Saturday, July 28, 2012

If the BEST rumor of 1.5 C Warming is True

Ron Bailey at Reason Magazine says there is a rumor that next week the BEST group at Berkeley will announce a finding of a 1.5°C temperature increase since 1750.

OK, it's just a rumor. But what would it mean for climate sensitivity?

The CDIAC estimates of historical CO2 emissions start in 1751 for fossil fuel combustion, but 1850 for land use changes. What if you included land use emissions from 1750 - 1850?

Since 1850, the totals are:

cumulative fossil fuel consumption emissions = 374 GtC
cumulative land use emissions = 165 GtC

where I have filled in the last few years of fossil fuel emissions from IEA data and, since CDIAC's land use emission dataset goes only to 2005, assumed annual land use changes since then are equal to 2005's emissions of 1,467 MtC/yr.

Since HadCRUT4 global surface warming since 1850 (slope × interval length) is 0.73°C, that gives a carbon-climate response of 1.37°C/TtC, which can be compared to the Swart and Weaver CCR function of 1.5 °C/TtC (which comes from a 2009 Nature paper by Matthews et al, which finds the 5th - 95th percentiles of the CCR to be 1.0 - 2.1 °C/TtC).

So what might earlier land use changes, from 1750 - 1850, add to cumulative carbon emissions? Well, from 1850-1870 those emissions averaged 535 MtC/yr. Assuming that's what it was starting in 1750, and adding in CDIAC's numbers for fossil fuel emissions for that period (which are very small), cumulative emissions from 1750-1850 would be 54 GtC, or only 10% more than the total from 1850-2010.

That would only mean an extra 0.08°C (0.05-0.011°C) of warming (5th-95th percentile), for a total of 0.87°C (0.58-1.22°C).

So if the rumor is true and the warming since 1750 is 1.5°C, maybe the global average surface temperature is on the high side of the CCR.

These historical carbon accountings are difficult and the error bars are significant. On the other hand, the CCR should hold over any time period where natural fluctuations average to zero, and its value for recent decades does tend to be high:

CCR of last 20 years: 1.99 °C/TtC
CCR of last 30 years: 2.28 °C/TtC
CCR of last 50 years: 1.91 °C/TtC

1 comment:

Paul S said...

BEST is reporting only land-surface temperature at this point. All the major indices report ~1.2ºC land warming over their full series (starting 1850-1880).

I would wager the remaining few tenths can be amply explained by a period of major volcanic activity around 1800.