Joe’s Weather Blog: Forecasting History (TUE-1/27)
Good morning…crisp and frosty out there today but the great weather continues through tomorrow before we start a gradual chilling process that will eventually lead to some arctic air dumps over the next 10 days or so. Winter isn’t over and there’s an increasing chance of snow developing over the weekend as well.
Today: Sunny and mild with highs again well into the 50s to near 60°
Tonight: Fair skies and not too cold with lows in the 30s
Tomorrow: I’ll be sticking to my guns and forecasting a record high. The record now is
65° set back in 1917 74° set back in 2013. I’m sticking to my previous forecast of 70° with some upside potential to that. Gusty S/SW winds of 30 MPH will settle in to later in the morning and into the early afternoon. A cold front will move through during the evening.
It can be an easy thing forecasting a historical event…and a very difficult thing too. Say for example a record high temperature and for whatever reason there is a low side record high for a particular date and you have 100% confidence that temperatures are going to break that record. You go for it and hopefully it happens. You just forecast history and it played out. You essentially made a forecast for something that hasn’t happened before on a particular day. Perhaps even for a particular month. The all-time JAN record is 75° and while doable I’m NOT forecasting that at this point. Would it surprise me tomorrow…no as I’ve tweeted and blogged about since Saturday.
For that to happen there are a few other parameters that HAVE to happen and I’m not sure they will happen so I’m NOT forecasting that likelihood. There is a difference in forecasting “could” and “will”. I try to base my forecasts on the “will” part and blog and talk about the “could” part here.
(Update at 4PM): So here I’m writing a blog about the NE part of the country and the mistakes made and I make a mistake in KC for weather that has already happened. The irony does not escape me. The record tomorrow is actually 74° set in 2013. Why did I think the record was 65°? I was sourcing my records from the NWS web page. Notice anything about that web page? I sure wasn’t paying attention…it was updated through 2011. It doesn’t have updates apparently beyond that. I didn’t even look at 2012-2014…what were the odds? Well they hit. 2013 is the record and it’s 74° and while still doable as I’ve mentioned…it’s not likely. One thing though is that it’s sure warm now in western KS
The monthly record is still 75° set in 1950…I double-checked.
Look at the 3PM map of weather through the Plains…the temperatures are in RED.
A close up of KS temperatures via the KS Mesonet…show these numbers for 3PM…click on that image to make it larger.
So yes I’ve just face-palmed myself :(
Meanwhile model data today is still suggestive of the potential of an accumulating snow over the weekend. Again not a big storm but something to watch especially with colder weather coming in overnight SAT into Sunday. Should we have snow on the ground MON AM…there is the potential for sub-zero cold in KC. Just a heads up and certainly contingent on snowcover.
So now we have the case of the NE US Blizzard that is still cranking away for parts of Long Island and into the eastern reaches of New England where the “will” part of the storm was always 24″+ of accumulation. This seemed to me to be a very safe forecast even from the past weekend. It’s verifying nicely + the insane winds and everything else that is happening…all accurately forecast and being delivered by the monster Nor’easter.
It’s an impressive storm for sure. Yet as I type this there are some pretty loud grumblings about what happened from essentially the NYC area and southwards down the coast of NJ. It was still a storm…it was still a significant storm for at least the NYC area 4-10″ seems common but there seem to be a lot of lower scale numbers there and not too many of the higher numbers there.
Then once you get west of NYC proper the numbers tank…likewise to the north. Fore a region expecting 24-36″ of historic snow…that is NOT a good forecast.
This morning some meteorologists are apologizing publicly for the failed forecast.
He’s a good follow on twitter and the MIC or meteorologist-in-charge of the NWS in Mt Holly, NJ. In other words he’s in charge of the office there.
The thing is…forecasting history when it comes to snow can be a MUCH more difficult thing. It’s one thing to do it like the NYC office did…it’s another to do it and insert caveats so the folks there understand should the storm NOT pan out as expected. It seemed before and during the event that was the missing item in the discussions that I read. There was soooo much certainty.
