Extended Abstract (128K)P3.6
4D assimilation of cloudy satellite radiances
T. Vukicevic, CIRA/Colorado State University, Ft. Collins, CO; and M. Sengupta, T. Vonder Haar, A. S. Jones, and K. F. Evans
Statistical cloud properties are required to test and improve cloud parameterizations in weather prediction and climate models and to study role of clouds in the atmospheric system. The cloud properties are also desired in initial condition in the weather prediction. Cloud property retrievals from site observations (e.g., the Atmospheric Radiation Measurements) and global satellite remote sensing measurements (e.g., International Satellite Cloud Climatology Project) and simulations of cloud resolving models (CRMs) have been used to estimate the statistical cloud properties. Neither of the approaches have been fully successful. Studies have shown that the retrievals at the single site are not representative of large domains, while the cloud analysis based on the global satellite measurement retrievals do not have sufficient resolution to represent high variability at spatial scales which control the cloud properties. Furthermore, other information required for parameterization work and understanding of cloud evolution, such as advection of cloud profiles and the relationships between clouds and 3D dynamics are not available from these analyses. The CRM simulations, on the other hand, by definition represent cloud microphysical and the associated dynamical properties but are poorly constrained if at all with cloud observations. The required analysis of cloud properties could be improved considerably with aid of objective 4D data assimilation which involves simultaneous utilization of both the cloud sensitive remote sensing measurements and cloud resolving model. The purpose of data assimilation, in general, is to use all available information to determine as accurately as possible state of the system under study (i.e. the Atmosphere with clouds). The all available information consists of the observations and models. We developed a 4D variational (4DVAR) data assimilation numerical algorithm for the Regional Atmospheric Modeling System (RAMS) with cloud resolving capability. This project supported by the Army Research Lab is intended to improve high resolution atmospheric state estimation with clouds in 4D. The current applications of the new 4DVAR system emphasize assimilation of cloudy satellite radiances from GOES in visible and IR wavelengths. The early results of experiments over the U.S. show skill and feasibility of cloud resolving data assimilation.
Poster Session 3, Data Assimilation
Tuesday, 21 September 2004, 9:30 AM-11:00 AM
Previous paper Next paper