Ongoing development and testing of generalized cloud analysis package within GSI for initializing Rapid Refresh
Ming Hu, ESRL Global Systems Division, Boulder, CO; and S. S. Weygandt, S. Benjamin, and M. Xue
An initial version of a generalized cloud analysis package which combines the strengths of the RUC and ARPS cloud analysis packages has been developed within the GSI for the purpose of initializing the Rapid Refresh, which will employ a version of the WRF model as a replacement of the current operational RUC. This package will be a key component in the initialization of ongoing cloud and precipitation systems within the hourly cycled Rapid Refresh. As with the RUC, the Rapid Refresh cloud analysis is critical for the accuracy of short-range ceiling and visibility forecasts, which are extremely important for aviation.
We have completed a preliminary set of tests of the cloud analysis packages using the 13 March 2006 case (including widespread stratiform clouds and a central US squall line) under the planned Rapid Refresh configuration, but on a smaller CONUS domain. These experiments are a prelude to hourly cycled real-time tests over the full CONUS domain (covering all of North America), which will begin this fall. Particular emphasis in these real-time tests will be on evaluating and modifying the cloud package to improve analysis and prediction skill for cold season clouds and icing. This cold season evaluation will complement ongoing warm season tests focused on improving forecasts of convection.
Recent work has included improving the efficiency of the generalized cloud analysis package through parallelizing the data ingest and cloud analysis itself. Future work will include adding new observation data sets to the GSI cloud analysis (lightning and satellite derived water paths) and coupling the radar reflectivity portion of the cloud analysis to the WRF model pre-forecast integration to better initialize convection. At the conference, we will describe the cloud analysis in detail, present verification statistics, and show specific case study illustrations.
Extended Abstract (1.7M)
Session 7, Nowcasting and Modeling Part III
Tuesday, 22 January 2008, 1:30 PM-3:00 PM, 226-227
Previous paper Next paper
Browse or search entire meeting
AMS Home Page