SECTION 1 PROJECT OBJECTIVES AND ACCOMPLISHMENTS
The objectives of this partners project were to analyze the meteorological conditions that led to the development of the rain event of 20-21 June 1997 in Milwaukee and to assess the capability of a very high resolution mesoscale model to predict these conditions at lead times of 00-18h, 12-30h and 24-42h (using forecasts initialized at 0000 UTC 21 June, 1200 UTC 20 June and 0000 UTC 20 June, respectively). The results from these very high resolution simulations were compared to that obtained from the then current suite of operational models available from the National Center for Environmental Prediction (NCEP).
Project participants were Paul Roebber, Associate Professor in the University of Wisconsin at Milwaukee Department of Geosciences and John Eise, Science and Operations Officer (SOO) at the NWS/MKX.
The contributions of and collaborations between UWM and the NWS were constituted by the work of the two project PIs (Roebber and Eise, respectively). Roebber (with assistance from student hourly help) conducted the analysis of the surface and upper air data and performed the MM5 simulations. Eise obtained the WSR-88D level II archive data, Wisconsin mesonetwork data, PCGRIDDS data and high resolution satellite data. Synthesis of all these data was accomplished through a series of person-to-person meetings held at UWM. Additional communications were conducted via telephone and e-mail. Preparation of a manuscript for submission to an AMS journal (Weather and Forecasting) is underway and is expected by fall 1999. Eise is collecting the study datasets for submission to COMET.
SECTION 2 SUMMARY OF UNIVERSITY/NWS EXCHANGES
Professor Roebber participated in a portion of the AWIPS training program held at NWS-MKX in April 1999. This program was useful as a view towards the future use of model datasets by NWS forecasters.
SECTION 3 PRESENTATIONS AND PUBLICATIONS
Professor Roebber has presented preliminary results from this research to the American Society of Mechanical Engineers (December 2: 100 Years of Experience in Forecasting the Weather; Milwaukee, WI) and the NWS-Green Bay forecast office (December 7: The realtime MM5 and operational utilization of mesoscale models; Green Bay, WI). Additional aspects will be presented in June 1999 at the MM5 Users' Workshop: The role of Lake Michigan in modulating upstream weather conditions: Experience with forecasts using MM5 in warm and cold seasons and priorities for future model development.
SECTION 4 SUMMARY OF BENEFITS AND PROBLEMS ENCOUNTERED
4.1 University's Perspective
The primary benefit to the University has been a formal working relationship with the NWS office in Sullivan. This has improved interactions between our groups and eased the exchange of scientific data and ideas. This collaboration has also led to the clearer identification of regional forecast problems to researchers at UWM.
As a result of this information, a full-scale evaluation of the operational utility of the UWM realtime MM5 system will be undertaken under a COMET Coop grant with NWS-GRB. This evaluation will allow us to determine for specific forecast scenarios the strengths and limitations of very-high resolution (6 km) model forecasts in the Great Lakes region.
The most significant problems relate to inherent limitations in the scope of work imposed by the COMET partners program. For example, funding for partners programs, although by design much more limited in scope than the Coop program, should be increased to allow more substantial participation by graduate students.
4.2 NWS Perspective
With the information gleaned from the mesoscale model runs of the June 20-21, 1997 Great Milwaukee Flood, real-time access to this information would have proven useful in a more timely decision on warning for flooding. This work has also demonstrated that mesoscale modeling, to be most effectively employed by forecasters for convective phenomena, will need to be done at the 1-2 km scale. To accomplish such modeling will require substantial infusion of computational resources at some level in the NWS.
The University of Wisconsin-Milwaukee has put online a web page with the near
realtime run at 0000 UTC of the MM5 model. Also, VIS5D files are being FTPd
to the NWS office in Sullivan for their use. These datasets are potentially
of use to NWS forecasters. However, Internet bandwidth continues to be problematic,
as the time necessary to download files/transfer information from the Central
Region servers to the Sullivan forecast office slows down the incorporation
of these data into routine usage by NWS forecasters.