Optimization of Radiometric and Geophysical Covariances in a 1DVar System for All-Sky Data Assimilation

In this study optimization of radiometric and geophysical covariance matrices are presented. In order to test the matrices, an existing 1DVar system is used for both retrieval and data assimilation pre-processing. The main focus is to optimize the matrices for data assimilation in cloudy and precipitating conditions. Several optimization procedures were applied to construct the radiometric and geophysical covariance matrices. For radiometric covariances, the process includes perturbing the cloud optical properties in multiple Community Radiative Transfer Model (CRTM) simulations of cloudy and precipitating conditions. The new geophysical covariance matrix was built by using a standard data-set made available by the ECMWF which has geophysical as well as hydro-meteorological data. A thorough assessment of the impact was undertaken using radar, radiosonde and dropsonde data. The assessments were performed on the hydrometeors as well as temperature and moisture profiles in precipitating condition. Results including the impact on temperature and water vapor soundings and emissivity against collocated radiosonde, dropsonde, and analysis references will be presented.