Probabilistic Forecasting of Snowfall Amounts Using a Hybrid between a Parametric and an Analog Approach

Forecast uncertainty associated with the prediction of snowfall amounts is a complex superposition of the uncertainty about precipitation amounts on the one hand, and the uncertainty about weather variables like temperature and wind that influence the snow-forming process on the other hand. While Scheuerer and Hamill (2015) have demonstrated that their parametric, censored, shifted gamma distribution (CSGD) approach compares favorably with the non-parametric analog method when the same amount of training data is available and the focus is on more extreme precipitation amounts, the complexity of the relation between temperature and wind with snowfall amounts conversely makes the very flexible analog method an attractive choice.
In this presentation we show how these two different concepts - a parametric, regression-type approach and the analog method - can be combined in a way that leverages the respective advantages. Predictive distributions of precipitation amounts are obtained using the CSGD method, and quantile forecasts derived from it are then used together with ensemble forecasts of temperature and wind speed to find historical dates where these quantities were similar. The corresponding, observed snowfall amounts on these dates can then be used to compose an ensemble that represents the uncertainty about future snowfall. We demonstrate this approach with reforecast data (Hamill et al. 2012) from the Global Ensemble Forecast System (GEFS) and NOHRSC snowfall analyses.

References:

Hamill, T.M., Bates, G.T., Whitaker. J.S., Murray D.R., Fiorino, M.., Galarneau, Jr., T.J., Zhu, Y. , and Lapenta, W. (2013): NOAA's second-generation global medium-range ensemble reforecast data set. Bull Amer. Meteor. Soc., 94, 1553-1565.

Scheuerer, M. and Hamill, T.M. (2015): Statistical post-processing of ensemble precipitation forecasts by fitting censored, shifted Gamma distributions. Monthly Weather Review, 143(11), 4578-4596.