September 19, 2019 by Andrew P. Jones
Note the authors of the paper discussed below have issued an erratum retracting some of their claims that we had pointed out initially. Additionally, En-ROADS has been updated since this post and so some scenarios may appear differently. (June 11, 2020)
By Andrew P. Jones, Dr. Lori Siegel, and Dr. Florian Kapmeier
Research published recently in Science by Jean-Francois Bastin et al. (2019) entitled “The global tree restoration potential” received high media attention all around the world. The authors found that “ecosystem restoration [is] the most effective solution at our disposal to mitigate climate change” (p. 78). Newspapers and journals reported with great excitement about this study, including leading outlets such as The New York Times and The Guardian, with eye-catching headlines like “The trees might save us yet” or “Tree planting has mind-blowing potential to tackle climate crisis”.
Planting trees sounded like THE silver-bullet solution to meet the Paris goals – obvious, cheap, effective, and apparently easy and fast to implement. Our team at Climate Interactive has assessed what is behind Bastin et al.’s findings.
In their research, Bastinet al. analyzed “how much tree cover might be possible under current or future climate conditions” (p. 76). The resulting tree carrying capacity map shows both the presence of forest in all existing forested land and the tree cover that could exist beyond already existing forested land. The authors estimate the total potential land area to be about 4.4 billion hectares (under current climate conditions).
The study goes on to state that these newly planted trees could remove an additional 205 Gigatons of carbon. In the official ETH press release, the authors stress the following two insights:
Our team at Climate Interactive has analyzed these two statements below with our En-ROADS climate and policy simulation model. We first reflect on the assumptions themselves, and then test the policies’ impact on global temperature by 2100 – will it really have “mind-blowing potential”?
1)**Assessing this first statement: **“Once mature, these new forests could store 205 billion tonnes of carbon: about two thirds of the 300 billion tonnes of carbon that has been released into the atmosphere as a result of human activity since the Industrial Revolution.”
We confirm the underlying calculation:
205 tons C/300 tons C = 68% (1)
For our analysis with En-ROADS, we translate carbon emissions into CO2emissions, representing roughly 750 Gtons of CO2removal and 1100 Gtons CO2for the total emissions (a CO2 molecule is 3.7 times heavier than only the C atom (Le Quéré et al., 2018)), thus
750 Gtons CO2/1100 Gtons CO2= 68% (2)
But the 1,100 Gtons CO2emissions in the denominator is misleading. Humans have emitted much more than this since the beginning of the industrial revolution. Bastin et al. removed, we assume, the natural carbon uptake by the world’s ecosystems, subtracting 1,000 Gtons CO2that oceans and biomass (plants and soils) had already absorbed since the beginning of the industrial revolution. As these emissions polluted the environment, e.g., by acidifying the oceans, a more reasonable number to use is all cumulative anthropogenic carbon emissions to date without subtracting absorption by oceans and biomass. These are about 2,100 Gtons CO2:
750 Gtons CO2/2100 Gtons CO2= 36% (3)
Hence, the maximum of the removals from afforestation are about 36% of total anthropogenic CO2emissions from 1870 until 2018, equaling closer to one-third as opposed to two-thirds.
But still, this is not the full picture. The authors report (p. 78) that the uptake will not happen instantaneously, as it would take several decades for forests to grow. We agree. There are considerable time delays involved, including the acquisition of land and the growing of the trees (see also Sterman et al., 2018a and 2018b, two papers by three of our team members on bioenergy dynamics). We therefore consider the total amount of CO2emissions from the beginning of the industrial revolution until the future when the trees have all grown – let’s assume until 2100.
Under the BAU scenario in En-ROADS, human activity will emit another 5,700 Gtons CO2. When adding past and future anthropogenic CO2emissions, total cumulative CO2emissions will be around 7,800 Gtons, thus:
750 Gtons CO2/7800 Gtons CO2= 10% (4)
The carbon removal from the newly planted trees represents, at the very high end, approximately 10% of total cumulative human emissions.
