I often ask: ”Do weather forecasters feel guilty accepting their pay?”
Most people would say: “They certainly should”.
After clearing the streets of NYC for “an unprecedented blizzard of epic proportions”, weather forecasters had to eat crow when the snow-that-would-end-the world turned out to be, well, a garden-variety winter snow storm.
At least one weather-dude had the decency to apologize.
According to CNBC:
“Gary Szatkowski, the meteorologist in charge of the National Weather Service’s office in New Jersey, stunned people in the wee hours Tuesday with a heartfelt apology for the blown forecast.”
How did the forecasters get things so wrong?
Paraphrasing Nate Silver from his book The Signal & the Noise:
Weather forecasters have a pretty good (and improving) track record given that weather forecasting is tough to do.
Why so tough?
Reason 1: think, chaos theory.
Uncertain starting points and dynamic interactions that cause relative small changes in the “system” to have big, sudden impacts on outcomes.
That makes weather – like snow storms – hard to call.
Silver praises weather guys for being able to forecast where hurricanes hit land within 100 miles … down from plus or minus 350 miles a couple of years ago.
This week’s storm veered 50 miles east of NYC where it was expected.
That’s within Silver’s margin of error.
Couple that with something known as “wet bias”: the tendency of forecasters to overforecast the effect of storms … figuring people will be happy and forget if a storm underperforms … but hold a grudge if a storm is worse than predicted.
Chaos Theory + Wet Bias = Blown Forecast
My favorite excuses so far have been:
1. Based forecast on a European (French) computer model rather than a newly developed U.S. model.
2. Climate change causing havoc on all the computer models … all historical relationships are being disrupted.
Nice try guys
The greatest damage from the storm appears to be on weather forecasters already marginal credibility.
According to a CNBC poll, the busted blizzard has caused 2 out of 3 East-coasters to lose confidence in weather forecasters.
Wonder how many had confidence before the storm missed?