Post by grega94 on Jan 23, 2024 23:40:58 GMT -5
Not sure if this has already been discussed, but it's something that's been on my mind over the years and recently I decided to look at the numbers.
So, for simplicity it seems like the wiki weather boxes use the daily median to calculate the daily mean, but considering the differing latitudes and changing day lengths as the seasons go by, the median can be a degree (F) or so off. Should the wiki boxes then label them more accurately as daily median as that is what they are technically depicting, and in that case wouldn't that throw off the actual climate classification of some borderline places? Or are the differences so negligible that it's not worth bothering with?
For example, here is Seattle.
Month: Mean | Median | Difference
Jan: 42.1 | 42.8 | -0.7
Feb: 43.3 | 44.0 | -0.7
Mar: 46.4 | 47.1 | -0.7
Apr: 50.6 | 51.3 | -0.7
May: 56.8 | 57.5 | -0.7
Jun: 61.4 | 62.0 | -0.6
Jul: 66.5 | 67.1 | -0.6
Aug: 66.7 | 67.4 | -0.7
Sep: 61.9 | 62.6 | -0.7
Oct: 53.0 | 53.8 | -0.8
Nov: 45.8 | 46.5 | -0.7
Dec: 41.4 | 42.0 | -0.6
Year: 53.0 | 53.7 | -0.7
here are some more examples for yearly mean.
City: Mean | Median | Difference
Los Angeles: 66.3 | 65.8 | +0.5
San Francisco: 58.1 | 57.7 | +0.4
Miami: 77.6 | 77.4 | +0.2
St. Louis: 57.6 | 57.4 | +0.2
Anchorage: 37.7 | 37.6 | +0.1
Phoenix: 75.5 | 75.6 | -0.1
Fairbanks: 28.2 | 28.3 | -0.1
New York City: 55.6 | 55.8 | -0.2
Atlanta: 63.3 | 63.6 | -0.3
Portland: 54.8 | 55.1 | -0.3
Chicago: 52.0 | 52.4 | -0.4
Seattle: 53.0 | 53.7 | -0.7
Source: www.weather.gov/wrh/climate
* I used 1991-2020 data and made sure to match the location with the wiki box.
Furthermore, I was looking at my own data that I collected from my backyard thermometer which makes a data point every minute (not sure how frequently NOAA collects theirs?) and the biggest difference I found was for Oct 2023 of -1.6 (51.9-53.5). *Only started collecting Apr 2023.
So, for simplicity it seems like the wiki weather boxes use the daily median to calculate the daily mean, but considering the differing latitudes and changing day lengths as the seasons go by, the median can be a degree (F) or so off. Should the wiki boxes then label them more accurately as daily median as that is what they are technically depicting, and in that case wouldn't that throw off the actual climate classification of some borderline places? Or are the differences so negligible that it's not worth bothering with?
For example, here is Seattle.
Month: Mean | Median | Difference
Jan: 42.1 | 42.8 | -0.7
Feb: 43.3 | 44.0 | -0.7
Mar: 46.4 | 47.1 | -0.7
Apr: 50.6 | 51.3 | -0.7
May: 56.8 | 57.5 | -0.7
Jun: 61.4 | 62.0 | -0.6
Jul: 66.5 | 67.1 | -0.6
Aug: 66.7 | 67.4 | -0.7
Sep: 61.9 | 62.6 | -0.7
Oct: 53.0 | 53.8 | -0.8
Nov: 45.8 | 46.5 | -0.7
Dec: 41.4 | 42.0 | -0.6
Year: 53.0 | 53.7 | -0.7
here are some more examples for yearly mean.
City: Mean | Median | Difference
Los Angeles: 66.3 | 65.8 | +0.5
San Francisco: 58.1 | 57.7 | +0.4
Miami: 77.6 | 77.4 | +0.2
St. Louis: 57.6 | 57.4 | +0.2
Anchorage: 37.7 | 37.6 | +0.1
Phoenix: 75.5 | 75.6 | -0.1
Fairbanks: 28.2 | 28.3 | -0.1
New York City: 55.6 | 55.8 | -0.2
Atlanta: 63.3 | 63.6 | -0.3
Portland: 54.8 | 55.1 | -0.3
Chicago: 52.0 | 52.4 | -0.4
Seattle: 53.0 | 53.7 | -0.7
Source: www.weather.gov/wrh/climate
* I used 1991-2020 data and made sure to match the location with the wiki box.
Furthermore, I was looking at my own data that I collected from my backyard thermometer which makes a data point every minute (not sure how frequently NOAA collects theirs?) and the biggest difference I found was for Oct 2023 of -1.6 (51.9-53.5). *Only started collecting Apr 2023.
