An investigation of lake-effect snow warning size in relation to snowfall extent

Thumbnail Image
Jaszka, Keith D.
Call, David A.
Issue Date
Thesis (M.S.)
Department of Geography
Other Identifiers

Every late autumn and winter, lake-effect snow (LES) disrupts the livelihoods of those who reside in the lee of the Great Lakes. This phenomenon often generates significant localized snowfalls that can cripple one section of a community while the remainder is spared. Detrimental impacts on travel from snowfall rates of 1 inch hour-1 or greater are not uncommon. Storm-total snowfall gradients of 6 to 8 inches mile-1 have also been documented. Twenty four-hour snowfall measurements taken by National Weather Service (NWS) observers during Lake Ontario LES events were collected. Utilizing a geographic information system, snowfall was interpolated from spotter reports with the kriging method. Then, the area of warning-criteria snowfall was compared to the area of the corresponding county-based LES warning for calculation of percent of false alarm. This warning performance evaluator identified the proportion of a county-based LES warning that received sub-warning-criteria snowfall and was therefore unnecessary. A total of thirteen events from the 2009-2010 through 2011-2012 LES seasons were analyzed. The average percent of false alarm was approximately 90%. This suggests the NWS should consider smaller warning polygons. Limited-size warnings would enable meteorologists to provide greater specificity regarding the location of anticipated significant snowfall, while simultaneously reducing the number of people who are unnecessarily warned.