Winter snowfall prediction in the United States using multiple discriminant analysis

Abstract

This study seeks to determine the skill of multiple discriminant analysis for predicting seasonal snowfall. Winter total snowfall amount and frequency of snowfall events are examined for 440 stations in the United States from 1930 to 2006. The independent variables used to create the forecast include ocean-atmosphere teleconnection patterns [such as the Pacific Decadal Oscillation (PDO) and El Niño Southern Oscillation (ENSO)], large-scale atmospheric patterns [such as the Arctic Oscillation (AO), North Atlantic Oscillation (NAO) and Pacific/North American (PNA)], land cover (such as Arctic sea ice extent and Eurasian snow cover extent), and temperature. Based on a jackknife analysis, forecasts are correct 20-80% of the time for categories of 'below normal', 'near normal', and 'above normal'. When broader categories are used of 'normal or below', 'near normal', and 'normal or above' the forecasts are correct as much as 90% of the time at some stations. The Central United States, Ohio River Valley, Great Lakes, and Upper Midwest regions show the highest level of skill. Results not only confirm relationships previously documented between atmospheric phenomena and US snowfall (such as with the PNA, NAO, and ENSO), but also expand our understanding of factors that influence decadal-scale snowfall variation (such as Arctic sea ice extent and Eurasian snow cover extent).

Spatio‐temporal November and March snowfall trends in the Lake Michigan region

“…There are also lagged correlations among snowfall and teleconnection indices, as found by Kluver and Leathers () in their study utilizing teleconnection patterns for seasonal snowfall prediction. For the Lake Michigan region, November composite snowfall is significantly correlated with the October NAO, while the March snowfall is anti‐correlated with the February PDO (Table ).…”

Regionalization of snowfall frequency and trends over the contiguous United States

“…The teleconnection indices used are selected due to their impact on either the location of precipitation or temperature anomalies that influence snowfall in the United States. They are also selected because they were identified by Kluver and Leathers (2014) as having significant correlations with individual snowfall stations. Included in the analysis are: the Arctic Oscillation (AO) (Thompson and Wallace, 2000), the North Atlantic Oscillation (NAO) (van Loon and Rogers, 1978;Wallace and Gutzler, 1981;Hartley and Keables, 1998;Bradbury et al, 2002;Morin et al, 2008;Climate Prediction Center, 2010;Ghatak et al, 2010), Nino 3.4 region sea surface temperatures (Rasmusson and Wallace, 1983, Ropelewski and Halpert, 1986, Kunkel and Angel, 1999Smith and O'Brien, 2001;Patten et al, 2003;Climate Prediction Center, 2010), the Pacific North American index (PNA) Gutzler, 1981, Leathers et al, 1991;Serreze et al, 1998;Notaro et al, 2006;Coleman and Rogers, 2007;Morin et al, 2008;Climate Prediction Center, 2010;Abatzoglou, 2011), and the Pacific Decadal Oscillation (PDO) [Mantua and Hare, 2002; Joint Institute for the Study of the Atmosphere and Ocean (JISAO), 2010].…”

Influence of regional Arctic sea ice extent on lagged snowfall in the contiguous United States

“…The establishment of a positive AO during high Kara Sea SIE years produces higher pressure anomalies over the mid-latitudes, which shifts storm tracks to the north as they cross the Pacific (Thompson and Wallace, 2000). This results in more snowfall in the western United States and less snowfall in the eastern United States (Kluver and Leathers, 2015).…”

“…Some examples of these established relationships are between US snowfall and the North Atlantic Oscillation (NAO), PNA, PDO, and ENSO (Hartley and Keables, 1998;Serreze et al, 1998;Kunkel and Angel, 1999;O'Brien, 2001, Bradbury et al, 2003;Patten et al, 2003;Hunter et al, 2006;Notaro et al, 2006;Coleman and Rogers, 2007;Morin et al, 2008;Kunkel et al, 2009;Kluver and Leathers, 2015). Kluver and Leathers (2015) also included surface land cover when they used Arctic sea ice extent (SIE) and Eurasian snow cover in a multiple discriminant analysis of winter snowfall frequency and amount. This corroborates modelling results from Deser et al (2010) and Liu et al (2012) which proposed a connection between Arctic sea ice and northern Canadian/US snowfall during winter months.…”