Jordan Dale shared his initial findings from two WRF simulations of TC Ernesto initialized at 06Z and 12Z 8/30/06 using 1̊ resolution GFS forecast data. The 12Z simulation was run on a 36-km grid with nested 12 and 4-km domains. Both WRF simulations used 47 vertical levels up to 50 mb, WSM 6-class graupel MP scheme, MYJ TKE BL scheme, the Noah Land-Surface Model, and the BMJ CP scheme (for the 36 and 12-km grids; no CP scheme was used for the 4-km grid). Results from the two simulations both showed a weak, marginal representation of the PRE ahead of landfall and bad track error with a consistent westward bias.
Experiments are currently under way using different initialization times with GFS forecast data, ECMWF ERA-Interim data, GFS Analysis data, and different model physics schemes. The purpose of these simulations is to produce a control run that replicates the actual storm. Once an adequate control run is simulated, the importance of the PRE that occurred on 8/30/06 ahead of landfall will be investigated.
The next call is scheduled for Wednesday, August 15th .