Read on Twitter
Many years ago I had the idea that we might be able to use tropical forest plot data to constrain the tropical land contribution to Equilibrium Climate Sensitivity (following something @PFriedling said in a meeting I was at). The resulting paper is out in Science today. A thread:
First it took a long time to piece together the data - 500,000 trees across 813 forests in 24 countries in South American, African, Asian and Australasian tropical forest. And second find a talented researcher to lead the analysis, @mjpsullivan, see https://science.sciencemag.org/content/368/6493/869?_ga=2.157484733.602342098.1590141391-1781683786.1468596102
Tropical forests are carbon-dense and productive, so small changes can have big global impacts, as we showed in this recent Nature paper, intact tropical forests are a large carbon sink slowing climate change today, but that sink is in decline: https://www.nature.com/articles/s41586-020-2035-0
But what might happen long-term as we warm the planet? Tropical forests are expected to lose carbon as the planet warms, as hotter forests support, on average, smaller trees as respiration costs increase and photosynthesis slows in the middle of the day as optimum conditions.
This loss of carbon from tropical forests for each degree of warming is what we hoped to get from the plot data, and see where it falls in relation to climate models.
We did this by comparing forests growing in different climates, i.e. in geographical space, as this allows us to see how tropical forests respond to climate after having had time to adapt. That is, after species best adapted to those conditions have got there.
We found that maximum daily temperature is the most important determinant of tropical forest carbon storage, above all other climate and soil variables. This means it is the curtailment of photosynthesis rather than higher respiration costs that is important in the long-term.
That is counter to the view from 1 yr studies of individual hot years, that seem to show night-time minimum temperatures are important, suggesting that there is some acclimation of respiration to warmer temperatures.
The basic result is a loss of 14 billion tonnes carbon per degree C increase maximum temperature. But, there is a threshold, ~32 deg C maximum daily temperature, where carbon losses become large. Global temperature rises push more and more forest over the threshold.
Below the threshold, our approach shows a lower sensitivity to warming in these tropical forests than most CMIP5 models, but above the threshold, our sensitivity is higher than CMIP5 models. Its is tricky to compare, due to our use of max temp, looking at live biomass carbon...
But, overall the non-linearly is important as more and more forest will be pushed beyond a high-temperature threshold limiting photosynthesis as the world warms.
The analysis assesses temperature (and drought), but those losses are offset by boosted tree growth due to more carbon dioxide in the atmosphere in the future, which cannot be obtained from a space-for-time study, but we can use published estimates.
What happens between now and stabilizing the climate at 1.5C or 2C or 3C ? How do the losses from higher temperatures match up to the gains from CO2?
At 1.5C most of Amazonia's forests are pushed above the 32C threshold, and and at 2C warming no matter how strong the CO2 fertilization effect, they will lose carbon. While the Congo basin stays below the threshold for longer, and tends not show large losses at 2C warming.
These results align well with the analyses in the Nature study of the saturation of the carbon sink, showing saturation first in Amazonia, with Amazonia becoming a carbon source in the decades ahead, while the decline in the African sink is more modest.
Overall, to see high levels of carbon storage in tropical forests in the long-term, we need to stabilise the climate in line with the Paris Agreement, protect forests from logging and stop fragmenting forests to species can move as the climate changes.
Lastly, there are 225 authors and literally thousands of people involved in making 2 million tree diameter measurements comprising 10,000 species, so big thanks to everyone for a huge team effort. Lets hope it helps efforts to protect forests and drives down fossil fuel emissions
