Twitter | Search | |
Andrew Dessler
Colleagues and I have a new paper out evaluating methodology for estimating equilibrium climate sensitivity (ECS) from the 20th-century historical record: 1/
Reply Retweet Like More
Andrew Dessler 15 Jan 18
Replying to @AndrewDessler
Such analyses use the first law of thermodynamics — energy is conserved — to estimate ECS from estimates of forcing, temperature change, and change in planetary energy balance. 2/
Reply Retweet Like
Andrew Dessler 15 Jan 18
Replying to @AndrewDessler
To evaluate the methodology, we ran a climate model 100 times over the period 1850-2005; for each ensemble member, we calculate ECS. Here is what the resulting ECS distribution looks like 3/
Reply Retweet Like
Andrew Dessler 15 Jan 18
Replying to @AndrewDessler
We know the actual ECS of the model from a 2xCO2 run, and it’s 2.9 K. Thus, this methodology does a poor job — even if you know forcing, temperature change, and change in planetary energy balance perfectly. 4/
Reply Retweet Like
Andrew Dessler 15 Jan 18
Replying to @AndrewDessler
This adds to growing list of problems with these calculations: large & evolving uncertainty in forcing (Forster, 2016), different forcing efficacies of greenhouse gases and aerosols (Shindell, 2014; Kummer and Dessler, 2014), 5/
Reply Retweet Like
Andrew Dessler 15 Jan 18
Replying to @AndrewDessler
geographically incomplete & inhomogeneous observations (Richardson et al., 2016), and 20th century ECS being low-biased with respect to the actual ECS (Armour, 2017; Proistosescu and Huybers, 2017). 6/
Reply Retweet Like
Andrew Dessler 15 Jan 18
Our paper does not tell us where our 20th century falls within the theoretical ensemble of all possible 20th centuries. However, work presented at the recent AGU meeting by and show that the actual 20th century gives us an ECS that is low biased. 7/
Reply Retweet Like
Andrew Dessler 15 Jan 18
Replying to @AndrewDessler
Based on all of this work, I would expect future assessments to raise the low end of the ECS range to >= 2 K. 8/8
Reply Retweet Like
John Mitchell 15 Jan 18
Replying to @AndrewDessler
Engineer napkin test: ECS=TCR/0.6 ; current temp = 0.5 of 2XCO2 TCR (due to aerosols, lag time etc) so, 1.1C/0.5/0.6 = ECS = 3.7C. But not applicable to future conditions, actual 2XCO2 likely closer to 6.0C (sea ice & cloud albedo + DMS red due 2 acidification. . .)
Reply Retweet Like
Brad Rosenheim 15 Jan 18
Replying to @AndrewDessler
Your distribution does show the heavy tail that so many different estimations do. Realizing this is partly functional form, your method seems more empirical yet still shows it.
Reply Retweet Like
Alexander Ač 15 Jan 18
Replying to @AndrewDessler
And maybe the central estimate as well ;-)
Reply Retweet Like
STOP BREXIT NOW 15 Jan 18
Replying to @AndrewDessler
Huh? That looks like a pretty good job to me. Is your complaint that there is some uncertainty?
Reply Retweet Like
Andrew Dessler 15 Jan 18
Replying to @jamesannan
If you have 100 ensembles of the 20th century, then it looks great. If you only have one (which we do), then not so great.
Reply Retweet Like
MichaelCrow 15 Jan 18
Using known change in insolation over the year in the extratropics, compared to change in temp puts it below 0.02F/W/m^2. Approach and data here
Reply Retweet Like
Deep Simplicity 18 Jan 18
Replying to @AndrewDessler
What's that mean to us commoners, like F and C?
Reply Retweet Like
Grassland 18 Apr 18
Are this papers findings relevant to or 's recent papers on ECS?
Reply Retweet Like