To summarize: The absorption potential is not 2/3, but more likely, at the high end, 10% of total cumulative anthropogenic emissions through 2100.
While this does not sound as promising as formulated in the original paper, we strongly emphasize that this is still significant and should be enough to justify investment in this climate policy.
Bastin et al. made another strong assumption in their analysis, when assuming that “there is room for an extra 0.9 billion hectares” (p. 76) of land available. The authors state that “they span the entire range of environmental conditions, from the lowest to the highest possible tree cover” (p. 76).
Considering the entire range of possibilities and thus a more conservative perspective, the Royal Society Report (2018) has issued possible ranges of global CO2removal potential. The upper limit for afforestation is at an annual uptake of 20 GTCO2, roughly in line with Bastin et al.’s findings of 900 million hectares of land. The lower end of the Royal Society Report lies, however, at an uptake of 3 GtCO2 annually. Considering this lower end, we arrive at only 1%:
70 Gtons CO2/7800 Gtons CO2= 1% (5)
To summarize again: The absorption potential is not 2/3, but more likely 1-10% of total cumulative anthropogenic emissions through 2100.
While afforestation plays a significant role in mitigating climate change, it seems as if the numbers used by Bastin et al. are somewhat misleading. The range of anthropogenic CO2 emissions removal from afforestation varies from 10% at the high-end to 1% at the low-end. Now, even if all the available land that could be afforested will actually be afforested, growth of trees is not certain. Trees in new (and existing) forests can be cut (legally or illegally), burned in wildfires, and/or die from diseases and insects. All this is more likely in a warming world with growing population in which climate zones are moving.
We also need to keep in mind that relatively lower atmospheric CO2 also means less flux of carbon into other biomass stocks and into ocean (Jones et al., 2016). Therefore, atmospheric CO2 will not fall by the cumulative amount of carbon held in new forests, but less, reducing the impact on warming and climate change even more.
Now, let’s test whether afforestation really has the “mind-blowing potential,” as stated by The Guardian. To test the high-end scenario in En-ROADS, maximum available land for afforestation is increased to 900 Mha and a rapid afforestation planting time of 12 years is assumed, starting in 2020.
This scenario leads to a warming of 3.9 °C instead of 4.1 °C above preindustrial levels by 2100 (Figure 1 below).
On the left hand graph, we observe that net CO2removals from afforestation first increase very quickly at an exponential rate, then reach an inflection point to increase at a decreasing rate until reaching a peak of about 10.7 Gtons in 2080. The graph in the upper-right hand shows greenhouse gas emissions by gas. An afforestation policy does not affect the other greenhouse gases, thus they are the same as in the BAU scenario. But, CO2emissions from land use & forestry start to decline slowly after 2020 and become negative after the 2040s: the forests take-up more CO2 than they emit and the forests become “negative emissions” (The Royal Society, 2018). CO2 removals from afforestation do not stay on this high level (left graph) but decrease because of natural processes of stored carbon after sequestration (typical biomass and soil release rates, through bacterial respiration (decay) and wildfire), which is why land-use CO2 slightly increases towards the end of the century.
The scenario settings used in En-ROADS to test the impacts of this policy are:
The key insight here is that while afforestation does remove a critical mass of CO2and reduces temperature and afforestation efforts are a very crucial contribution to climate change mitigation, it is not a silver-bullet. It does not have “mind-blowing potential”, the trees do not “save us”, and they do not “save the climate”. Afforestation helps, and it is an important part of a silver-buckshot suite of policies and behavior change.
2)**Now, assessing this second statement: **“[T]his would be the most effective method to combat climate change”, which is similarly formulated in their paper: “This places ecosystem restoration as the most effective solution at our disposal to mitigate climate change” (p. 78).
