Tuesday, 31 July 2001
Analysis of the mesoscale precipitation band associated with the 24-25 January 2000 storm
The blizzard of 24-25 January 2000, that produced record snowfall totals in areas stretching from the Carolinas to New England, was handled very poorly by operational numerical weather prediction models. The surface cyclone associated with the blizzard was poorly forecast with errors of several hundred kilometers too far to the east in placement of the center, particularly with lead times in the range of 72 to 96 hours. With diminishing lead times, the forecasted positions of the cyclone at 1200 UTC 25 January 2000 were within 100 kilometers of the analyzed position. Despite improvements in the cylcone position, the forecasted distribution of precipitation remained too far east.
In this presentation, the results of an analysis of the intense mesoscale precipitation band associated with the event will be presented. The relationship of this band to the cyclone position and intensity will be described and the forcing mechanisms for the precipitation band will be
diagnosed. In addition, the NCAR/Penn State non-hydrostatic modeling system (MM5) and the MM5 adjoint will be used to study the sensitivity of errors in both the location and the forcing of the precipitation band to the initial condition of the model forecast. In addition, the structure and evolution of singular vectors calculated for this event will be diagnosed to assess synoptically and dynamically the effect improvements to the initial conditions would have on the forecast of the event.