Passive microwave remote sensing of extreme weather events using NOAA-18 AMSUA and MHS

The ability to provide temperature and water-vapor soundings under extreme weather conditions, such as hurricanes, could extend the coverage of space-based measurements to critical areas and provide information that could enhance outcomes of numerical weather prediction (NWP) models and other stormtrack forecasting models, which, in turn, could have vital societal benefits. An NWP-independent 1D-VAR system has been developed to carry out the simultaneous restitutions of atmospheric constituents and surface parameters in all weather conditions. This consistent treatment of all components that have an impact on the measurements allows an optimal information-content extraction. This study focuses on the data from the NOAA-18 satellite (AMSUA and MHS sounders). The retrieval of the precipitating and nonprecipitating cloud parameters is done in a profile form, taking advantage of the natural correlations that do exist between the different parameters and across the vertical layers. Stability and the problem's ill-posed nature are the two classical issues facing this type of retrieval. The use of empirically orthogonal-function decomposition leads to a dramatic stabilization of the problem. The main goal of this inversion system is to be able to retrieve independently, with a high-enough accuracy and under all conditions, the temperature and water-vapor profiles, which are still the two main prognostic variables in numerical weather forecast models. Validation of these parameters in different conditions is undertaken in this paper by comparing the case-by-case retrievals with GPS-dropsondes data and NWP analyses in and around a hurricane. High temporal and spatial variabilities of the atmosphere are shown to present a challenge to any attempt to validate the microwave remote-sensing retrievals in meteorologically active areas.