Weather Hype For Next Week’s Phantom Storm

There’s no shortage of weather hype available online. Some of it is funny and some of it manages to peddle a lot of misinformation. The Capital Weather Gang had a great article about bad forecasts from suspect sources and their proliferation on social media a few weeks ago.

Last night’s European model run developed a significant nor’easter and snowstorm for Thursday across a large portion of the northeast. Yesterday, it was targeting the southern Appalachians and even areas near Atlanta as potential recipients of heavy snow and now it’s New York City and parts of New England.

ecmwf_apcp_f162_us

 

While there is a mall chance this will actually occur – most signals are pointing in the other direction. The operational GFS, the Euro’s American counterpart, has virtually no storm at all.

In a day 6 (or 7) forecast it’s generally best to not look at operational/deterministic model forecasts but rather look at the ensembles (read more about ensemble modeling here). The GFS ensembles for next Thursday don’t show much of a storm at all.

f162

 

You can see one solution there with an amped up coastal low – the rest have nothing. For the Euro ensembles (of which there are more than 50 members) only a handful have a coastal storm that impacts us.

So while you may read or see posts about the European’s big snowstorm next week – keep in mind it’s just one possible solution that is quite unlikely. Posting a map of ensembles that show no storm at all isn’t very exciting – and that’s why you don’t see many of them!

I do think that we run the risk for some snow squalls on Tuesday as the Arctic front races through. The phantom storm on Thursday, while not impossible, isn’t something I’m awfully excited about right now.

Advertisements

5 thoughts on “Weather Hype For Next Week’s Phantom Storm

  1. I would love to have you develop these thoughts further. What are ensembles, and how does one decide which members to take seriously, and which to ignore? It seems to me that it was just about a year ago that you signaled, rather tentatively, and self critically, a terrilble storm that might impact the east coast. I am grateful that I took you seriously and alerted friends and family from DC to Boston. You were right, in that case, and you are presumably right this time. But what is the difference in the two situations that makes you so sure. Can you share more of the details of your thinking? Inquiring people want to know. N

    • Check out the link in the article about ensembles… also check out the wikipedia article on ensemble modeling. Basically, since we will never be able to get the initial “snapshot” of the atmosphere 100% the errors in that initialization grow exponentially in time. Ensemble modeling takes the initial snapshot or best guess and tweaks it a bit in a number of ways. That way we can see a range of possible solutions based on the inherent uncertainty of the initial conditions. In this case the Euro had a snowstorm for NYC but it had virtually no ensemble support. That told me the odds of the Euro solution verifying were quite low. Especially when you’re talking about a forecast between days 4 and 7… ensemble modeling is the way to go and how I base most of my forecasts.

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s