@wxornotBG Facebook
@wxornotBG Twitter
@wxornotBG Instagram
As many of you know, parts of the Northeast, and especially New England, have been impacted by a crippling blizzard. Boston has entered the top 5 snowfalls all time, and areas around Boston have been slammed by this system. However, this system produced what is already being called a botched storm by many in areas further west, such as NYC and Philadelphia.
The NWS in New York and Philadelphia vastly overestimated what this storm would produce. This wasn’t just a regular blown forecast like mine from last Thursday, this can be considered about as significant a miss as we have seen in the weather world in quite sometime. Here is what happened in the few days prior to this:
- Models bomb out a clipper system off the Northeast coast
- We see consistency from all models
- Snowfall totals look absurd
- High winds + lots of moisture + heavy snowfall = mass hysteria
- NWS follows ridiculous model totals to the end, even as models began to diverge more as we got closer
The end result of all that is a forecast bust unlike one we have seen in several years. And the funny thing is, snowfall totals ended up being close to a foot of snow in NYC! Why in the world was this a bust? Well, it is easy to see just a few days in advance.
NWS forecast always being updated. Amazing #blockbuster amounts on the scoreboard. 22-31” from NYC to Boston pic.twitter.com/1qurcelonc
— WeatherBELL (@weatherbell) January 26, 2015
The ECMWF model absolutely destroyed the northeast. Like 1978-on-steroids destroyed. It basically said, “Hey, 60 mph+ gusts and at least 2.5 feet of snow on top of two of the biggest hubs in the US” and the meteorological world decided to systematically lose its collective mind. When the NWS in NYC is issuing things like this for a solid 48 hours in a row, it is hard not to pay attention…
So, what did everyone do? For maybe the first time in world history, the public did what was SMART in light of the forecasts. The Gov. of New York issued a state of emergency. The NYC Mayor issued travel bans, school closing, etc. This system was the storm that could be the end all be all. It would be like The Day After Tomorrow, only in real life. The scary thing is that very reliable models were showing this until the storm itself came to fruition. The NAM and ECMWF models were painting almost unrealistic snowfall totals across the region, and when you have those two showing the system dumping on the northeast, it is hard to ignore it.
The GFS (which nailed this system) was calling for it to go east a bit, and take the snowfall with it. However, it was recently upgraded, and the skill that it has with systems of this magnitude was unknown, as it could have biases that we didn’t know about. Additionally, the ECMWF has outperformed the GFS for quite sometime now, and that is one of the main models that the NWS weighs in their guidance. Because of this, it led to the NWS NYC banking on snowfall totals that would have been historic had they panned out.
But they didn’t. The system that was supposed to shut everything down, did. It closed major airports, canceling thousands of flights and keeping people in New York and from getting there. It closed roads and schools and businesses. But the resulting snowfall didn’t warrant this, and there is now a major backlash against the NWS for this storm system.
However, the NWS in NYC does not deserve this type of backlash. They just simply do not. Why? Because this is meteorology. We are in the art of forecasting the incredibly dynamic processes of the atmosphere. The way precipitation shields develop and intensify within a rapidly deepening surface cyclone is very, very complex, and it just isn’t right to place all of the blame on the NWS for this bust. Imagine seeing the two most reliable models on the market consistently paint historic snowfall totals over the largest city in America? But not only that, it is painting winds gusting upwards of 60 mph, and that it would last for over a day? What else are they supposed to forecast and tell the public?
“A LIFE THREATENING…POTENTIALLY HISTORIC WINTER STORM EXPECTED FROM LATE TODAY INTO TUESDAY…”
Because this was such an intense storm, the risk-reward was much higher, and they were almost forced to give the worst-case-scenario. If the NWS forecasts 10-14″ of snowfall across the NYC area, and they did end up receiving 24″ of snowfall, then the consequences would have been dire and much, much worse than they were when they overestimated forecast totals. Am I saying that the NWS in New York needs to do this all over again next system? Absolutely not. However, there are things they really need to improve upon.
First off, unless there is a true model consensus that snowfall totals will be in the 36″ range, there is no need to forecast totals that high. NYC has had many great snowstorms in the city’s history, and not one has been greater than 27“. Getting 36” of snowfall is incredibly hard, and as often as the models may have shown that during this event, you just can’t forecast that. It causes way too much overreaction.
So, even if you “hugged” ECMWF deterministic run lovingly, 30-inches for NYC was still outlandish. That’s really going for broke.
— Ryan Maue (@RyanMaue) January 27, 2015
The biggest thing they need to improve upon is communicating that their forecast was the worst case scenario. The worst case. There was so much uncertainty leading into this storm, and even as this storm was developing, there was little that actually showed that the precip shield would develop westward enough to dump 3 feet of snow on NYC. The NWS needs to become better at communicating the uncertainty associated with systems like these so that the public can subsequently make decisions that are logical.
Was this system a bust? Yes and no.
The NWS office in New York City has taken a lot of heat for this, but they even owned up to their mistakes in the forecast. This leads me to hold true to the idea that while the NWS office in New York has things to improve upon, they don’t deserve the heat they have taken for this storm the past couple days.