We looked into whether this Maine lobster can predict the weather

This story first appeared in the Midcoast Update, a newsletter published every Tuesday morning. Sign up to receive here to receive stories about the midcoast delivered to your inbox each week, along with our other newsletters.

The days are getting shorter and cooler in Maine. But it’s never clear when exactly the shift from summer to fall to winter will truly start.

One midcoast community has devised its own quirky system for answering this. Just as one Pennsylvania community has long used a groundhog seeing its shadow as a predictor of winter’s remaining length, the city of Belfast now has its own form of Punxsutawney Phil.

This coming Labor Day will mark a decade since the Belfast Area Chamber of Commerce first began plucking a lobster out of the Passagassawakeag River for its annual forecasts. Known as “Passy Pete,” the crustacean is caught by a local fisherman in advance, housed in a trap and brought forward.

It’s then presented with two scrolls that feature alternating poems by a local writer, one indicating that there will be an early winter, and another forecasting six more weeks of summer, according to the chamber’s executive director, Dorothy Havey.

“Whichever one he seems to hit with his claw first is the one that we read,” she said. “It’s all done in good fun.”

The lobster is then set free to enjoy his life, with a new one receiving the Passy Pete mantle the following year. “Maybe he comes back year after year,” Havey added. “We don’t know.”

But even if it’s done in good fun, all of this raises a question: just how strong are Passy Pete’s meteorological skills? The Bangor Daily News compared nine years of predictions with the available temperature records that the National Weather Service collects in the Belfast area.

It wasn’t a simple task. In a way, the scrolls presented to Passy Pete on Labor Day really measure two different things: whether summer will last for six more weeks or whether winter will come early. It’s entirely possible that both of those things could happen in a single year, depending on how brief autumn is.

Here’s how we ran the numbers: for the years in which Pete predicted six more weeks of summer, we found the last day of that year in which the temperature reached at least 75 degrees Fahrenheit, then measured whether it was six weeks after Labor Day.

And for the years Pete predicted an early winter, we identified the first day in fall or winter when the temperature never rose above freezing, and judged whether that came earlier than the actual start of winter on Dec. 21.

The verdict? Passy Pete is not a good weatherman, with a success rate of just 22 percent.

He was better in his predictions of early winter, successfully forecasting them in 2022 and 2023, when fully freezing days first came in October, but he mistakenly made early winter calls in 2018 and 2019.

On the question of longer summers, though, Passy Pete probably can’t be trusted. He’s forecast them five times, and the only year he came close was 2017, when the last day to hit 75 degrees was Oct. 5 — just over four weeks after the prediction.

So Passy Pete will need to work on his forecasting skills if he really wants to upstage that famous groundhog from Pennsylvania. But to be fair, Punxsutawney Phil does not have the most stellar record either: His predictions have come true just 36 percent of the time, according to LiveScience .

Regardless of Passy Pete’s accuracy, people will still be looking out for which scroll he chooses next week.

“We’re excited to have a decade of predictions, and we just look forward to what he’s going to say about this year,” Havey said. “Of course, we hope it’s going to be a late summer.”