SoylentNews

The evidence is presented in the linked journal papers that discuss each of the three events that were listed. In short, because you don't want to click through and actually look at the papers, you're denying that there's any evidence. It's plainly obvious that you haven't even looked at any of the three papers that are linked on that webpage, and because you couldn't be bothered to click the links, you're pretending there's no evidence provided. That is incredibly lazy, and your reason to claim that no evidence was provided is asinine and disingenuous. The evidence is literally two clicks away from this comment: one to https://www.ametsoc.org/ams/index.cfm/publications/bulletin-of-the-american-meteorological-society-bams/explaining-extreme-events-from-a-climate-perspective/three-extreme-events-that-were-not-possible-in-a-preindustrial-climate/ [ametsoc.org] [ametsoc.org], the second to one of the three journal papers. You do understand how to identify and click links on a webpage, don't you? The linked articles are BAMS articles, which are intended to be more accessible to people with a general science background than, say, an article in Journal of Climate. BAMS is, of course, still a peer-reviewed journal.

I see you present no evidence. If it's easy for me to do, then it probably is easy for you to do as well. As a result, I think there's a slight error in your claim as to who is being dishonest here.

Because they're simulations not actual observations of reality. Reliance on models instead of reality is a common symptom of the problems we have with climatology.

This is vague and unscientific. It is not a useful criticism. The models are verified against past climates, and generally do a good job of simulating the past. In this case, the models are simulating climates of the past and present day, which exactly what they've been verified against. This is not the same as the scenario of simulating future climates, which have more uncertainty because the models are being extrapolated into conditions for which there's currently no data to verify them against. But attribution studies involve simulating past and present climates. Models have been verified in these climatic conditions. They do a good job of replicating them.

No, it's quite scientific. The huge problem being missed here is that it's quite easy to build models that fit against existing data, but extrapolate to anything you want. And this practice of claiming that something is too extreme for non-climate change scenario is worse than the scenario of simulating future climates because there is no way to check the claim.

Now, you could say that there's a problem with a specific model. There are flaws and biases in each model, sure. But you'd also be wrong to involve that argument, because the researchers were using data from CMIP5. It's a collection of data from many models [llnl.gov], so they didn't just rely on one model.

Now, we're making a silly argument that because there's a lot of models, then the models must be right. The obvious rebuttal here is confirmation bias. For example, we have a several decade history [soylentnews.org] of that mass of models understating carbon sinks - models running hot and making predictions that overshoot actual warming but interpreted otherwise once one adjusts the model for actual greenhouse gases in the atmosphere.

Your criticism is unscientific and would be dismissed out of hand if you provided it as a peer reviewer.

Note, that doesn't make it a false claim. Or for that matter, something that would be dismissed out of hand if I were to present it as a peer reviewer. If the journal really would ignore my concerns about paper two, for example, then they would be heavily biased. It's such an elementary concern.

TL;DR: These studies are simulating climates of the recent past. The simulations can be trusted because the models have been verified against those climates. There are many models, so the results aren't because one single model is flawed.

No, they haven't been so verified. You merely claim they have. There's only one genuine test here - testing against unknown, future data rather than fitting to existing data. And as I noted above, they're failing that test collectively.