#scisky #climate
https://phys.org/news/2025-08-sea-1990s.html
[ So were the late-80s projections about climate change effects on weather I read in the WSJ ]
There was a write-up a few years ago about how various climate models since the first ones have performed. For the earliest ones, we have 50 years of prediction. Generally, they have held up quite well.
The first projections were much earlier, but they were not models, they just said how much warming to expect from a given amount of CO2. That has held up too, since the 1930s I think.
In general, the early models were overly optimistic and predicted less warming than we have seen, not because the models were off, but because they built in the assumption that, since the science was ironclad, policymakers and public would respond by cutting emissions. It's hard to remember, but prior to 1990 there was no partisan divide on the science of climate change in the US and relatively minor differences on what the proper response should be.
In other words, they got the physics right but the psychology/sociology wrong.
The other interesting thing is that as models have improved and gotten dramatically more details and precise, they haven't gotten more accurate - the old simple models and the new complex models turn out very similar conclusions.
Finally... it turns out one of the hardest things to model from a physics perspective is clouds. High wispy cirrus clouds turn out to have a significant warming effect while big puffy cumulus clouds have a cooling effect. So figuring out how a given amount of warming will affect the total water vapor in the atmosphere isn't *that* hard, but figuring out how much will form cumulus clouds and how much will form cirrus clouds is maddeningly hard.