Dear Tom,
Why not use weather averages based on the whole record for Chicago rather than a 30-year period, as is now done?
Allen Benson, Chicago

Dear Allen,
The use of a 30-year data set (currently 1991 through 2020) for computing climatological averages is a compromise between a huge data set, like the period beginning with 1870 (the inception of Chicago’s weather records), and a short one, such as the year 2021. Averages based on huge data sets are little affected by extreme weather events, but “old” data may not be representative of today’s climate. Averages based on a short data set (such as from 2021) will be representative of recent weather, but will be unduly distorted by extreme events. Therefore, the international convention of a 30-year data set seems to be a good compromise.