Bastin et al. find that extremely high levels of afforestation could remove 750 Gtons CO2(or, 205 Gtons of carbon). We use En-ROADS to assess what other actions, policies, or behavior changes are required to remove about 750 Gtons CO2 by 2100. We find that this uptake is similar to:
We conclude that afforestation “would be the most effective method to combat climate change” is likely not true. There are many approaches that all accomplish mitigating climate change of which afforestation is a very important way to tackle the climate crisis.
Using the same scenario settings as outlined for Figure 1 above, we take a look at CO2emissions by source (Figure 2 below). The green area below the “0” line represents the cumulative uptake of CO2 from forests (750 Gtons). The area above it represents the 7,800 Gtons that we will emit in the BAU scenario through burning coal, oil, and gas (grey area), CH4(blue), N2O (pink), and F-Gases (yellow).
We see that afforestation has a modest effect. However, when we apply other policies as well (Figure 3 below), the green area (carbon dioxide removal) becomes larger in comparison to the other areas (carbon dioxide emissions). In this scenario, fossil fuels are kept in the ground; coal peaks around 2022, gas around 2030, and oil around 2035.
Note that, in this two degree scenario, the afforestation proposed in Bastin et al. would remove approximately 18% of cumulative CO2 emissions from 1870-2100.
The scenario settings used in En-ROADS to test the impacts of this policy are:
Transport Energy Efficiency
Buildings and Industry Energy Efficiency
Methane & Other Gases
Furthermore, restoring forests is important for many reasons other than their (important) role in removing carbon from the atmosphere. First, they provide a natural habitat for many species that are currently under threat, thus ensuring biodiversity. Second, they stabilize soil and thus prevent land degradation. Finally, they absorb groundwater and provide important resilience against flooding as a result of extreme rainfall, which is more likely to occur in a warming world (our team at Climate Interactive calls these systemic solutions that protect the climate while also improving health, equity, and well-being “multi-solving”).
Our assessments show that afforestation efforts could remove likely only 1-10% of anthropogenic CO2emissions from the beginning of the industrial revolution until 2100, (up to 18% in a low emissions trajectory) in contrast with Bastin et al’s estimates of two-thirds. Additionally, the idea that planting trees will save us from climate change is incorrect. Afforestation is a big, important part of a broad suite of policies that address climate change, but it is not a silver bullet solution. If managed well or kept wild, trees (beyond removing carbon) can increase biodiversity, stabilize soil, retain groundwater, and increase resilience against flooding.
Bastin, J.-F., Finegold, Y., Garcia, C., Mollicone, D., Rezende, M., Routh, D., … Crowther, T. W. (2019). The global tree restoration potential. Science, 365(6448), 76-79. doi:10.1126/science.aax0848
Jones, C. D., Ciais, P., Davis, S. J., Friedlingstein, P., Gasser, T., Peters, G. P., … Wiltshire, A. (2016). Simulating the Earth system response to negative emissions. Environmental Research Letters, 11(9), 095012. doi:10.1088/1748-9326/11/9/095012
Le Quéré, C., Andrew, R. M., Friedlingstein, P., Sitch, S., Hauck, J., Pongratz, J., … Zheng, B. (2018). Global Carbon Budget 2018. Earth Syst. Sci. Data, 10(4), 2141-2194. doi:10.5194/essd-10-2141-2018
Sterman, J. D., Siegel, L., & Rooney-Varga, J. N. (2018a). Does replacing coal with wood lower CO2 emissions? Dynamic lifecycle analysis of wood bioenergy. Environmental Research Letters, 13(1), 015007. doi:10.1088/1748-9326/aaa512
Sterman, J. D., Siegel, L., & Rooney-Varga, J. N. (2018b). Reply to comment on ‘Does replacing coal with wood lower CO2 emissions? Dynamic lifecycle analysis of wood bioenergy’. Environmental Research Letters, 13(12), 128003. doi:10.1088/1748-9326/aaf354
The Royal Society. (2018). Greenhouse gas removal. Retrieved from https://royalsociety.org/-/media/policy/projects/greenhouse-gas-removal/royal-society-greenhouse-gas-removal-report-2018.pdf: