Date: Wed, 01 Apr 1998 11:09:26 -0600 From: "Paul J. Roebber" Organization: University of Wisconsin at Milwaukee Mime-Version: 1.0 To: eegle-mod@wings.glerl.noaa.gov Subject: New MM5 run for Mar 9 at 1600 UTC X-Attachments: 1998/attach-19980401.gif Attached you will find a new gif file of the MM5 wind field for 1600 UTC on 9 March 1998 for comparison with the GLCFS Winds and the earlier MM5 simulation. This map is also produced from a 4h forecast on a 12 km model domain using the 1200 UTC AVN initial analysis and the contours in the center of Lake Michigan are 20 m/s (with a contour interval of 2 m/s). The winds have been reduced to the 10m level using similarity theory and the landuse and terrain resolution has been increased (used 0.9 km data interpolated to the 12 km grid). In general, the biggest change is the removal of the "blocky" structure caused by the coarse grain landuse that was used previously. The 20m/s contour now extends most of the major axis of Lake Michigan and pushes closer to the shoreline. 1) Continued good model matches with obs are: station obs old.model new.model 45002 37kt 37kt 37kt GSRW3 39kt 37kt 37kt ESC 19kt 23kt 23kt MDW 19kt 23kt 23kt MNW 23kt 25kt 25kt TVC 25kt 23kt 25kt GRB 26kt 25kt 27kt 2) Improved model matches are: station obs old.model new.model SGNW3 41kt 27kt 35kt 20C 40kt 27kt 34kt CGX 37kt 27kt 33kt 3) Worsened model matches are: station obs old.model new.model 17C 24kt 27kt 35kt 4) Still unsatisfactory matches are: station obs old.model new.model Y09 15kt 25kt 27kt The only additional modification that can be done (with these data; this does not preclude running the case using the actual obs as boundary conditions) is to run the model at higher resolutions (e.g. 6 km, 4 km etc.). I'd like to hear your thoughts (especially Dave). Paul Roebber