Document Type: Original Research Paper
Department of Geography, Geology, and the Environment, Slippery Rock University, Slippery Rock, PA 16057, USA
The climate and weather patterns of Buffalo (New York, U.S.A.) are strongly influenced by the
city’s proximity to Lake Erie. Total monthly snowfall in Buffalo is forecasted using neural network techniques
(Multi-Layer Perceptron = MLP) and a multiple linear regression (LR) model. The period of analysis comprises 28 years from January 1982 to December 2009. Input data include: zonal wind speed (u-wind), meridional wind speed (v-wind), air temperature, the geopotential height (GPH) over Lake Erie and the surrounding regions at the 1000 mb -, 925 mb -, 850 mb -, and 700 mb - levels as well as the surface pressure and air temperature, mean water temperature, lake surface water temperatures (LSWT) and the amount of ice coverage of Lake Erie; the 500 mb GPH over James Bay, Canada; and the surface pressure over the North-Central Great Plains. Different lead times of the input variables are tested for their suitability. The most accurate result is obtained by using the MLP with an optimum lead time approach (lead times vary for the different input variables between one and six months). The results of the MLP with six months lead time are in good agreement with observed precipitation records over the study period.