For non-specialists, I think the most important view on papers is to not view them as nuggets of truth, but communications of a group of people who are trying to establish truth. No single paper is definitive!

Peer review is an important part of scientific publication, but it's also important for the general public to not view peer review as a full vetting. Peer reviewers look for things like reproducibility of the analysis, suitability of the conclusions given the methods, discussions of the limitations of the data and methods, appropriate statistical tests, correct approval from IRBs if there are humans or animals involved, and things like that. For many journals, the editors are also asking if the results are interesting and significant enough to meet the prestige of the journal.

Peer review misses things like intentional fraud, mistakes in computations, and of course any blind spots that the field has not yet acknowledged (for example, nearly every scientific specialty had to rediscover the important of splitting training and testing datasets for machine learning methods somewhat on their own, as new practitioners adopted new methods quickly and then some papers would slip through at the beginning when reviewers were not yet aware of the necessity of this split...)

Any single paper is not revealed truth, it's a step towards establishing truth, maybe. Science is supposed to be self-correcting, which also necessitates the mistakes that need correction. Climate science is one of the fields that gets the most attention and scrutiny, so a series of papers in that field goes a long ways towards establishing truth, much more so than, say, new MRI technology in psychology.

I'd say that for a non-scientist, you should treat it as a non-event -- a paper that hasn't happened yet.

The climate is not something for which you need daily, weekly, or even monthly updates. Rather, this paper is just one more on top of a gigantic pile of evidence that that climate change is serious, something that we can and should do something about.

If the paper passes muster, you'll hear about it then, though all it'll do is very slightly increase your confidence in something that is already very well confirmed. Or, the paper may not pass review, in which case it doesn't mean anything at all, and you fall back on the existing mountain of evidence.

If the paper had reached the opposite conclusion, that might merit more investigation by you now, since that would potentially be a significant update to your beliefs. And more importantly, it would certainly be presented as if it were a fait accompli, even before peer review.

Instead, you can simply say, "I don't know what this paper means, but I already have a very well-founded understanding of climate change and its significance."

Peer review is still very relevant in climate science. But given it is from well-respected authors, I am more inclined to trust the results at this stage.

It is already published at Geophysical Research Letters, a highly (if not the most) reputable source in the area. But that journal is behind a paywall: https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/202...

For one thing, some of the places which would publish this kind of thing will authorize authors to provide anybody and everybody pre-prints but not the final copy they published.

In principle you could go (pay to†) read the actual final published copy, maybe it's different, but almost always it's basically the same, the text is enough to qualify.

If you go to https://eel.is/c++draft/ you'll find the "Draft" C++ standard, and it has this text:

Note: this is an early draft. It's known to be incomplet and incorrekt, and it has lots of bad formatting.

Nevertheless, the people who wrote your C++ compiler used that "draft" document, because it isn't reasonable to wait a few years for ISO to publish the "real" document which is identical other than lacking that scary text and having a bunch of verbiage about how ISO owns this document and it mustn't be republished.

And you might be thinking "OK, I'm sure those GNU hippies don't pay for a real published copy, but surely the Microsoft Corporation buys their engineers a real one". Nope. Waste of money.

† If you have a relationship with a research institution it might have this or be willing to help you order it from somewhere else at no personal cost.

Pre-prints exists because it can take up to 18 months to get a paper published in a journal or reputable conference. Since lots of people can publish pre-prints[1] what you should think depends on the authors. If they have a record of publishing good research you should think highly of the paper.

[1] - Actually, there are hoops on pre-print repositories, such as arXiv, so not everyone can post there. I guesstimate that 99% of the public has no means of posting on arXiv.

Are you? How many preprints are posted here every day?