Climate Sensitivity and timing

[Snow_Miser]

My apologies for not seeing this article posted and re-posting it in another thread.

 

I think you actually posted it before me .

[Snow_Miser]

My guess is that it reflects updated data through the most recent decade discussed in this recent paper.

 

http://www.uwe-merckens.com/bilder/Wetter/ngeo.pdf

 

The ECS and TCR values in the Otto et al. paper were noticeably lower than the values in AR4.. still not as low as some skeptic climate sensitivity values, but definitely a downward revision.

[PhillipS]

The ECS and TCR values in the Otto et al. paper were noticeably lower than the values in AR4.. still not as low as some skeptic climate sensitivity values, but definitely a downward revision.

 

Several things to consider:

 

1)  The Otto et al paper is not peer-reviewed, it is a letter to the editor.  They can say whatever they want.

 

2)  Most of the lower climate sensitivity comes from the lower radiative forcing value for a doubling of CO2 (2xCO2).  The authors used the value 3.44 watts/m2 instead of the widely accepted 3.7 watts/m2.  That's a 7% reduction in the radiative forcing due to doubled CO2, so it's not surprising that the resulting sensitivity values are similarly lower.

 

3)  Almost half of the references, 7 of 16, were from the authors themselves, including the reference for the lower 2xCO2 forcing they used.  Given that authors often try to show a broad set of references to support their hypothesis it is puzzling that they had to use so much of their own prior work.

 

4)  The two equations they provide, for ECS and TCR, are odd - particularly equation 2.  At a CO2 level of 560 ppm (double pre-industrial 280 ppm) the equation becomes TCR = deltaT because F_2x / delta F = 1.  In other words, the transient climate response is whatever the observed temperature change is.  And that value is unbounded.

 

I've only had time to give the letter a quick read so I may have missed points.  I appreciate any analysis that can dispel my puzzlement.

[Snow_Miser]

I could also be that the recent papers with lower sensitivity ranges are due to short term natural climate variability 

with the PDO.

 

There was a recent paper published in Nature Climate Change that found that even over the last 20 years, the models substantially overestimated warming. The 1990s warming was likely juiced up substantially by the +PDO, and the last 20 years included this period of rapid warming. Even including the rapid period of warming, the models still simulated too much warming by over a factor of two over the last 20 years. The PDO has played a role in this discrepancy, but there may be other factors at work as well, such as a lower climate sensitivity than in the GCMs.

 

http://www.see.ed.ac.uk/~shs/Climate%20change/Climate%20model%20results/over%20estimate.pdf

 

The CMIP5 models simulating too little warming in the early-20th Century and oversimulating the cooling with Volcanic eruptions like with Mount Pinatubo indicates that they are probably too sensitive to changes in radiative forcing.

 

I also find it amazingly remarkable that the IPCC sensitivity estimates are no more certain than they were in 1990. Now we have a most likely sensitivity of 1.5-4.5 Degrees C again from the IPCC. What I also find remarkable is how they could increase their confidence on the causes of the late-20th Century warm period when their sensitivity estimates have become less certain than they were in AR4.