better to know...?, climate-debate

The Weather Stations We Never Had

by Dr M. Wielicki, Aug 11, 2025 in Clintel

A central pillar of the climate-crisis narrative is simple enough to fit on a bumper sticker… today is the hottest in human history. That line only works if you accept, without question, that we have reliable, global temperature data before satellites. We do not. What we have is a patchwork of land stations concentrated in a few developed regions, a lot of ocean guesses from ship tracks, and then, later, generous statistical infilling.

Everyone agrees the 1930s were brutally hot across the United States… the Dust Bowl was a humanitarian and ecological disaster. Crops failed, soils blew away, and heat waves killed thousands. NOAA’s own retrospectives still call out 1936 as a benchmark summer, and July 1936 remains a singular month in the U.S. record.

https://www.weather.gov/arx/heat_jul36?utm

…

The global map we never measured

Before 1950, most thermometers were in the United States, Europe, and parts of the British Commonwealth. Large parts of Africa, South America, the Arctic, and the Southern Ocean had little to no routine coverage. Even the NOAA-led overview of GHCN-Daily notes how the core database is a collage of many sources with varying periods of record… that is the raw material modern analyses inherit.

Now the uncomfortable part. When there are no thermometers, you either leave grid boxes blank, or you paint numbers in from far away. HadCRUT historically left many boxes blank, explicitly avoiding interpolation, which means the “global” mean depends on where you have observations. NASA’s GISTEMP goes the other direction and spreads anomalies up to twelve hundred kilometers from a station, filling the gaps with 1200 km smoothing. Those are not trivial choices, they are the ballgame.

If you overlay the 1930s anomaly map with the station density maps, you see something obvious… warm where the thermometers were numerous, cool or neutral where coverage was threadbare. A compilation of historical station distribution between 1921 and 1950 makes the same basic point… the network was sparse and badly unbalanced.

…