Models like the one mentioned (without references) are created by humans and the parameters they use are arbitrary. Scientists make some assumptions about parameters and run them through the system and get some results. They do this many times and vary the parameters to see what happens. It's the change in results vs the change in parameters that is the interesting stuff. Claiming they can predict the future is dishonest.
Why do all references to climate models refer to 'The models or model'? It's like they are supposed to be oracles -- they are NOT. There is not just one or a few. They do not all show the same things.
Models are a useful way to learn, but picking one result, not showing any work or stating assumptions, and then publishing click-bait referencing it is not ethical.
> Climate models project the first 35°C TW occurrences by the mid-21st century.
Which Climate models? Oh sorry it's the ONE so I won't question further. It could never be wrong.
Do other models show anything different? Oh there are no others I see.
What parameters were used? What datasets? There are zero references to any of this in the original paper but that didn't slow down the click-bait climate-fear-porn author who is not a scientist but a corporate communications person.
> Looks like the data points have more to do with how many weather recording stations a region has than anything else.
As others have said the 'survey data' they do reference seems to be mostly about how many stations a region has.
Who funded this?
> The study was supported by NOAA’s Regional Integrated Sciences and Assessments (RISA) Program.
Big government looking to get bigger.
For my part I have many years experience designing and running HPC scientific computing models in related domains.