Sunday, February 9, 2014

February 12-14 Potentially Significant Snowstorm

Model guidance is beginning to jump onboard the idea of a potentially significant snowstorm in the February 12-14 period.

We begin with the snowfall forecast from the CMC (Canadian) model. The CMC model goes all-out on this storm potential and throws down amounts over 12" in northern Georgia and a wide swath of North Carolina into Virginia, before laying down amounts near 24" in the Mid-Atlantic and New England regions as the storm bombs out to a minimum sea level pressure value of 961 millibars.

This is the CMC model forecast of 6 hour precipitation values and sea level pressure contours. We can clearly see the storm in all its might, dropping intense snows of nearly 4" per hour on coastal areas affected by the wintry side of this storm. The storm then deepens to 961 millibars 6 hours later as it pulls away from the East Coast. Taking the paralyzing snowstorm aspect out of this storm for a moment, the CMC would be printing out a potentially life-threatening event, due to the strength of this storm. We would be talking about severe wind damage, possibly catastrophic if this storm actually verifies. Luckily, I highly doubt we see a solution just like what the CMC here says. However, I do think we have to watch out for a storm in this time period.

This image above shows 500mb observed height anomalies on the left, and sea level pressure/high pressure-low pressure denotations/cloud cover denotations on the right. Both panels were valid on February 4th. If you've been with The Weather Centre for a while, you know how we use a rule created by Joe Renken that states a storm system in East Asia then results in a storm in the United States 6-10 days later. Looking at the panel above, we see that a strong storm system was observed cutting north across Japan on February 4th, as the deep blues in the left panel over Japan tell us. If we move ahead 6-10 days from that February 4th timeframe, we find ourselves with a potentially significant winter storm in the US in a time period of February 10-14. I have a good feeling this CMC model projection is indicating that this is the correlating storm. Since the Japan storm was strong, I have a feeling this storm in the US will be strong as well.

The CMC isn't the only model showing a big winter storm in this timeframe...

The ECMWF model also puts down significant snowfall from the Mid-Atlantic to the New England regions, though we see snowfall amounts reduced and moved a bit east compared to the CMC. The CMC model has always retained a bias of being too strong with storm systems, and it's more than likely that this bias is affecting its forecast at the top of this post as well. However, we still can take into consideration the idea that the CMC is printing out a general storm hitting the East Coast, and that idea is something the ECMWF agrees on. We see snowfall amounts of over 12" slamming northern Georgia and into the Carolinas, before a strip of 10-15" snow amounts strike coastal areas in the Northeast.

The ECMWF precipitation forecast at the storm's maximum impact shows the heaviest precipitation on the eastern flank of the storm system, with the system overall weaker than the CMC (which can be explained by the CMC's aforementioned bias), and cold sector precipitation (snow) much more scant (once again attributed to the CMC's bias).

The only model guidance system that isn't jumping on the idea of an East Coast snowstorm is the GFS.

The GFS model actually has the storm begin as two low pressure systems that eventually congeal into one system at the time this forecast graphic is valid (February 13). We can see that the majority of precipitation here is rain, which is located offshore, with snow hitting only coastal areas.

Snowfall amounts from the GFS model aren't even worth mentioning, with 2-4" accumulations only found near the coast.

So why is the GFS so different from the ECMWF?

When comparing normalized 500mb height anomalies from the 12z ECMWF and 12z GFS, both at forecast hour 96, it's not so much that the synoptic pattern is different, but more about how far south the storm pushes before it makes its jump up the coast. The GFS model on the left is noticeably weaker and further south with the storm in terms of these normalized 500mb anomalies when compared to the ECMWF, which has a stronger, slightly further north system. Another issue also appears to be how ridging is aligned downstream of the storm in the Atlantic. We see high pressure a bit suppressed in far eastern Canada, towards Nova Scotia in the GFS on the left, with the ridging more pronounced and stronger in the ECMWF forecast on the right. What's interesting is that the pattern upstream is in near-complete agreement among the models. We see a strong system over Alaska, with strong-yet-suppressed ridging across the Southwest US. Normally, this upstream agreement ought to result in agreement on the evolution of the storm, but in this case, it looks like it will be an issue of just how far south the storm goes. If we count the higher accuracy of the ECMWF and CMC models versus the GFS model as of late, as well as model guidance tendencies for the storm to track too far south and the ECMWF being rather consistent on this idea of an East Coast snowstorm, I think it's best to support the ECMWF/CMC solution in terms of track, NOT amounts. Amounts will be determined further in the future.