基于集合卡尔曼滤波(ensemble Kalman filter,EnKF)方法和分布式水文模型SWAT(soil and water assessment tool),构建了一个土壤水分状态与参数同时更新的土壤湿度数据同化方案,通过遥感观测土壤湿度数据同化的仿真试验,研究土壤湿度数据同化在优化土壤水分参数、改进模型产汇流过程模拟方面的效果及潜力.结果表明:通过表层(0～5 cm)土壤湿度数据同化可实现土壤持水能力参数的准确估计;当给定的参数更新平滑因子在合理范围时,基于EnKF方法的参数优化效果具有较好的稳定性;表层土壤湿度数据同化对SWAT模型产汇流过程模拟有一定改进,但受降雨误差的影响,其对流域出口径流过程改进效果有限,表明基于遥感土壤湿度数据同化改进流域水文过程模拟还有赖于降雨输入精度及可靠性的提高.
Soil moisture data assimilation experiment based on EnKF method and SWAT model
A soil moisture(SM)data assimilation scheme in which soil water status and parameters were updated simultaneously was constructed based on the ensemble Kalman filter(EnKF)method and the distributed hydrological model of SWAT(soil and water assessment tool).The effects and capability of SM data assimilation in optimizing soil parameters and improving runoff generation and convergence process simulations of the model were investigated through the simulation experiment of remote sensing SM data assimilation.The results showed that the parameter of soil water holding capacity could be accurately estimated by surface(0～5 cm)SM data assimilation.The EnKF based parameter optimization effect was stable when the given parameter updating smoothing factor was in a reasonable range.The runoff generation and confluence processes had certain improvement by surface SM data assimilation.However,the improvement of the basin outlet runoff process was limited due to the impact of the precipitation errors.It can be inferred that the improvement of watershed hydrological modeling based on remote sensing SM data assimilation depends on the improvement of accuracy and reliability of precipitation input.
NETL
NSTL
万方数据
维普
