Climate change models part 2: Dealing with worst-case scenarios (That could come true!)

  • Published on February 18th, 2020

Today’s denier roundup is going to be a little different, in that it’s not really addressing deniers directly (although they are certainly watching the issue we’re going to talk about closely, and will pounce on the idea that models are unreliable). Today we’ll be addressing an emerging issue in climate science: that some of the models in the newest generation of models are showing that the climate may be even more responsive to carbon dioxide than we thought.

IPCC models for business as usual


By Climate Denier Roundup

As you may have seen in Vice and Bloomberg, the climate modeling community has been steadily releasing the latest batch of model results from the Coupled Model Intercomparison Project, Phase 6 (aka CMIP6) and around one-fifth of the results are telling us that if we double CO2 concentrations in the atmosphere, the climate would warm by something in the ballpark of 5C, as opposed to the current consensus of around 3C.

If these models are correct, then it would suggest we need to do even more to reduce emissions, even faster, if we are to avoid even-more-dangerous climate conditions.

On the flip side, if they’re wrong, then it means some of the models just run too hot, which of course deniers will seize on as evidence that models can’t be trusted. Those who work with climate models will tell you none of them are perfect, but most all of them are useful in helping us understand the many different parts of a very complex climate system.

In this case, the average of all the models’ results may be the most accurate reflection of reality. As noted, only 20 percent or so of the model runs that have been completed so far are significantly hotter, and those particular models also run warm when hindcasting. Hindcasting is one way to check the accuracy of a climate model, by setting the initial conditions at what the world was like in 1900, for instance, and then running the models forward from there and seeing if the model accurately reflects how the climate actually changed over the century. So the fact that the model runs that show we’re in for more warming in the future also say we should’ve seen more warming in the past may suggest something’s not quite right.

Additionally, the reporting and concern so far has focused on the high end of the warmest model runs. But when you look at all the CMIP6 runs together, the bulk of them are much closer to past projections, and the average degree of warming they suggest is still within the range of what other lines of evidence suggest would be accurate. So overall, especially once the runs that take longer to play out get published, the CMIP6 models may not actually differ all that much from the past aside from perhaps pushing the worst-case, high-end possibility a little higher and making the best-case low-end warming scenario less likely

We should all be careful, then, about focusing on these preliminary results. After all, in some ways the amount of warming we’re going to get in the future is irrelevant, when we already know the warming we have is dangerous.

Sea level rise is already swamping coastlines. Wildfires are already burning through forests. Droughts and floods are hitting farmlands with a one-two punch, and heatwaves are killing vulnerable populations.

So there is every reason to reduce emissions as fast as possible, as lives are already on the line. Many more will be in the future, especially if these preliminary results stand up to scrutiny. There’s no reason to focus on what might happen then considering what’s already happening right now.

In either event, every bit of effort is important, every effort saves something valuable.

More on climate change modeling:

RCP 8.5: World’s Stupidest Debate Makes Jump From Denialsphere To Mainstream

RCP 8.5: Business-as-usual or a worst-case scenario?

About the Author