Use of a GIS application to evaluate the accuracy of
forecaster and model predictions of snowfall in eastern New York and western
New England (265)
Michael Evans, NOAA /
NWS, Albany, NY
Joe Villani, NOAA /
NSWS
Vasil Koleci, NOAA /
NWS
Charle Gant, NOAA / NWS
Abstract:
A collaborative project
between the National Weather Service Forecast offices in Albany, New York and
Greenville-Spartanburg, South Carolina has resulted in the development of a
GIS-based application that produces high-resolution analyses of observed
snowfall, and calculates errors and biases of corresponding gridded snowfall
forecasts based on the observations. This presentation will introduce the
application, and summarize the results of multiple associated projects that
examine patterns of observed snowfall, and the accuracy and biases of various
corresponding snowfall forecasts.
An evaluation of several
years of observed snowfall in eastern New York and western New England
indicates that observed snowfall is often strongly tied to terrain features,
depending on characteristics of the lower-tropospheric flow pattern. In
addition, it appears that the accuracy and bias of National Weather Service
snowfall forecasts in this area may also be dependent on interactions between
lower-tropospheric flow and terrain. Preliminary results from a comparison of
observed snowfall and forecasts from high-resolution models will also be shown,
to test whether the models exhibit biases similar to the official forecasts.