This implies that at 11:59 PM tonight, the temp will be 10F. It will then jump to 11 at exactly midnight and climb to 14 tomorrow. But then it has to miraculously jump by at least 8 degrees (from 14 to 22) early Wednesday, since Wednesday’s low is 22. Then something similar happens on Thursday as well.
Only if you assume the low temperature occurs right at midnight…why would it need to be 10 degrees at 11:59? Can’t the lowest temperature of the day occur at any point throughout the day?
Yes, but my point is that the temperature ranges of adjacent days have to intersect.
For example, look at tomorrow and Wednesday. Among those 2 days, the temperature never hits 18. How can it move from 14 tomorrow to 22 on Wednesday without crossing 18 at some point?
I see…I was just looking at today and tomorrow, and focusing too much on the “11:59” part of your post. I see what you mean, and it appears ‘high’ and ‘low’ either don’t mean to meteorologists what they mean to the rest of us, or that forecast is a load of b.s.
My guess from reading the chapter on weather forecasting in Nate Silver’s book.
Those aren’t actual temperatures. They are the most likely lows and highs in a probability distribution of lows and highs from the various computer forecasts. With a possible bias toward lower temps to make people feel warmer. (There is a bias for predicting rain - he didn’t mention one for lows, but it seems reasonable.)
Therefore, you can’t expect the intersection point to be reasonable.
If these were real temps, then the explanation would be that they forgot to chip the ice off the sensor.
Obligatory California obnoxiousness: our high was in the mid-60s yesterday.
True, but why not make part of the prediction come from using the prediction from the previous day?
BTW, it can get a lot worse. I know a lot of them will have the highs and lows be completely different from the hourly predictions. They’ll have a temperature at midnight that is higher than the predicted or lower than the predicted low. And this can be a significant difference. I remember a low of 30-something when midnight was at 18.
Maybe I am reading the question wrong, but I was under the impression that when they give predictions to the teeming millions, they don’t use the precise definition of a “day.”
So when Joe the weather man gets on TV tonight and says “the low for tonight will be 15 degrees…” he is not just saying what the low will be before midnight, he is counting all of the period of darkness as “tonight” weather or not that technically includes tomorrow.
I could be incorrect, but if I see this:
Wednesday (high/low): 85/46
I assume the high will be experienced at about 2:00PM on Wednesday and the low of 46 to actually occur at about 530 AM Thursday (because people think of that as the low temperature of “Wednesday night”).
I think this is the normal scheme for predicted temperatures.
The temps listed in the OP can be interpreted as:
Today’s high temperature will be 49; this will be followed by a low of 10 tonight.
Tomorrow’s high temperature will be 14; this will be followed by a low of 11 tomorrow night.
…
The Australian Bureau of Meteorology reports the daily maximum as the highest temperature reached between 6am and 9pm, and the daily minimum as the lowest temperature reached between 6pm and 9am. So “Thursday’s” minimum temperature could occur at 7pm on Wednesday (but it’s unlikely).
Yep. They use the same definition of “day” as TV Guide used to, wherein the next day doesn’t start until 6. Except in cases in which the temperature steadily drops from 12:00:01 AM.
You are entirely correct. As Chicago weather guru Tom Skilling never tires of explaining, forecast highs and lows are always given in chronological order.
If it is between the afternoon high and the morning low, the forecast will take the form:
Tonight’s low / tomorrow’s high
Tomorrow night’s low / day after tomorrow high . . . and so on.
If it is between the morning low and the afternoon high, the forecast will take the form:
Today’s high / tonight’s low
Tomorrow’s high / tomorrow night’s low . . . and so on.
In rare cases, temperatures will warm through the night or decline during the day. In those cases the forecast will say something like:
