Assimilation of Hyperspectral Infrared Sounder Radiances Under Cloudy Skies in a Regional NWP Model PDF Download

Author:
Publisher:
ISBN:
Category :
Languages : en
Pages : 238

Get Book

Book Description
Satellite measurements are an important source of global observations in support of numerical weather prediction (NWP). The assimilation of satellite radiances under clear skies has greatly improved NWP forecast scores. Since most of the data assimilation models are used for the clear radiances assimilation, an important step for satellite radiances assimilation is the clear location detection. Good clear detection could effectively remove the cloud contamination and keep the clear observations for assimilation. In this dissertation, a new detection method uses collocated high spatial resolution imager data onboard the same platform as the satellite sounders to help IR sounders subpixel cloud detection, such as the Atmospheric Infrared Sounder (AIRS) and Moderate Resolution Imaging Spectroradiometer (MODIS), the Crosstrack Infrared Sounder (CrIS) and Visible Infrared Imaging Radiometer Suite (VIIRS). The MODIS cloud mask provides a level of confidence for the observed skies to help AIRS Field-of-View (FOVs) cloud detection. By reducing the cloud contamination, a cold bias in the temperature field and a wet bias in the moisture field are corrected for the atmospheric analysis fields. These less cloud affected analysis fields further improve hurricane track and intensity forecast. The availability of satellite observations that can be assimilated in the model is limited if only the clear radiances are assimilation. An effective way to use the thermodynamic information under partially cloudy regions is to assimilate the “cloud-cleared” radiances (CCRs); CCRs are also called clear equivalent radiances. Because the CCRs are the equivalent clear radiances from the partially cloudy FOVs, they can be directly assimilated into the current data assimilation models without modifications. The AIRS CCRs are assimilated and compared with the AIRS using stand-alone cloud detection and collocated cloud detection. The assimilation of AIRS cloud-cleared radiances directly affects the atmospheric fields in the surroundings away from the hurricane center at the analysis time. With the improvement of the atmospheric fields, the assimilation of cloud-cleared AIRS gives the smallest root mean square error (RMSE) of hurricane tract 72-hour forecast compared to the AIRS radiances using stand-alone cloud detection and collocated cloud mask cloud detection.