Hold the Hype, Please

For the past few days, the models have been very inconsistent with one another, and from run-to-run (meaning the morning version of a model disagrees with the same model's afternoon version which disagrees with the night version and so on). The problem storm is a nor'easter that is expected to develop over the Gulf of Mexico before storming northward on Monday. Some models had been pushing the storm off the east coast, while others bring it further inland. As is always the case with these types of storms, the track is everything.

The evening models last night (Thursday) finally showed some agreement with a more inland track. This had some major implications on the meteorological community: the hype machine was put into full throttle. To understand this, and the forecasts that are gushing out, the public needs to glimpse into the mind of a meteorologist for a moment.

Most of us love storms...it is just part of our nature. Active weather is exciting. Unseasonable weather gets our adrenaline pumping. Average temperatures and sunny weather are, dare I say, boring. So when there is a big storm on the horizon, there is always a temptation to sensationalize it and force our predictions into more of what we want to happen and less what the models actually say will happen. This is called 'wishcasting'.

It is difficult not to wishcast, and it is something I struggled with a lot during the early days of Grotonweather. I easily got caught up in hype and would force situations to what I wanted to happen. I would latch onto little parts of the model that supported my desire, and ignore larger parts that would indicate I was about to make a bad forecast.

So when the models came into some semblance of agreement last night, the floodgates were opened. Many now felt justified in making big predictions-- after all, the models are all agreeing now!!

Hold the hype, please.

This is a nor'easter we are talking about, a proverbial thorn in the side of meteorologists for as long as they have been predicting weather in the Northeast. The models three days out for a nor'easter often do not capture the details that make or break the storm. Sometimes even the models a day out are no good. Consistency is the key. Just because the models yesterday evening agreed does not mean they will not disagree again this morning, afternoon or evening. It would be wise to see a few more runs of consistency before calling for a monster storm.

But now, to the wishcasting.

As I said above, meteorologists love active, out-of-season weather. What is more out of season than a heavy snowstorm in late April? And that is exactly what some meteorologists are calling for over parts of the region. However, given the current models, I feel such a forecast is uncalled for. Could it happen? Sure- history shows it can. I just do not think anyone is justified in making that call yet.

The first problem is the "540 thickness line". Meteorological jargon for sure, but let me explain. One of the (many) equations we have to learn in school related the average temperature of the atmosphere above a given point to the pressure at certain levels in the atmosphere. A warmer atmosphere will cause higher pressures, and thus a "thicker" atmosphere.

There are certain rules of thumb for how "thick" an air column above a point can be before the air is too warm for snow. One of those rules is the 540 line. To keep things simple, it suffices to know that values below 540 are favorable for snow and values above 540 are going to be rain. This rule has a catch though: it works best in the dead of winter. In late fall and early spring, values at or just below 540 probably are going to be rain, or perhaps a rain/snow mix at best. I like to see values closer to 534 before I get excited about snow.

Let us now look at Monday morning at 8am, the time that many meteorologists are posting images of the models for. Surface temperatures will be coldest at this time, with daytime heating kicking in for the afternoon Monday.

Major snow storm? The models don't think so.

As you can see, these three major models all indicate little, if any, snow confined to the far western fringes of the storm. Additionally, temperatures have been unseasonably warm all winter and spring. The ground, therefore, is also unseasonably warm. Late April storms of the past (2007 and 1993 for example) have come following very snowy years where the ground was likely cooler than normal. I have a hard time believing snow would accumulate very easily except perhaps over the highest elevations unless it got very cold (surface temperatures in the 20s).

Models try to directly guess the precipitation type and snowfall amounts, and some will undoubtedly point to those model outputs as their justification. (The NAM model, for example, has 6" of snow in Ithaca on Monday). Those model predictions are basically made by taking the model predictions, putting them through more equations and conditions, and then trying to make a new prediction. They can be useful, but the more processed the model data is, the greater the opportunity for error. The precipitation type and snowfall amounts, especially from the NAM model, I have found to be rather unreliable this spring. A few weeks ago, this same model insisted on bringing Ithaca another storm of 6-12". It rained.

In conclusion, this is what I want the public to take away from this. We don't know what will happen yet. Confidence is low and this system needs to be watched. While I enjoy the loyalty my patrons have to Grotonweather, I encourage them to look at other forecasts, too. I have been wrong before and will be again in the future.

What I want other meteorologists to take from this is as follows: we, as a meteorological community, need to focus more on uncertainty and communicating that in our forecasts. We need to cut the hype and sensationalizing every large storm. There are times when we NEED to really play up a storm- like last weekend's tornado outbreak. There was a lot of justified hype last weekend and it saved lives. This weekend's hype is not justified. Where do we draw that line? It is impossible to say for sure. But I can say it needs to be drawn better than it currently is, or when it does matter, people are not going to listen.