We looked at some of the most disruptive Thanksgiving snowstorms of years past in Part I. Now, it’s time to turn forward. Let’s see which way the signs point for this year’s traditional celebrations.
Snowstorm Signals for the Peak Travel Days of 2019
Using NCEP/NCAR Reanalysis 1 data courtesy of ESRL*, I’ve looked at the percentage of years with disruptive snowfall conditions, comparing all years 2000-2018 with a selection of ‘analogue’ years**.
Due to a lack of direct snowfall observation data, disruptive snowfall conditions have had to be estimated. This was using a combination of temperature and precipitation. The results presented below (Figure 1) aren’t to be taken too literally! Their main purpose is to snow how the ‘analogue’ years compare to 2000-2018. From this, signals can be inferred with respect to Thanksgiving 2019.
Figure 1: Maps showing the percentage of years with disruptive snow conditions in all years 2000-2018 (left) and in a selection of ‘analogue’ years (right). Qualifying days have a max temp of no more than 3°C, a diurnal mean temp less than or equal to 0°C, and a total precipitation of at least 0.3 mm (approx. 3 inches snowfall assuming 1:10 water to snow ratio). These results are for broad-scale guidance only; do not use them for local-scale decision-making.
Reasons to Be Cautious
Without taking the details too literally, there are some noteworthy differences between the two plots.
The ‘analogue’ years suggest a much higher than usual risk for elevated (about 1200 feet+) north-western parts of the Midwest, along with most of Oklahoma, a lot of Missouri, and part of Illinois.
Meanwhile, a slightly above normal risk is suggested for elevated parts of the Northeast and Mid-Atlantic states. The same is true for the northwest of the West (mainly Washington and Oregon), away from the Pacific coast.
By contrast, Utah, Wyoming and Colorado are proposed to see a lower risk than usual.
Flanking Lows Drive Increased Threat of Snows
By looking at a plot of 500 hPa geopotential height (GPH) anomalies (Figure 2), I can deduce what the ‘analogue’ years are suggesting for the overall movement of the atmosphere compared to usual. How, you ask? Well, that can be the subject of a dedicated blog piece in future 😉. These then give me explanations for the differences in snow risk pattern being suggested by those years (compared to all years 2000-2018).
Figure 2: Plot of mean 500 hPa geopotential height anomalies for 27th-28th Nov, taken across the ‘analogue’ years only, annotated with the associated anomalous atmospheric movements that are responsible for the snow risk patterns suggested by those years.
Broadly, the positioning of areas of below-normal 500 hPa GPH either side of the U.S. suggests an above-normal chance that areas of low pressure will be pushing moisture-laden air into the Northwest, Northeast and Mid-Atlantic, or both. Not the best of patterns to have in place during Thanksgiving.
Meanwhile, the slightly above-normal 500 hPa GPH over the eastern Northwest implies a higher than usual chance that cold air will move from Canada across the Midwest and on down toward the Southeast. Associated cold weather fronts can bring significant snow, especially to elevated areas.
Readers may now be contemplating what may unfold should that cold air interact with an area of low pressure pushing moisture across the Northeast and Mid-Atlantic states. Yes – this is a pattern that brings an increased risk of a nor’easter storm! I’ll certainly be on the lookout, in the coming few weeks, for any sign of such an event from forecasting models (e.g. the GFS and the ECMWF model) and other guidance tools.
To my readers in the United States – wherever you are and however you do it, I wish you a Thanksgiving to remember fondly! I can already smell the pumpkin pies from here… on the other side of the pond!
James Peacock MSc
Head Meteorologist at MetSwift
* Ideally, this would have been done using data from weather stations, but there are thousands of those. Analysing all of those would be unreasonably time-consuming for a one-time use in a blog piece. Although the reanalysis data is relatively coarse, it’s still of high enough resolution to provide useful regional-scale information.
** This is a selection of historical years from 1950-2018. They’re identified by looking at key drivers of the global weather patterns. In the analogue years, these behaved most similar to what’s expected during the guidance period.