Exploring Snowfall Predictability with the HRRR model for the January 16-17, 2022 Mixed Precipitation Event in Pennsylvania

Open Access
- Author:
- Myskowski, Eric
- Area of Honors:
- Meteorology
- Degree:
- Bachelor of Science
- Document Type:
- Thesis
- Thesis Supervisors:
- Steven J Greybush, Thesis Supervisor
Chris E Forest, Thesis Honors Advisor - Keywords:
- meteorology
winter storms
predictability
precipitation transition
HRRR - Abstract:
- Snowfall predictability in a mixed precipitation event is analyzed, with an emphasis on the effects of the forecasted timing and extent of the transition from snow to other precipitation-types on snowfall forecasts. This is done by comparing High Resolution Rapid Refresh (HRRR) model forecasts to observations during a winter storm that moved through Pennsylvania on January 16-17, 2022. During this storm, snowfall amounts throughout Pennsylvania were significantly overpredicted by the HRRR, with areas to the west of the storm track having the most error. Mapping the observed and forecasted precipitation transition timing shows that precipitation changed from snow to a wintry mix quicker than the HRRR forecasted, especially in areas to the west of the center of the low. These western areas also remained as sleet and freezing rain for longer than areas to the east of the low, where precipitation changed to rain after changing over from snow. Comparing the transition times at specific locations made a more quantitative analysis possible, showing that the forecast error was only in the range of one to three hours and occurred during the time of the heaviest precipitation. Analyzing soundings showed that temperatures around 850 mb were higher than forecast around the time of the transition. After the HRRR forecasted the precipitation to change over, it added additional snowfall accumulation in some locations, increasing the forecast snowfall error. An analysis of the synoptic scale progression of the storm discovered that elements such as the track and strength of the low pressure, as well as the position and strength of the high pressure to the north supplying the arctic air, were forecasted well and likely were not the primary causes of snowfall error. Overall, quicker than expected transition to sleet was a source of error, especially at initialization times farther from the storm arriving in Pennsylvania.