Close Window

University of Wisconsin (Brad Hoggatt, COMET graduate student fellow): "Assessment of the potential for mesoscale numerical prediction of lake effect snows"

Final Report

An important goal set by the United States Weather Research Program is to develop technologies capable of predicting precipitation on the meso-beta scale and on time scales of 24-48 hours. In the near future, it is expected that numerical models will increase in resolution to the point where these variations can be resolved. Unfortunately, the conventional observing system only resolves three-dimensional variations on the largest of the meso-alpha scales. Although these observations provide insufficient information for conventional model initialization on the mesoscale, the great hope has been that much of the mesoscale variability is derived from the interaction between synoptic scale flows and small-scale surface variability. The Great Lakes lake effect snow storms are one such example of this interaction. Although the principal causes of lake effect snows have been extensively studied, it is still not well known to what extent specific mesoscale numerical forecasts can be used by the operational forecaster and to what extent the numerical forecast represents a single realization of many possible solutions to a chaotic flow regime.

The purpose of this project was to understand the practical limits of the mesoscale predictability of a nonhydrostatic model in lake effect snow situations. For his Masters thesis work, Mr. Hoggatt collaborated with the Sullivan NWS office to numerically investigate two lake-effect snowstorms which exhibited peculiar structural variability. Using initial and boundary conditions provided by the Eta model, high resolution, multiple, two-way nested UW-NMS simulations with a minimum grid spacing of 4.5 km were performed. These simulations were able to capture much of the small scale variability (i.e., gravity wave initiation, banded precipitation structures, transient sinusoidal disturbance), despite the use of initial conditions deficient in meso-beta scale information. The results indicate that lake-effect snow structures, including the small scale variations, are likely to be predictable out to several days provided the large scale forecast is accurate.

Mr. Hoggatt also presented several seminars at the Sullivan NWS office as part of his work. Topics included a review of basic thermodynamics and dynamic principals, differences between hydrostatic and nonhydrostatic models, and the use of various diagnostic tools is a means to better understand fundamental forcing mechanisms.