Uncertainty using High Resolution Ensembles & Probabilistic Datasets for an
Anomalous Early Season Snow Event in New England
Frank Nocera, NOAA/NWS Weather Forecast Office Boston, MA, Taunton, MA
Communicating Uncertainty using High Resolution Ensembles & Probabilistic Datasets for an Anomalous Early Season Snow Event in New England
Frank M. Nocera
NOAA/National Weather Service Boston, MA
An anomalous early season snow event impacted southern New England during the afternoon and evening hours of 27 October 2016. Snowfall amounts of up to 6 inches were reported across the high terrain (above 1500 ft) of western Massachusetts with accumulating snow observed all the way down to the valley floor of the Connecticut River Valley, including the highly populated city of Springfield, MA. Given trees were still partially leaved the weight of the heavy snow resulted in localized areas of down tree limbs and isolated power outages.
Unlike snowstorms in the middle of the winter season (December, January & February), this storm presented uncertainty with snowfall amounts given the lack of cold air resulting in concerns regarding precipitation type and intensity, duration of snow and accumulations on paved vs unpaved surfaces given the time of year. High resolution ensembles and associated impact graphics along with operational model guidance showed the potential for an anomalous early season snow event in the days leading up to 27 October.
As part of its Decision Support Services program, the National Weather Service (NWS) in Boston provided email briefings to its core partners and social media posts to the general public in the days leading up to this snow event. However, there were challenges in communicating the uncertainty in the forecast even 6-12 hours before the snow began.
This presentation will focus on the science behind this anomalous event including a review of how high resolution ensembles and associated impact graphics increased forecaster confidence on a reasonable “worst” case scenario becoming more likely. In particular the use of experimental probabilistic snowfall forecasts will be discussed.