On the use of 1-min ASOS data to predict the onset of snow

The onset of precipitation and precipitation type is important for aviation applications, specifically for aircraft deicing during winter storms. This study examines the use of 1-min ASOS data to predict the onset of snow. Cloud base height, visibility, and dewpoint temperature depression are investigated as potential predictors. Various methods of temporal smoothing are used to reduce variance in the cloud base and dew point depression data. It is shown that the variance in visibility, however, may be a useful predictor. Nearly 50 episodes of the onset of snow are examined using ASOS data from the east coast and Denver areas. About half the cases show a decrease in the cloud base height prior to onset of snow, with as much as 30 min of lead time. Variance of visibility is also a good predictor but with less lead time. Dewpoint temperature depression rarely falls before the onset. Details of the data analysis, including attempts to identify freezing rain, will be shown. In addition, efforts to compare the ASOS data to numerical weather prediction data will be described.