The Guardian has found a new report that has caused it much excitation, and it is easy to see why:

The paper, published on Friday in the journal Nature Communications Earth & Environment, found that from 2006 to 2020, the climate crisis contributed to about 15,000 deaths from exposure to small particulate matter from wildfires and cost about $160bn. The annual range of deaths was 130 to 5,100, the study showed, with the highest in states such as Oregon and California.

Scary stuff, I’m sure you will agree. But before discussing the report’s findings in any detail, it is perhaps worth reflecting upon the nature of the study undertaken. Whilst you might be forgiven for believing that scientific measurements lay behind the grim proclamation of a climate change-driven increase in pulmonary deaths, the truth is that the paper is based heavily upon mathematical modelling – or more to the point, mathematical modelling layered upon mathematical modelling. In fact, they used a clinical model to predict the health impacts resulting from wildfire smoke dispersion that had itself been modelled. This modelling took as its starting point a posited climate-driven increase in burn area as calculated from ecoregion-level empirical models using fire weather index (FWI) and precipitation from ERA5, whilst using counterfactual analysis based upon a climate model ensemble (CMIP6). With so many models to work with, what could possibly go wrong? Well, plenty from where I am standing.

Before going any further, I should point out that I’ve got nothing against mathematical modelling. I would say that if it is all you have to work with, then work with it. But should you do so, it is very important to keep a clear idea as to how much credence one can place in the modelled results. Even worse, when modelling is layered upon modelling, you end up with an exercise that is likely to be riven with compounding uncertainties. Given that we are dealing here with three layers of modelling, I think it is fair to say that the result could be characterised as modelled speculation.

Predictably, the Guardian article chooses not to highlight the study’s speculative nature, preferring instead to lead with confident statements that emphasise the hazards posed by the “climate crisis”. It isn’t until one reads towards the bottom of the article that the reality starts to find its voice. Specifically, the article reports that Marshall Burke, global environmental policy professor at Stanford University, feels that whilst evidence linking climate change to burned areas was “rock solid”, the modelling layered on top was a bit flaky:

“Linking burned area to smoke is trickier because you never know exactly which way the wind’s going to blow,” he said, and he wondered how the death estimates compared with fatalities tied to general air pollution.

Even then, we are only beginning to unearth the truth with this quote. I can easily see why the smoke dispersal modelling should be highly dubious, but the idea that this contrasts with a “rock solid” foundation of FWI calculations is simply absurd. That may be the hype issued by outfits such as the World Weather Attribution (WWA) centre but the reality could not be further from the truth. In the recent WWA attribution study of the 2025 LA wildfires the uncertainties associated with the FWI calculations were such that they actually rendered the results statistically insignificant. I’ll repeat that: whilst the paper said that the FWI was made 35% more probable under climate change, the results were equally consistent with the idea that climate change played no role whatsoever! So let us be realistic here; there is actually nothing in the Communications Earth & Environment study that can be referred to as rock solid.

So what exactly did the paper say? If you recall, the Guardian spoke of a “climate crisis” that has contributed to about 15,000 deaths between 2006 and 2020. Let’s put aside for one moment the vagueness behind the concept of contributing towards death, and focus instead upon how such a contribution might fit into the scale of things. After all, even in a counterfactual world in which climate change didn’t happen, wildfires would still have blown their smoke over populated areas leading to pulmonary illnesses. So what is the scale of the increase attributed to climate change? The answer to that question can be found in figure 4 of the report: ‘Increasing climate change contributions to wildfire PM2.5 mortality for CONUS between 2006 and 2020’.

I don’t wish to sound complacent here, but apart from a very odd looking glitch in 2020, the climate-driven contribution doesn’t look that spectacular to me. In fact, what this graph does rather well in my opinion, is draw attention to the scope of contributions from non-climatic factors. For example, what would the graph have looked like if the counterfactual analysis addressed instead the impact of trends in arson, or trends in urban sprawl adjacent to fire risk areas? But of course, in asking those questions, one comes up against the revelation of a misplaced focus, for which we have Professor Patrick T. Brown to thank. It is no surprise whatsoever that the study quantifies only the climatic contribution – after all, that’s what all similar studies do. You will never see a figure akin to the above that quantifies the wildfire mortality due to human decision-making, other than, of course, the decision to burn fossil fuels.

To be fair to the Guardian, they do actually quote Professor Brown, albeit deep within the bowels of the article:

Patrick Brown, a Johns Hopkins University lecturer in climate and energy policy, said he had some concerns about the study. One was conceptual. The study acknowledges the power non-climate drivers have on wildfires, but it doesn’t give them proper weight, he said in an email. Brown, who was not involved in the study, worries decision-makers could wrongly conclude that mitigating planet-warming carbon emissions is the only solution. “Yet in many regions, the more immediate life‑saving action may be fuel breaks, prescribed burns, ignition‑source regulation, public health efforts, etc,” he said.

Even so, the delicate Guardian reader is provided with an immediate antidote to deal with Brown’s heretical poison:

But ultimately, the study notes, the problem of deaths from wildfire smoke will only get worse without the reduction of greenhouse gas emissions.

And I suppose, to the Guardian reader at least, that is all that you should be worrying about. Besides which, it’s all science, don’t you know. You may scoff, but you’re probably just an orc like me. We just need to understand that ‘a study has shown’. And that’s an end to it.