地下水位在非饱和水流数据同化中的应用

为理解地下水位观测信息在非饱和水流数据同化中的数据价值,建立了基于地下水位动态观测信息的一维饱和-非饱和水流集合卡尔曼滤波,通过虚拟数值实验检验了地下水位观测信息在非饱和水力参数估计和水分校正中的潜在价值。研究结果表明:在以地下水位为唯一观测数据时,同时更新参数和水头比仅更新水头能更好地校正土壤剖面的水头分布;当多层单个水力参数未知时,地下水位观测可以为参数估计提供有效信息;当多层多个参数未知时,地下水位与多层多个参数之间的复杂关系导致观测信息难以估计出最优的(唯一的)参数值;地下水位可作为辅助信息,与含水量观测等信息联合运用改善参数估计和含水量预测精度。

Application of groundwater level data to data assimilation for unsaturated flow

Groundwater level is a low-cost measurement that may contain valuable information about unsaturated flow. To provide an insight into the value of groundwater level data, the ensemble Kalman filter for one-dimensional saturated-unsaturated flow is developed to assimilate the dynamic groundwater level data. The potential value of groundwater level data is investigated in a series of synthetic numerical experiments. Results show that when assimilating the groundwater level data alone, the model parameters and state variables give better estimations of the pressure head distribution when updated simultaneously than when only the state variables are updated. Groundwater level data leads to satisfactory estimations if one of the soil hydraulic parameters in the multi-layer soil is identified. However, groundwater level data cannot produce optimal (unique) estimates if multiple parameters are determined simultaneously because of the complicated relationships between the groundwater level data and other parameters. Groundwater level is a source of supplementary information that, when jointly used with soil water content data, can be used to improve parameter estimations and predictions of soil moisture content.