Thursday, 26 January 2017: 11:15 AM
607 (Washington State Convention Center )
Scientists from Cooperative Institute of Meteorological Satellite Studies (CIMSS) at University of Wisconsin-Madison have developed a near realtime regional Satellite Data Assimilation system for Tropical storm forecasts (SDAT) (http://cimss.ssec.wisc.edu/sdat) since 2013. The system is built with the community Gridpoint Statistical Interpolation (GSI) assimilation and advanced Weather Research Forecast (WRF) model. With GSI, SDAT can assimilate all operational available satellite data including GOES, AMSUA/AMSUB, HIRS, MHS, ATMS, AIRS, CrIs and IASI radiances and some satellite derived products. In addition, some research products, such as hyperspectral IR retrieved temperature and moisture profiles, GOES imager atmospheric motion vector (AMV) and GOES sounder layered precipitable water (LPW), can also be assimilated in the system. Using SDAT as a research testbed, studies have been conducted on better handling the different satellite data to improve the high impact weather forecast. A couple of years’ realtime experiments show that 72 hours tropical storm track forecasts in the Atlantic Ocean region are reasonably well compared with that from the operational model forecasts. However, there is often a large error of track and intensity forecast at the initial time due to the lack of a special vortex initialization scheme. We have recently added a vortex relocation algorithm into SDAT system. The algorithm is based on WRF eta-surface and model horizontal grids, and has improved vortex isolation near the surface. Tests from last year’s hurricane cases show that the first 18 hours storm track forecasts have been greatly improved. We are also working to develop a dynamical initialization scheme to improve the initial vortex intensity forecast. The impact of vortex initialization on satellite data assimilation is also investigated, results from these studies will be presented at the meeting.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner