See Comments down arrow

#ECS in the real world: Spencer and Christy 2023

14 Feb 2024 | Science Notes

We conclude our series on #ECS estimates derived by observing the real climate rather than running models with a September 2023 paper by University of Alabama climatologists Roy Spencer and John Christy, the same experts who invented the modern method of using satellites to monitor atmospheric temperatures. Here they use surface thermometer data, despite the known problems with those data sets, and a new aspect of their analysis is they look at how surface heat diffuses down into the ground. Climate models take into account the way heat diffuses deep into the ocean, but they only include land heat storage down to 10m. Yet studies from boreholes show heat diffuses hundreds of meters down, so Spencer and Christy incorporate estimates of heat diffusion down 200m. Noting that the IPCC now says ECS is at least 2.5C, Spencer and Christy conclude it is more likely 1.9C. So yet again less than 2C, as so often in this reality-based series.

They calculate a slightly different version of ECS than most other studies called Effective (rather than Equilibrium) Climate Sensitivity, or EffCS. But for practical purposes it’s the same thing unless feedback mechanisms somehow get stronger as time passes after an increase in CO2. Spencer and Christy proceed by constructing what they call a 1-dimensional energy balance model to take account of energy inputs and outputs in the atmosphere. They note that for all their sophisticated design, climate models still don’t “conserve” energy even though conservation of energy is a fundamental property of physical systems and a fundamental axiom of physics. Energy can’t just appear out of nowhere nor can it magically disappear. What goes in must either stay in or come out. Yet climate models lose track of energy inputs and outputs and end up with too much or too little.

Remember that next time someone says they’re just following “the science”. Especially as Spencer and Christy point out that that 80 percent of modern climate models predict too much warming at the surface since 1970, possibly due to this non-trivial problem.

In order to compute the parameters of their simple model the authors of the paper we’re examining here needed estimates of the energy inputs, or “forcings” as they are called by insiders, over the 1970-2021 interval. They adjusted their model’s parameters so the predicted temperatures matched, as well as possible, the global average temperature estimate from 1940 to 2021. Using an older set of forcings (from the RCP6.0 scenario) they got an EffCS estimate of 2.5C, right at the low end of the IPCC range. But using the newest forcing series, called SSP, the estimate drops to 1.9C.

Well, there it is. We started this series by noting that if climate sensitivity is 2C or under, economists have shown that the Social Cost of Carbon is too low to justify any of the costly climate policies we’re being hit with, or any costly climate policies at all. And year after year studies published in the peer reviewed literature have shown that, despite models having high ECS values, the real world does not.

Do not let them tell you otherwise.

2 comments on “#ECS in the real world: Spencer and Christy 2023”

  1. Hmmm….models that don’t “conserve” energy….If you ignore energy balance, you can envision heating your house with a candle….of course, you’d be mistaken.

  2. "Yet studies from boreholes show heat diffuses hundreds of meters down, so Spencer and Christy incorporate estimates of heat diffusion down 200m." Ok, but the rate of change of temperature below 5m all the way to 200m is essentially ZERO (see https://www.researchgate.net/figure/Depth-dependence-of-annual-range-of-ground-temperatures-in-Ottawa-Canada-source_fig1_228683230 ) and so, the net effect of this minutiae on the temp of the atmosphere is also ZERO ! This is your typical manufactured research project from academia that proves nothing and goes nowhere. It's high time to defund the lot of them.

Leave a Reply

Your email address will not be published. Required fields are marked *