Tag Archives: Weather Forecasting

What happened to the Blizzard of 2015?

What happened to the Blizzard of 2015? Well, it happened. Despite breathless complaining about how the forecasters got it all wrong, they didn’t. As the storm was predicted, there should have been close to about two feet of snow in the New York City metropolitan area, but as it turns out, there was between 8 and 12 inches. That means that New York City experienced a typical winter month’s worth of snow in one day. Also, most snow that falls on The City falls a few inches at a time and melts more or less instantly, as few cities can match New York in its heat island effect. So, 8-12 inches of snow all at once is a meaningful, crippling snow storm. Two feet would have been much worse, but it is not like The City did not experience a memorable weather event.

More importantly, the forecast was for a huge blizzard with up to three feet of snow across a blob shaped region of the Northeast approximately 475 miles along its longest dimension (see graphic above). The blob ended up being off, on the southwest end, by about 40 or 50 miles. So the spatial extent of the storm was misestimated, days in advance, by about 10%. An object the size of a country was off by the distance a healthy adult can walk in a long day. That was, ladies and gentleman, an excellent, accurate prediction.

nyt-march-29-1976But, since the storm’s outcome was different than predicted in the world’s most inward looking city (you’ve seen the self-effacing maps produced now and then by the New Yorker magazine), it is assumed by many that the forecast was bad, that forecasting was bad, that weather models are bad, and so on.

As meteorologist Paul Douglass told me yesterday when I asked him if he was going to be kneeling on any carpets today over the difference between prediction and reality, “No kneeling, Greg. Just because we tap supercomputers and Doppler radar doesn’t mean we can predict snowfall down to the inch. Models are good and getting better, but they’re not perfect and never will be. People expect perfection in an imperfect world. Boston picked up 20-30” snow, Long Island saw 15-23”, so did much of Connecticut. There was an 8 foot storm surge on Cape Cod where winds gusted to 78 mph.”

Paul also told me something he shared later that day on the Ed Show. “Over 30 years I’ve worked with a series of anchormen in the Twin Cities and Chicago. When they invariably gave me a hard time for busting a forecast I reminded them that a monkey in a sport coat could report on what happened yesterday. Look at the trends and predict tomorrow’s news headlines!” He indicated that when sportscasters started to routinely predict tomorrow’s scores rather than report today’s scores, they would be on a level playing field with the meteorologists.

Here is that Ed Show piece:

The Blizzard of 2015 was in some ways comparable to the Blizzard of 1978, which was one of the first storms of the modern era of increased storminess. The snowfall totals may have been greater for 2015, but coastal winds were greater for 1978. But, in 1978 over 100 people died, and most of them died of exposure because they were caught in the snow. So, in terms of cost of human lives, the two storms are very comparable despite the differences in winds.


Why did over 100 people die in New England’s 1978 storm, but either zero or one person died (depending on attribution of a single sledding accident related death to the storm) in 2015?

Weather forecasting. It got better because the science and technology behind it got better. And, frankly, that is partly a result of storms like the ’78 storm and various hurricanes, which prompted an interest in advancing this technology, which includes on one hand satellites producing piles of data and on the other hand advanced computer and software producing powerful models.

You should buy your local meteorologist a beer.

The image comparing 1978 and 2015 is a chimera of images that come from NOAA and the Boston Globe.