11 F Warming by 2050?? No way.

Note added 5/31/12: Reuters has issued a correction to the article this post was based on. Click here for details.

--

The new Romm Record of 11°F apparently comes from a Reuters article where Faith Fatih Birol, the chief economist of the IEA, says we are on track for that amount of warming by 2050:

"When I look at this data, the trend is perfectly in line with a temperature increase of 6 degrees Celsius (by 2050), which would have devastating consequences for the planet," Fatih Birol, IEA's chief economist told Reuters.

That seems preposterous to me, on its face and on further analysis, which seems to me implies an additional warming (from now) of about 1.5 to 2°F by 2050. Here's why.

The carbon-climate response (CCR) is about 1.5°C per trillion metric tons of carbon emitted. That's the number given in the recent Swart and Weaver Nature Climate Change paper analyzing expected from the Alberta tar sands, with a range of 1.0–2.1 °C per Tt C (5th–95th percentile). That result comes from a 2009 Nature paper by Matthews et al., and this function includes feedbacks.

So let's add up the emissions.

Fossil fuels and cement production: 
Here is historical carbon emissions data from CDIAC; it includes fossil fuel combustion and cement production, and comes to 347 GtC as of 2008. Add in the recent IEA figures -- which are for fossil fuel consumption only -- and using the CDIAC data to estimate cement-carbon for the last few years, and the total emissions in that sector are about 375 GtC.

Land use changes:
Here is CDIAC's estimates of carbon emissions from land use changes. It the 1800s it was several times combustion emissions, but since 1933 it has been less and is now about 16% of combustion emissions. Add these up and this sector's total emissions (since 1850) are about 165 GtC.

Total Emissions:
So total emissions since 1850 are about 540 GtC, for which the CCR gives an expected warming of 0.81°C , which is what's been observed. Here's the plot:

So cumulative emissions by 2050 look to be around 2000 GtC, give or take. [That's the upper limit of applicability of this CCR.] That gives an expected [total] warming of about about 3°C (2.2°C more than today), or a total of 5.4°F, half of Birol's number.

Even that seems large. It means the next 38 years will see an average temperature increase of 0.58°C/decade, which is three and a half times the GISS warming rate of the last 30 years. Be honest: does that seem likely to you?

Even taking the upper bound of the CCR range, 2.1°C (which, let's note, overestimates the warming that's taken place so far by over 30%), gives a total warming of 7.6°F -- an additional warming of 0.9°C/decade from now until 2050.

I just don't see how anyone can get 11°F out of these numbers. 2050 cumulative emissions would need to be about 3,500 GtC for that to happen, which seems very unlikely. (At the CCR's upper bound it would be 2,500 GtC, which seems unlikely, though perhaps not "very" unlikely.)

Conclusion: Romm is rong -- in this case, the "alarmist" label definitely applies.

Note added: On second thought, 2050 cumulative emissions could be much less than I assumed -- projecting the line in the plot above, starting at 1950, gives cumulative emissions of 1300 GtC in 2050, not 2000 as I guessed. That gives a total warming of only 3.5°F (2.0°C) by 2050.