Here in KC, as you know, we “mostly” worry about whether a storm will track too far south or perhaps too far north and place us in the wrong spot for the heaviest snow. Then when it comes to nor’easters it mostly whether or not the storm will be too far east or too far southeast for the NYC area to be heavily impacted. IF you watched my forecast on Sunday evening I talked about this during one of my evening weather hits. It took me awhile to find a way to display a product that was giving NYC about 9″ of snow and Boston over 2 feet of snow. I found one and showed it relaying the story above to our viewers. I mentioned that something just didn’t seem right about the NYC area in particular about the storm and the model output. I was very happy with this “armchair” forecast. The funny thing is “armchair” forecasting is a heck of a lot easier than being the person who has to do this for real. In this day and age of twitter and FB and all this data that’s freely available to all who know where to find it there are a lot of “armchair quarterbacks” out there. It’s like being the back-up quarterback. You hear a play called in the headsets…you’re chatting with another teammate and say blank would be the better call…the called play fails miserably and you realize YOUR call would’ve probably been the game winner given the defense that was played.
Would I have done this and been the sole voice of reason while the insanity was running around me IF I was working in NYC? Don’t know. I’d like to say yes to that but by the same token I know what the models were outputting. Colleagues who I respect may have been wanting to go “all in” and they too would’ve added more pressure to what you were thinking about doing/saying/rationalizing/forecasting. A career could’ve been made though by sticking to those guns. People would always remember you because in a sea of crazed snow forecasts you would’ve been pretty darn close. In fairness though I don’t think I would’ve forecasted such a sharp gradient to what happened. At least in the 2 days leading up to the event. Perhaps that afternoon maybe. What actually has me fired up a bit more today is not the criticism some are getting…but its the “stubbornness” that was evident when it seemed, at least to me yesterday, that something was amiss. The short-range model data was portraying far lessor amounts in NYC area…and they weren’t even referenced from what I can tell by looking at the discussions. Perhaps they were tossed aside. Could happen. They aren’t gospel either and when you’re “set in your ways” with a forecast you start eliminating contradictory data. This was VERY early in the afternoon. Granted the horses were out of the barn as I like to say…but still it was a warning shot that something wasn’t “quite right”
I’ve been in the same situation as well as many of my colleagues. We/I call it denial. You stick to your guns and your forecast longer than you should (if they won’t admit it they’re likely lying). You start seeing things on radar that may or may not be there. You start saying more of the “could” happen stuff and less of the “will” happen stuff. Like I said been there…done that…and it stinks when you finally have to give up the ghost. My feeling is that the “give up the ghost part” perhaps could’ve come 12 hours+ earlier. At least the potential of 24-36″ of a forecast part. I also checked in with some of the TV mets there and they too were “mostly” toeing the party line about the storm…although some did cut the totals somewhat. I was interested in the later shows especially when it was VERY apparent that something wasn’t happening right.
It’s far too easy to be a critic from 1000 miles away I realize…but wow IF you’re going to forecast 24-36″ of snow and you’re wrong for millions upon millions of people…that’s not a good thing. The storm is everything and then some from eastern Long Island up through eastern New England…it was always going to be “ground zero” for the blizzard. Again I’ve had my “busts” too…except my busts weren’t forecasting 2-3 FEET of snow and not getting 12″. I did see LaGuardia airport coming in at 11″ I think. You live and you learn. Will those forecasters ever issue a 2-3 FOOT snowstorm forecast again…maybe. Maybe not though. Will they remember this storm for the rest of their lives…probably. The next time a storm is “modeled” to do something similar will they pull the 2-3 FOOT trigger again…or will something in the back of their minds say…you know a 12-24″ forecast with some caveats might be a better play on this.
Of course I write all this with some snow possible here this weekend. It doesn’t look like a big snow but something perhaps sloppy through SAT evening…then something a little more “sticky” later SAT night into Sunday morning. How much? FAR too early to spin out numbers. Like I said it doesn’t look too major though.
Have a great day and let’s see if we can get “history” here in KC tomorrow!