Trees and CO2

You know they love the stuff. New in PNAS:

Shevliakova E, Stouffer RJ, Malyshev S, Krasting JP, Hurtt GC, Pacala SW. in press. Historical warming reduced due to enhanced land carbon uptake. PNAS doi: 10.1073/pnas.1314047110

Tiny excerpt: “…we estimate that enhanced vegetation growth has lowered the historical atmospheric CO2 concentration by 85 ppm, avoiding an additional 0.31 ± 0.06 °C warming.”

What they did: Very roughly speaking they ran an Earth Systems model from 1861 to 2005, with fossil-C being burned at the known rates, with or without trees being fertilized by +CO2. Also varied land-use changes during the period, but that takes longer to explain. The avoided warming is based on the 3 oC per CO2 doubling that gets its own discussions here at PC.

A similar rate carbon sequestration on land has been shown by ‘bottom up’ accounting, and we have talked about some of those (like Yude Pan’s) before. This is the first (or at least the broadest) ‘top-down’ study. At their 2005 endpoint, there was no indication that the trees were slowing down. So while we know that trees also need water, nitrogen, etc., and prefer not being killed by beetles, this ‘ecosystem service’ was still humming along. I really don’t know why they ended it in 2005.

Another new one:

Ostberg S, Lucht W, Schaphoff S, Gerten D. 2013. Critical impacts of global warming on land ecosystems. Earth Syst. Dynam., 4, 347–357. doi: 10.5194/esd-4-347-2013

As you might guess from the title, this modeling study (based on the several IPCC RCP scenarios) anticipates large changes in what vegetation types grow where. What they didn’t do was any prediction of how the global plant-carbon sink might change over this century. I think this could have been done.

The first falls in the category of good news (sub category already known). Second one, I can only squeeze into that category in that a really large hit was needed to convert Amazonia to savannah. That has been a topic of interest. -. -. --.- News from the edge of terrestrial C global modeling research.