Friday, November 21, 2014

Thanksgiving Potentially Strong Storm System

I'm still watching for the threat of a Thanksgiving winter storm, this time with renewed caution concerning its evolution. Please read here for the discussion on how this potential came about in the first place.

We'll begin with the GFS model forecast for November 27th. The top left panel shows 500mb vorticity values in shaded colors, as well as 500mb height contours superimposed. The top-left image depicts 1000-500mb thickness values (people north of the red dashed '540' line generally can expect snow) as well as sea level pressure. The bottom-left panel gives an indication of low-level relative humidity, basically a measure of moisture in the air, while forecasted precipitation is on the bottom right in conjunction with 850mb temperature values. Glancing over this image, we see a significant storm evolving in the East US, with the combined energy of a trough in the southern jet stream, as well as a potent clipper system dropping down from Canada. These energies combine into one strong trough, traveling northeast and dropping intense snowfall along the East Coast. I'll provide analysis of this scenario after the GEM model, what I'm doing now is explaining what it shows.

The GEM model is different. Using the same definitions for each panel in this graphic as the GFS model, we see the storm is separated into two pieces of energy. We see our primary southern-stream storm in the Southeast US with the expansive precipitation shield, but that clipper is now located to the northwest, not conjoined with the Southeast-US storm. As a result, we see accumulating snow strike the Midwest and Plains, putting down precipitation amounts as follows.

This image shows snowfall accumulation in its liquid-equivalent from right now to 10 days from today, but the clipper's snowfall is shown by the swath of greens and light yellows stretching from the Dakota into Indiana and Ohio. Doing a quick conversion tells us 10 millimeters is equal to 0.39 inches, 15mm is 0.59 inches, and so forth. At face value, the GEM would be kicking out a good 4-9" snowstorm from this clipper, with the highest amounts in those yellow shadings. With higher ratios taken into account, we would probably be facing a 5-10"+ snowstorm. Since the GEM is notorious for exaggerating snowfall amounts, this isn't something to hang your hat on. However, it does give breadth to the idea of two solutions to this storm.

For our model analysis, I want to focus on the problems with the GFS model. I've mentioned a handful of times that the GFS model (and most models in general) are prone to a progressive bias. This means that forecasts will move storms along quicker than they should, which might be leading to that merging of the clipper and southern-stream storm on the GFS. Dropping/correcting this bias might lead to a forecast not unlike the GEM model.

For now, purely due to uncertainty, I'm not willing to side with one model over the other. However, keeping in mind this bias which does appear to be showing itself in the GFS model (for now), a solution similar to the GEM forecast might be expected, something that could easily change in coming updates.



Anonymous said...

Hi Andrew... coming to you from the Northwest suburb of Barrington in the Chicago area.. It has seemed to me that especially during the months of November, December, and January the ECMWF has generally had the best handle on clipper type systems coming out of the Northwest.. Is there any truth to that and if so is the European falling along lines closer to that of the GEM or GFS or possibly a compromise of both. Just curious because I'm one of those guys who loves a good snow on a Thanksgiving in my neck of the woods .. Thanks for all you do for us novices who love weather because you have taught us all so much.. Vince

Tristan NACHBAUR said...

Hi Andrew,

I live in Northern France (latitude of Winnipeg) but I follow up very carefully what happens trough the other side, I continue to think that most of your weather events in winter impact ours. Let's say we share the same climate of Belgium.
So my question is the following : Are you fully confident in the predictions of ukmo, GEM or CFSv2 who seem each of them to forecast a much milder winter in Western Europe and over a large part of US? Some people doubt of it but their models are robustly computed and accurate. However, I hear more and more that our winter would be actually splitted in two parts in Europe. Mild and wet till mid-december then colder thanks to indexes like strong negative values in the OPI index in october (and so a negative NAO that is a good new for a colder winter :)) or an earlier SSW than expected, disturbing the polar vortex. So what are your predictions for Europe and eastern USA when you compare those data with the past? My question may have been already sent to you. If any, all my apologizes.
Thanks a lot in advance for your response.
Best Regards,

Anonymous said...

4-9" not quite buddy. We're looking at 1-3" here in NWI Wednesday into Thursday. Note that the map you posted is millimeters and not inches. 10 millimeters of snow is less than a half inch.

Andrew said...

Anonymous: Please note on the image that it says 'water equivalent', and also note that this is one model forecast, certainly not the end-all, be-all.

Vince: The ECMWF is generally most accurate among all guidance, but recently with general poor model performance, I've found it best to use a mix of models for now.

Anonymous said...

any ideas when kansas my see its first decent snaowfall