Estimates of observation error based on shortwave infrared perpendicular water stress index for regional assimilation
短波红外垂直失水指数观测误差估计方法及其同化方案

Confirming observation operator and its spatial form of error variance is especially important for regional data assimilation scheme. In previous studies, observation error derived from remote sensing data is assumed only correlated with time while its spatial heterogeneity is often ignored. This assuming alleviates the computing pressure while also brings error to the assimilation process. In this study, new method for estimating the observation error based on shortwave infrared perpendicular water stress index was presented. The new observation error estimation method was further used into a two stage data assimilation scheme. From data assimilation experiment, it was demonstrated that the improved data assimilation scheme can fairly reveal spatial variations of surface soil moisture resulted from spatial and quantitative heterogeneous of vegetation and further improve assimilation accuracy.