revisit statistics on contributions to global C sink

[teixeirak]

looking back at the Pugh reference, I'm not sure where I got the cited 30%.

Harris_global_2021 presents this info in Table 1: 47% of global total net GHG flux!

I'd like to look back at Pugh and maybe some others to get a better understanding of the different estimates.

[teixeirak]

Pugh et al. 2009

Careful calculations based on Pugh

• Total forest C sink estimated at 2.15 Pg C yr (7.88 Pg CO2e yr-1)

  

• Their best estimate of C sink in temperate broadleaf forests (essentially all regrowth) is ~300 Tg C yr^-1^ (Fig. 3a and c), or 0.3 Pg C yr. It would be only ~100 Tg C yr^-1^ using alternate land use map.

• Doing the math, 0.3/2.15*100=14% of total global forest C sink. (I don't know how I calculated the current estimate of 30%, but that's not correct using Pugh's numbers. Perhaps I was counting only the portion of the sink attributable to environmental change, as opposed to regrowth.)

Evaluation of Pugh

• C sink estimate is similar to the net C sink estimate of Harris et al. 2021 (7.6 Gt CO2e yr-1), but much smaller than estimate of gross removals (15.6 Gt CO2eyr-1), which is where the comparison should be.

• It is also much smaller than other previous estimates:

  

• One reason for this discrepancy is that Pugh et al. use a coarser spatial scale that doesn't capture smaller-scale disturbances (e.g., shifting agriculture), but this isn't enough to explain the difference. Using a different land use history dataset, tropical forest regrowth would be much larger (so temperate broadleaf forest sink would be smaller):

  

• But the main reason for discrepancy seems to be a much lower estimate of the C sink potential of tropical regrowth forests-- their models count tropical regrowth forests <50 yrs old as an overall C source. My first instinct is that this is insanely low. ForC currently has 2 NEE/NEP records for young tropical broadleaf forests (ages 4 and 50), both strong sinks (3.6 and 4.4). Baldocchi et al. 2008 shows regrowth forests (mostly temperate) becoming sinks at ~ age 20. So I don't believe this.

• If tropical regrowth forests were better represented, they'd account for a much larger portion of the C sink, and therefore temperate broadleaf forests would be lower. So the 14% listed above is an overestimate. But I wouldn't trust the global forest C sink estimate from Pugh at all.

Harris et al. 2021, Nature Climate Change

This study uses a book-keeping method based on 30m-resolution land cover and IPCC accounting approach. I really like this study.

This estimates that temperate forests 28% of gross CO2 removals by forests and 47% of net removals:

The really unfortunate thing here is that temperate forests aren't broken up into broadleaf and evergreen.

Synthesis

Extratropical forest gross CO2 removals

Pugh (does not separate temperate vs boreal for needleleaf and mixed):

• ~1150 Tg C/yr (1.15Pg) for temperate and boreal forests combined
• temperate broadleaf ~ 26% of this

  Harris :

  • temperate: 4.4 Pg CO2/yr = 1.2 Pg C/yr
  • boreal: 2.5 Pg CO2/yr= .68 Pg C/yr
  • subtropical: 1.6 Pg CO2/yr= .44 PgC/yr
  • extratropical total: 2.32 PgC/yr

    comparison

  • Harris estimate for extratropical forest C sink is ~2x Pugh estimate.

• Estimates are difficult to compare, but very different
• Without trusting the Pugh global estimate, we can't give an estimate of % of global C sink.
• Best statement would be something like the following (all or part): "Temperate forests account for ~47% of the global forest net C sink (~1.0 Gt C yr; Harris et al), with an estimated uptake of >0.3 GtC yr in temperate broadleaf forests (Pugh et al.)"

[teixeirak]

I'm still not sure if we want to cite the 300 Tg C/yr statistic. I need to look more at the paper to decide whether to trust it.

[teixeirak]

Also check this: https://www.science.org/doi/10.1126/sciadv.abe9829

[teixeirak]

R1 comments: "L52. I read 30% in Fredlingstein."

This is pretty complex, and answers differ depending on how it's counted. I've started to go down a rabbit hole trying to figure out the best answer, but need to pause.

Friedlinstein et al. present the total land sink, which includes non-forests (2.9 GtC yr-1 in 2020). They also present fossil fuel emissions (9.5 GtC yr-1 in 2020).

Harris et al. 2021 separate out forests, and estimate global forests were a net C sink of -7.6 Gt CO2e yr-1 = 2.07 Gt C yr-1 (statistic in their abstract). The 20% currently cited was based on that figure (2.07/9.5=.2).

However, looking deeper at Harris, this number isn't necessarily what we want. I need to come back to this.

[teixeirak]

CO2 removals through GROSS biomass increases

This represents biomass increases and does not include losses to deforestation or natural disturbances.

from Harris et al. 2021:

(15.6+.16)*12/44=4.3 Gt C yr-1

In recent decades, tree growth in Earth's forests has sequestered CO~2~ from the atmosphere at a rate of ~r round((15.6+0.16)*12/44,1)Gt C yr^-1^ (gross removals from Ref. [@harris_global_2021]), equaling almost half of anthropogenic CO~2~ emissions (fossil fuels + cement[@harris_global_2021]) and thereby slowing the pace of atmospheric CO~2~ accumulation and climate change.

CO2 removals through NET biomass increases

This represents net biomass change, including losses to deforestation or natural disturbances.

[teixeirak]

I decided to focus on CO2 removals through net biomass increases, from Harris et al. , as gross removals is quite dependent on the scale of detectability and hard to relate to biological fluxes.

I've revised the text with careful documentation of sources in the .Rmd (as comments).

I'm pretty happy with this now.

[teixeirak]

Reopening because I still want to check this: https://www.science.org/doi/10.1126/sciadv.abe9829

[teixeirak]

Comparing Harris et al. 2021 and Xu et al. 2021

Harris: net forest sink = 5.8 GtCO2 /yr = 1.58 GtC/yr Xu: 0.23 to 0.88 PgC year−1

Xu et al provide a helpful comparison:

Given the context here (focus on forests), I think it's more appropriate to stick with Harris.