摘要
采用集合卡尔曼滤波同化水文模型与遥感模型反演的蒸散发(ET)结果,获得更接近真实情况的ET,并利用同化后的ET结果进一步优化水文模型,从而获取连续精度较高的区域ET.通过对北京市沙河流域1999—2007年的ET过程研究显示,运用本方案可以改善水文模型对ET的估算精度,模拟出精度较高的流域蒸散发过程.
A new integrated method was used to estimate continuous, high precision regional ET by data assimilation （DA） technology combined with remote sensing and hydrological models. Data from 18 typical TM/ETM＋images taken under clear skies from 1999 to 2007 in Shahe river basin（Beijing） were used. Two- layer remote sensing ET was used to calculate regional ETs for 18 days; Calculated values were used as observation values for DA. Distributed hydrological model was used to simulate daily ETs for these 8 years as background values. The two data sets were assimilated by ensemble kalman filter （EnKF） algorithm. Assimilated ETs were used to construct objective function to further optimize hydrological model in order to obtain a time-continuous ET process. The model was tested with observed daily ETs from flux observation stations in the study area. Optimization by assimilated ETs reduced average relative ETs error from 13.8% to 6.4%. Thus, integrated approach is appropriate for simulating ET process with high precision.
出处
《北京师范大学学报（自然科学版）》
CAS
CSCD
北大核心
2014年第1期73-82,共10页
Journal of Beijing Normal University(Natural Science)
基金
国家自然科学基金资助项目(41371043
41271003)
关键词
蒸散发过程
水文模型
遥感模型
数据同化
集合卡尔曼滤波
ET process
distributed hydrological model
remote sensing ET model
data assimilation
ensemble kalman filter（EnKF）
