Sunday, February 26, 2017

Temperature vs Log(CO2) Since 1970

Following Joe's suggestion (thanks), here I plot temperature since 1970 versus log(CO2):


It's a better fit, avoids the 1940-1970 aerosol mess, and for CO2's climate sensitivity gives S(CO2) = ΔT(x=1) = T(x=1) - T(x=0) = slope = 2.6°C. With no evidence that feedbacks have done much yet -- except for a negative feedback from aerosols (traditional pollution), which are reducing the apparent climate sensitivity to CO2.

8 comments:

DocRichard46 said...

Sorry, I don't understand. What are the axes?

David Appell said...

The y-axis is the temperature anomaly, and the x-axis is the log-base 2 of CO2 concentration divided by the initial CO2 concentration.

That help?

DocRichard46 said...

Thanks

climate data analysis said...

I was glad to see JoeT post this (hi, Joe!), and to see David Appell do a post about it. My own version is over here:

https://climategraphs.wordpress.com/2017/02/28/first-blog-post/

Some takeaways:

* It's a very intuitive way of visualizing the relationship between CO2 and temperature

* Over the past half-century it's pretty robust. The trends 1960-1990 and 1990-2016 are very similar, to each other and to the entire 1960-2016 trend.

* The trends are pretty similar across data sets. From 1960-2016 it's 2.6 in GISTEMP, 2.4 in Berkeley Earth and Cowtan&Way, and 2.9 in RATPAC-A (troposphere). From 1979-2016 it's 2.6 in RSS TTT (troposphere).

* Climate models also show this linear relationship persisting despite the presence of other forcings.

The big picture is that temperatures are marching upwards in close correlation with CO2, regardless of short-term fluctuations from ENSO etc.

Figure 5 at the linked site pretty much says everything that needs to be said.

JoeT said...

Great web site CDA! I look forward to reading more. Out of curiosity, did you do UAH v6? Have we run across each other before?

David, don't you think your graph shows that feedbacks are important? At the very least, the water vapor feedback. Possibly the ice albedo feedback as well.

David Appell said...

Joe, wouldnt positive feedbacks start showing up as a nonlinearity, an upward deviation from the straight line?

JoeT said...

Not necessarily. I think the water vapor feedback would show up right away. It was my motivation for taking the calculation back to 1850 or so. Even then one sees a slope greater than the Planck response, although clearly the fit isn't all that great. The interesting question is whether we are seeing the ice albedo effect really kick in now. I think that requires a far more sophisticated analysis than the B.o.E. estimates that we are doing.

climate data analysis said...

Great web site CDA! I look forward to reading more. Out of curiosity, did you do UAH v6? Have we run across each other before?

Thanks, Joe. UAH v6 has a slope that's quite a bit lower (1.8) and noisier. I am dubious about the value of UAH for a number of reasons, but will probably include it in later drafts. RATPAC-A provides data for the troposphere going back much further than UAH, and RSS TTT seems like a better product for the post-1979 period (I know you recommended that in your earlier comment here).