A little over three years ago, I was attending the new student orientation at the College of Engineering. And, if reality hadn’t struck when the time came to sign up for classes — I just couldn’t handle the prospect of taking multiple calculus courses — I might still be pursuing a degree in atmospheric sciences. Nevertheless, even after my decision to study public policy and political science instead, I still consider myself a weather nerd.

Two weeks ago, my meteorological geekiness was revived when a friend shared an article titled “‘A 30 percent Chance of Rain Tomorrow’: How Does the Public Understand Probabilistic Weather Forecasts?” The answer? It doesn’t understand them at all.

The article follows researchers who decided to survey random pedestrians in five cities: Amsterdam, Athens, Berlin, Milan and New York. The researchers asked the pedestrians to interpret what a 30-percent chance of rain means. In all cities except New York, most pedestrians provided an incorrect definition.

So, do you know what it means when there’s a 30-percent chance of rain tomorrow?

The correct interpretation is that, out of ten days like tomorrow, three of them will include measurable rainfall. Yet most respondents in the study interpreted “a 30-percent chance of rain” as meaning that it will rain 30 percent of the time or in 30 percent of the region for which the forecast applies. Oops.

The fact that weather forecasts are frequently misinterpreted raises an important question: Does the public lack an understanding of basic probabilities or does the fault lie with meteorologists’ lackluster abilities to clearly communicate their forecasts?

To some extent, both camps are at fault. Interpreting a 30-percent chance of rain as rain falling 30 percent of the time does misinterpret a basic probability: if it rains 30-percent of the time, it means there is a 100-percent chance of rain falling. The other misinterpretation, namely that it will rain in 30 percent of the region, seems to me a less serious offense. But I argue that meteorologists aren’t doing themselves any favors when they issue public forecasts.

Consider the following forecast provided by the National Weather Service (NWS) on Dec. 5, 2007: “Tonight…mostly cloudy. Snow likely this evening…total accumulation around an inch. Chance of snow 70 percent.”

Joe Bastardi, a senior forecaster at AccuWeather, responded to this forecast with a trenchant critique on the AccuWeather professional site (of which I am a loyal member): “Does anyone in the (NWS) understand they put out forecasts that make no sense? (T)he darn forecast says they will get an inch…but then has SNOW LIKELY THIS EVENING. How the heck can it only be likely? It has to snow to accumulate an inch, doesn’t it? How can it accumulate an inch, if there is a chance it doesn’t fall (30%)?”

Bastardi definitely has a point: how is the public, who already has difficulty understanding basic probabilities, supposed to correctly interpret forecasts that are ambiguous, contradictory and confusing?

After years of reading Bastardi’s blog, I know that Bastardi’s own forecasting methodology leaves much to be desired. Perhaps in an effort not to make the same mistake as the NWS, Bastardi seldom refers to probabilities in his forecasts. If he believes a hurricane will strike Miami, Florida, he’ll make that call as if it’s a certainty and stubbornly stick to his guns through thick and thin — at least until the storm strikes Charleston, South Carolina instead.

This problem is, frankly, a lose-lose situation. The public is left to choose between confusing forecasts that it often misinterprets or forecasts à la Bastardi that include no margin of error and are often wrong. On the other hand, meteorologists are faced with a public that doubts their predicting abilities and often mocks their profession. I can’t recall how many times I’ve heard the joke that meteorologists are the only people who can be wrong half of the time and still keep their jobs.

This problem might be dismissed as a minor inconvenience if weather forecasting weren’t such a vital component of our daily lives. In the end, meteorologists have an obligation to improve the clarity of their forecasts. As for us, perhaps we should make a little extra effort to interpret a forecast correctly before we crack another joke about it being wrong.

Tommaso Pavone can be reached at tpavone@umich.edu.