Climate is complicated, and no one faults climatologists for using simplifying assumptions to make their computer models manageable… within reason. But two scientists recently pointed out that among the worryingly long list of simplifying assumptions, given the precision claimed for the results, an especially curious one is... that the Earth is flat. In the early days computers were so costly that everything had to be drastically oversimplified. But over time, they noted, "while many of the original approximations have since been improved, one—that the Earth’s surface and atmosphere are locally flat—remains in current models." Seems alarmists were too busy castigating skeptics as "flat Earthers" to get around to making their own Earth models spherical.
Simplification in the models is acceptable within limits, provided it is accompanied by humility about the exactness of the outputs. But it’s a problem when they slash through all sorts of complexities in ways that matter and then claim infallible precision for the results. In fact the models dramatically oversimplify all kinds of things including the physics of clouds (in part by operating on a macro scale whereas cloud formation happens microscopically). And you can’t have more precision in the result than you had in the inputs, especially where complex feedback mechanisms are involved.
As it happens, the flat Earth assumption is pretty serious and, you won’t be astounded to hear, overstates the human contribution to warming. The reason is that one consequence of assuming a flat Earth is that the path of incoming solar rays in the model is biased and the model underestimates how much the sun heats the atmosphere.
The authors of this new study estimated that if climate models got the solar heating correct it would add about the same warming effect as that attributed to human greenhouse gas emissions since pre-industrial times. And whatever is attributed to natural causes isn’t left over to attribute to human causes, which requires us to revise our estimate of the warming impact of CO2 downwards which affects the models in important ways.
This complaint about the real Earth being round unlike the modellers’ Earth isn’t splitting hairs. It's a big enough deal to matter. And while they don't do the calculations themselves, the authors do set out a procedure that climate modelers could implement if they wanted to get closer to reality. Until then, we know who the flat Earthers really are.
"And you can’t have more precision in the result than you had in the inputs..."
But wait! I've seen this with my own eyes, many times! Lots of crime-drama detectives start with grainy old photographs, and through computer enhancement scientists are able to sharpen the image enough to see the tear-drop tattoo under the real killer's left eye... Climate science is like that.
The IPCC itself describes climate change science as "non-linear, coupled, and chaotic". Chaotic systems by their nature must be examined holistically, not incrementally. Consequentially, models that only look at parts of the issue simply do not work. For example, Andrew Weaver's model by his own admission only looks at possible human effects on climate change. In a chaotic system, this approach is doomed to failure.
If climate change is chaotic (and most people seem to agree that it is), then in reality we know very little about what drives it and we really don't know where it is going, except that chaotic systems tend not to have tipping points that lead to disaster.