数据同化对流域蒸散发过程模拟的影响研究

Effect of Data Assimilation on Basin Evapotranspiration Simulation

为研究数据同化系统对流域水文过程模拟特别是蒸散发模拟的矫正情况,构建一种基于时变增益水文模型和遥感蒸散发的确定性集合卡尔曼滤波同化系统。参照流域出口流量和子流域蒸散发观测分析水文过程的模拟精度。在时间过程和空间分布上探求流域蒸散发的模拟精度、变化过程和同化校准效应。结果表明,经过数据同化,时变增益水文模型的模拟精度在纳西效率系数和水量平衡系数2个指标上都得到提高;子流域尺度上蒸散发精度也得到改善。进一步分析发现,数据同化对极端气候和人类用水影响明显下的水文模拟实现了较为显著的修正。该方法具有潜在的实用价值,可作为水文过程模拟的备选工具。

In order to study the correction of hydrological process,especially evapotranspiration simulation by data assimilation,this paper constructs a deterministic ensemble Kalman filter assimilation system based on time variant gain model(TVGM)and remote-sensing evapotranspiration model,analyzes the simulation accuracy by the observation of watershed discharge and sub-basin evapotranspiration,and further studies the simulation accuracy,change process and assimilation calibration effect in terms of the spatial and temporal distribution of evapotranspiration.The results show that:The simulation accuracy of TVGM is improved in both Nash-Sutcliffe efficiency coefficient and water balance coefficient through data assimilation.The evapotranspiration accuracy for the sub-basin is improved.The data assimilation significantly modifies the hydrological process simulation under the influence of extreme climate and human water use.Therefore,the method is of potential practical value and an alternative tool for hydrological process simulation.