Y’know, I don’t expect perfection from weather forecasts– either forecasts by actual trained meteorologists or through a weather app. I want a vague idea of what the temperature is going to be like and a radar; I can figure out the rest on my own. I’ve discovered a new request over the last couple of days as the weather’s approached record-shitty levels: I would like the projections being made to make some kind of fucking sense. To not be, oh, impossible.
To wit:

This particular screenshot is taken from the weather channel’s app; I probably ought to have taken it from Yahoo’s, as Yahoo’s app is terrible, but it’ll do. Take a look at the forecast for today, Monday, and tomorrow.
Unless they are using definitions of what I consider simple words– “Monday,” “Tuesday,” “high,” and “low”– that I do not recognize, there is no goddamn way that these can be right, because temperature does not teleport. If today’s temperature is not going to go above -8, it is impossible for tomorrow’s low to be 3, because at some point the temperature has to be in between those two temperatures.
Note that Wednesday’s low is Tuesday’s high, which would be expected and consistent given a gradual warming trend over those two days. That one makes sense!
It gets worse: when I look at the hourly forecast, I find out that at 7:00 AM tomorrow, they’re expecting the temperature to be -11 and the wind chill to be in the neighborhood of -36. This is from the same weather service that says the low is going to be three degrees. Their hourly forecast doesn’t project a positive temperature until 2 PM tomorrow, and even then expects the wind chill to be -17.
Again: it contradicts the high and low data that their own weather service is providing. You can fit both of these inconsistent forecasts on the same screen.
Ordinarily, I’d blow this off– but it really doesn’t seem like it ought to be a terribly difficult programming challenge to get these numbers to pull from the same database, and there is a very real concern about whether my students are going to be expected to walk to fucking school, or wait for a bus, at 7:00 in the morning tomorrow. There is a big damn difference between three degrees and 36 below fucking zero.
I fully expect school to be cancelled again tomorrow. They haven’t done it yet, but we didn’t get the call until around 1:00 yesterday afternoon, so they’ve got a couple of hours. But they’re going to be getting very different data depending on which forecasts they’re paying attention to (and, again, this isn’t just one app— I’m seeing nonsense like this all over the place about tomorrow’s forecasts) and I would hope they’ve got somebody who understands how goddamn numbers work doing the deciding.
I don’t expect perfect forecasts, and I have some understanding of how these things work. I’m not the type of guy who sees a 20% chance of rain and then gets pissed off when he gets wet. But a little bit of internal fucking consistency would be nice.