December 15, 2005 Winter Storm Research

December 15, 2005
Winter Weather Event Review
Section 4: Conclusion

Back to the main research page for December 15, 2005

Section 1 -- Section 2 -- Section 3 -- Section 4

Forecasting precipitation type and amounts for Thursday morning, December 15, 2005 was very challenging. The GFS and NAM were in good agreement with the mass fields prior to the event. However, subtle differences in the forecasted lowest 300 mb thermal profiles and the strength of forcing made it difficult to determine both the precipitation type and amount of snowfall much before the onset of the event. In hind sight, the GFS handled the precipitation type and amounts just a bit better than the NAM. Both the area and energy methods suggested Dayton would quickly transition to snow around 12z. Interpreting the lowest 300 mb of the forecast sounding confirmed the quick transition to snow around 12Z with all but a small portion of the sounding being below freezing. The GFS did out perform the NAM in the overall 1000-850 mb thickness analysis by projecting the 1300 m contour just south of where the highest snowfall was observed. The toughest part of the precipitation type forecast was being able to properly assess how the boundary layer thermal profile would transition.

Forcing played a significant role in what type of precipitation fell. Strong upper level forcing enhanced by 700 mb deformation and frontogenesis contributed to evaporative and/or dynamical cooling of the lowest 300 mb of the atmosphere during the event. This cooling ensured the precipitation fell as wet snow to the northwest of I-71 instead of a rain/snow mix or just plain rain.

Perhaps the most difficult part of this event was forecasting how much snow would fall. Both models suggested around a third of an inch of liquid precipitation in the plan view using AWIPS. WWD suggested 2 to 4 inches of snow. However, the high resolution output using BUFKIT suggested around a half inch of liquid precipitation was expected at Dayton by both models. The Bourgouin PType (area) method suggested that most of the this precipitation would fall as snow. Would precipitation rates be strong enough to counteract diurnal daytime heating of the the near surface boundary layer? How much snow would actually fall with such a warm moist sounding which was just below freezing with surface temperatures expected to be just above freezing? The answer to the first question appears to be yes. The second question is a bit more problematic. In hind sight, it appears the the fine details of the models handled snowfall amounts well when assuming less than a 10 to 1 ratio. The coarser fields such as 1000-850 mb thicknesses and plan view precipitation suggesting more of a mix rain/snow event, especially in the beginning, with much less snowfall accumulation expected. However, as noted in the Meteorological Assessment section, we observed that looking at the finer details is essential. In the end the threat area received 3 to 5 inches of snow, which was closely associated with the the axis of best 700 mb deformation and 2D frontogenesis and oriented itself in a pocket of cooler low level thicknesses between 12Z and 18Z. In-depth analysis and trend assessment, using model sounding data, was show to be extremely valuable in this case.

In summary, forecasting the precipitation type and amounts for this event was problematic. There were several subtle signals up to 12 hours before the onset of the event suggesting a quick transition to snow around 12z. However, forecasting the amount of snow was difficult as surface temperatures were just above freezing and the lower 300 mb was saturated and just below freezing. This suggests the snow would be wet and melt/compact quickly upon landing.