In last week’s #ECS entry the authors came up with a model to estimate the minimum likely ECS. In this week’s, the author comes up with the minimum model needed to estimate the most likely ECS. It might be tempting to keep adding features, and kluges, to a computer simulation partly just because you can. But in pursuit of clarity, following Ockham’s razor, and trying to avoid a Frankenstein’s monster model that lurches about with frightening power doing things you can’t understand, there is something to be said for creating the simplest possible model you can, one that involves making the fewest possible assumptions compatible with plausible simulation. The author, Craig Loehle, published his work in the journal Ecological Modeling, and his unconventional approach was simply to take the surface temperature record, estimate the influences of natural climate oscillations, then simply assume that whatever trend is left over is entirely anthropogenic. And by doing so, and comparing that residue to changes in CO2, his method yielded an ECS estimate of... wait for it... 2 degrees C. That thing again.
Loehle began by looking at the temperature record from 1851 to 1950, noting that this century covers a period when manmade greenhouse gases are assumed not to have had much direct influence on the climate. He estimated a simple model consisting of a trend, a 60-year cycle and a 20-year cycle, and showed that it fit the data quite well. But when he projected it forward to 2013, a gap opened up between the model and the observations, which Loehle chalked up to greenhouse forcing.
He then compared the gap to the increase in CO2 levels that happened over the same period and found it implied a so-called Transient Climate Sensitivity (TCS) estimate of 1.1° C. Then he obtained the ratio between TCS and ECS from the IPCC, and it yielded an ECS estimate of 1.99° C for CO2 doubling, which we’ll round up to 2C. Having computed that estimate, Loehle noted that there seem to be rather a lot of authors getting an ECS estimate of 2C, and if they are right, it means there will be a lot less warming over the coming century than models have been projecting.
Which will come as no surprise to CDN readers, or anyone else who pays more attention to real world data than IPCC climate models.