The Wonderful Issues of Climate Change - Post #3 - 'Hottest Ever'

Original Date: January 10, 2024; Updated: January 14, 2024 and February 25, 2024.

Author: Tim Platnich

Following up on Post #2, I want to talk about headlines that make such claims as: 'hottest July ever recorded'.

We must be sensitive to what such a claim might be saying. For example, what does 'ever recorded' mean having regard to Post #2. Often times, 'ever recorded' means since satellite measurements have been made - about 50 years. Or, it could mean since 1850 based on thermostatic recordings. The point is to be mindful of what 'ever' means.

What does 'hottest' mean? Let's leave aside the whole month of July for a moment and focus on the days of July. Each day, at least two temperatures are recorded: the daily high and the daily low. These two temperature may be averaged yielding a mean temperature for the day. If we were dealing with a specific day of July and the claim is that that day, say July 10, was the hottest July 10th ever recorded, we would have to consider whether the claim means July 10th recorded the highest high, or the least low low, or the highest mean temperature. So, if the high on July 10 was 30C and the low was 10C, the mean would be 20C. What if the 30C was the highest recorded high, but the low was lower than 10C and the mean was lower than 20C? What can we take away from that? Does this particular July 10 represent anything relating to global warming especially if the days on either side of July 10 were not record settings?

Considering the whole month of July, the same analysis could be made except that every number for the month would be an average. We would have an average high, an average low and an average mean. What is meant by the hottest July every recorded?

Let's say that on all three bases, July was the hottest recorded since 1850. By how much? A degree, a tenth of a degree or one hundredth of a degree. Magnitude matters. Was this July a one-off? If so, what does it tell us? If all or most or many of the Julys in the last 30 years were hotter than those Julys of the 30 years from 1850 to 1880, we may have something. We may have warming and, to the extent warming is climate change, we may have that as well.

So far in the analysis I have assumed that the July in issue was for a particular place. Headlines tend to refer to specific places. In this case, what do we do if the same place, a month before, experienced the coldest June ever recorded? Or, if 100 km away from the particular place, another place did not experience any such hottest July?

If we are talking about the whole globe, the hottest July would be referencing the average mean temperature for the whole globe.

We could go even further and claim that the year 2023 was the hottest year on record. This claim would have to be based on averages.

One-off days, months and even years, are mere data points. It is averaging over a climate period, like 30 years, that matters. Beware of headlines that tempt you to forget that.

Update: The Weather Network has reported some interesting statistics about Canada's recent (January 2024) coldsnap. Keg River, Alberta, experienced a low of -50.1C, the lowest low recorded in Alberta in 20 years! The Edmonton International Airport reached a low of -45.9C. This is the third lowest temperature of that location since records began in 1960. Further, the high for that day was -34.4C - the lowest maximum high ever recorded at the airport. Does this make it the lowest maximum high 'ever'? In Calgary, on January 13, 2024, the maximum high was -30.4. This was Calgary's coldest day since February 1, 1989 (35 years). The low for that day was -35 which was the coldest low since January 11, 1998 (26 years).

Moving west, on January 11, Vancouver International Airport recorded its coldest January value since 1993 at -12.9C. It recorded the coldest windchill value for the airport since 1968. That day, Victoria recorded a daytime high of -6.6C. This was Victoria's chilliest day since 1972 (more than 50 years).

Evidence of global cooling or just data points? Obviously just data points.