDRUS-1D软件模拟结果（其与观测值的RMSE分别为0．03759，0．0647和0．0467cm3／cm3）相比，RE模型模拟的土壤水分具有更高的精度，也显示出RE模型的可靠性。探讨3个站土壤水分的时空变异规律及其影响因子并以大兴站为例，通过优化RE模型参数，探讨犁底层对土壤水分模拟结果的影响，进一步改善RE模型的模拟精度。

英文摘要：

A vertical soil moisture （θ）-based unsaturated soil hydrodynamic numerical model was developed and used to simulate soil moisture movement at Miyun （orchard woodland）, Daxing （Suburban farmland）, Guan- tao （plain farmland） sites, which represent different underlying surface types in the Hai River Basin. The model based on one dimensional Richards＇ equation （hereinafter referred to as the RE model） and the top boundary was input ground-measured rainfall and evapotranspiration data. Soil moisture during different growth stages was numeri- cally simulated based on the fully implicit Finite Differential Method to obtain time series of soil moisture profile. Soil moisture simulated by mature HYDRUS-1D software and observed at the sites was used for cross-validation and direct verification on the RE model simulation results, respectively. The result showed that the RE model was able to simulate the soil water dynamic changes of different underlying surfaces in Hai River Basin, with root mean square error （RMSE） compared to the observed soil moisture at three sites being 0. 0313, 0. 0359 and 0. 0409 cm3/cm3, respectively. Compared to the simulation results of HYDRUS-1D software （whose RMSE were 0. 0376, 0. 0647 and 0. 0467 cm3/cm3 , respectively） , the soil water simulated by RE model had higher precision which also showed the RE model reliability. The spatial and temporal variation of soil moisture and its impact factors were dis- cussed. For the case of Daxing site,we analyzed the impact of plow on soil moisture modeling by optimizing the RE model parameters, under which consideration improved the simulation accuracy of the RE model was improved.