17B.4
A Portable Mesoscale Prediction System for MODIS Direct Broadcast Sites
Robert M. Aune, NOAA/NESDIS/ORA, Madison, WI; and K. I. Strabala, S. S. Lindstrom, and A. Huang
A version of the PC/Linux based CIMSS Regional Assimilation System (CRAS) forecast model, at the Cooperative Institute for Meteorological Satellite Studies (CIMSS), has been adapted to assimilate retrieved products from the Moderate Resolution Infrared Spectroradiometer (MODIS) instrument onboard NASA Aqua and Terra satellites. The package is designed for use at MODIS Direct Broadcast (DB) sites using the International MODIS/AIRS Processing Package (IMAPP). Referred to as DBCRAS, the project adds an optional satellite data assimilation capability to the 200+ existing DB sites around the world, many in remote areas. IMAPP software processes MODIS radiances into meteorological parameters that are assimilated directly at the local DB site. A prototype DBCRAS has been running at CIMSS in Madison, Wisconsin since April 2007. It assimilates retrievals of cloud fraction, cloud-top pressure and total precipitable water (5 km) processed from MODIS radiances received by our own X-band antenna in real time (see http://cimss.ssec.wisc.edu/model/realtime/cras48_WI/daily.html). It has been shown that assimilating MODIS products has a positive impact on the DBCRAS cloud and precipitation forecasts. The data are assimilated into a 12-hour spin-up forecast that automatically searches for MODIS passes as they occur. Temperature, water vapor and cloud fields in the CRAS are adjusted in the process and then combined with the latest winds and temperatures from the NCEP Global Forecast System to complete the set of forecast initial conditions. A 72-hour forecast (48 km resolution) is generated on a typical Linux workstation in less than five hours time. The system outputs files in GRIB2 format and includes a version of the Man-Computer Interactive Data Access System (MCIDAS) that is used to create web page graphics. Besides North America, DBCRAS has routinely produced forecasts for Alaska, Antarctic, Beijing, Italy, Kazakhstan, Taiwan, South Africa, and Norway. A summary of DBCRAS accomplishments and future plans will be presented.
Session 17B, Forecasting Tools/Techniques Part III
Friday, 5 June 2009, 8:00 AM-9:00 AM, Grand Ballroom West
Previous paper