摘要：

Remotely sensed data, such as those from radar and satellite, provide rich information of the atmosphere and are especially valuable over locations with limited traditional observations. Successful use of these indirect observations in retrieval and data assimilation requires accurate handling of hydrometer scattering and radiation contribution from Earth surface. In this talk, I will first go over our past research on scattering properties of ice hydrometers in millimeter wavelength range, and their applications in radar retrievals. Then I will focus on two methods we recently developed to enhance satellite data assimilation: 1) to reduce uncertainties from errors in surface properties by combining multiple satellite channels, and 2) dimension reduction for hyperspectral observations in Ensemble Kalman Filter (EnKF) system using principal component analysis (PCA).