Predicting weather: mostly true with a chance of error

Tom Thunstrom of Phillyweather.net tells us what we can – and can’t – predict when it comes to weather. 

No matter how cold the temperatures, the weather is always a hot topic in this region. What’s coming down the pike? Which forecaster got it right, who got it wrong? Do they really know what they are talking about?

Tom Thunstrom is the editor of Phillyweather.net, a social media website and discussion network that discusses weather, and those who talk about weather in Philadelphia. Thunstrom is a banker during the day, but his passion for predicting the weather fills the rest of his time. He sat down to chat with us about what we can – and can’t – predict.

What do people want to talk about on your site?

They have a passion for weather and want to talk about the inside baseball behind weather – the info, the misinformation, we try to raise awareness at the right time. And, mostly snow, that’s the big driver. How much snow is going to fall, and how is it going to impact me – that tends to be the attention getter.

Can we really predict the weather?

To a degree, yes. You can tell with the storm last weekend, 50 miles made a huge difference. When everybody was watching the news on Friday night, we thought it was going to be Philly getting the most snow, but really, it was Atlantic City. Five days out, we knew it was going to snow, but exactly where, and how much, we didn’t know. The science has improved, but small details make a big difference, and the technological tools are not quite there yet, to say, “Hey, this is going to happen at this time and here is exactly what is happening.”

Where do you get your info from?

We monitor a few supercomputers that spit out information, and our job is to try to figure out which model is doing the best job in the current weather pattern. For example, if you are looking at precipitation, trying to predict how much of it will be snow, how much will be rain, what is happening to your west, or to whatever direction weather is coming from. Is the trend going to change?

So, how much is the forecast dependent on the special touch of the forecaster?

That makes the big difference. If you are just looking at the model, you just rip and read what the computer model is showing. We have to understand the computer models’ biases, and understand them. Some are known to predict more precipitation than others, and we also have to know what’s happening downstream. We have to update and change our forecast. Take the Eagles snowfall [last December], that started out at one to three inches, and then the city ended up with eight inches of snow, we had to respond to what was happening to the west and southwest of us, and change our forecast accordingly.

Is this like Wall Street, and knowing which stocks to buy?

A little bit, yes. There is a science to it, and there is speculation. We can get the general idea of a pattern, we can predict a storm approaching in advance, we can look at long-range modeling. But to say it’s going to rain, it’s going to snow, and here is how much – we are not there yet. But we can give people patterns, that part of the science has gotten a lot better. Details are still sketchy.

How did TV forecasters do in Philly this winter?

We grade them and that is a bit subjective, depending on which measure you use, temperature or snowfall. With snowfall, I say snowfall is a wild card, and very difficult to predict in this region. Nobody forecasted 62 inches of snow, nobody came close to that. The closest forecaster said 30, that was John Bolaris who is no longer in TV. The rest said 15-20 inches, and that’s what we typically average in any given year. Snowfall is so variable, it’s the nature of where we live, we have had winters where we had 80 inches of snow! I think it’s what drives this keen interest in weather around here. In Minneapolis, you know you’re going to get snow 20 times a winter, and it will sit there until the Spring!

Why do people LOVE to hate forecasters, it seems to inspire people when they get it wrong.

Watching weather – it’s a passion here. If you watch the newscast every night there is constant weather segments, even on a “normal” day. It brings everybody together. People want the forecast to be right, they expect it to be right, and when it isn’t, and forecasters aren’t humble, there is a big reaction to it. But – you can’t watch the first forecast and take it as gospel – you have to track as it develops.

Are there a lot of rogue weather people out there, on your site, who secretly think they can do it better?

The whole conversation is becoming much more social, now you have private forecasters, Facebook weather pages, a lot of different resources that are trying to forecast the weather. Some of them are really good, outside of the TV realm, which I think is doing a good job on the whole. Some Facebook pages are very good, but sometimes they get themselves in trouble when they get over-excited over a potential storm, and that goes viral. Recently, one person posted a map about a 30-inch snow fall – it went crazy, it was shared 10,000 times, I was getting tons of messages whether I had seen this, and we got three inches of snow that day.

It seems like there is a “I got a guy who gets it right” type thing going on here. 

Yes, my friends talk about it, and they believe they have found a forecaster who gets it right more reliably. But, most of us get it right more often than not. Sometimes we get it right, sometimes the other person. It all averages out, in the end.

So – speaking of getting things right – when is Spring coming?

On or around March 20th, and I’m not just saying that because that’s the official beginning of Spring. Also, it wouldn’t shock me if we hit 90 in April. When we have had this pattern over the winter, it typically allows for a very warm run in April. The good news is that this doesn’t usually mean that we will have a very hot summer, we’re not going to have this 90 degree stretch for three months.