offsets

Passive Forests?

by

by Brent Sohngen (Sohngen.1@osu.edu)

I’ve been puzzling over the paper by Allen et al. (2025) since it came out earlier this year. On land use and forestry, the authors seemingly dump all human agency into the “passive sink”, that is, nature-based carbon storage obtained without humans. Specific forest-based mitigation can be used to lower net emissions in the land-using sectors, but that’s about it.

The passive forest carbon sink, of course, has been an existential threat to climate progress from the beginning. Many will recall when Al Gore arrived at COP 3 in Kyoto with a plan to get the US below 1990 emission levels by using credits generated in part by passive forests. This plan was met with skepticism.

However, from Kyoto we got the Managed Land Proxy (MLP), which is a simple tool developed by the Intergovernmental Panel on Climate Change (IPCC) that allows countries to include carbon from land use in their UN Framework Convention on Climate Change (UNFCCC) accounts. Countries declare which forests they are going to measure, and they include that carbon in their accounts.

Carbon derived in these measured forests can be added with energy emissions and included in the net emission calculations of any country.

The “proxy” part was included in the name because what happens in forests is not always a function of direct human action. Planting and cutting trees are quite obviously anthropogenic, but bug infestations, forest fires, or CO2 fertilization? These may only indirectly be linked to humans. For simplicity, under the MLP, any carbon changes are assumed to result from human agency.

Consider forest fires. In the US, most forest fires are caused by human ambivalence and there is a strong argument that our history of forest management, or lack thereof, makes fires worse. So it seems reasonable that fires be included with other human drivers in the US.

In Canada, things are different. Way back in the late 90s and early 00s, some of Canada’s forests were delivering net sequestration, so government designated land storing carbon as part of the MLP. Almost immediately, things turned south as bugs swept through their forests, and then they burned. Canada responded by recognizing this nuance and adjusting their carbon accounting.

A bigger issue than forest fires, however, is the growing global carbon sink due to CO2 fertilization and climate change. Allen et al. argue that most of this carbon would be taken up without human intervention – that is, it’s happening passively.

More importantly, Allen et al. claim that because this carbon removal will happen with or without direct human intervention, including it as part of country or corporate net zero commitments will increase long-run carbon concentrations.

The physics is indisputable, but the policy is suspect. Where is the human agency?

Most forests have been influenced directly by human actions. A report by scientists at Winrock International, for instance, found that only 16.6% of the world’s current forests could be classified as “stable”, meaning “those not already disturbed nor facing predictable near-future risks of anthropogenic disturbance” (see Funk et al., 2019).

Of course, with all that human disturbance out there – especially tree harvesting and deforestation – one could argue that it really is just pure dumb luck that forests are sequestering anything.

Is it really just luck?

Predictions from some of the original assessments of how people adapt to climate change suggest that forest carbon fluxes are not all just a function of happenstance. Figure 1 presents a comparison in a 1998 journal article predicting the natural response to a doubled CO2 experiment with a model that predicted how humans would respond to the climatic and ecological perturbations, including increased forest dieback through fires, shifts in growth rates due to changes in CO2 concentration as well as temperature and precipitation, and shifts in the distribution of where forest species could regenerate.

Figure 1: US forest carbon flux prediction with climate change using predicted climate impacts from a doubled CO2 experiment on forests from the VEMAP project integrated into an economic model of forest management, derived from Sohngen et al. (1998). Positive values in this figure represent a removal from the atmosphere, while negative values represent an emission.

The experiment in Figure 1 was implemented assuming a linear adjustment over 70 years, which over-estimates the speed of the shift in climate. The economic model also over-estimates the portion of land that will be managed intensively or somewhat intensively. The results are nonetheless instructive, illustrating how simple adaptive mechanisms of harvesting forests early before they burn, salvaging what cannot be prevented from burning, replanting species better adapted to survive in the future can accomplish, and simply managing forests for increased growth due to fertilization and climate.

The takeaway: humans won’t sit idly by as forests are influenced by climate change, at least in places where they have agency and can manage their forests.

Whether those impacts are bugs, fires, or increased growth due to CO2 fertilization, we are now 40 to 50 years into this experiment on our atmosphere, and the evidence is clear that foresters are responding to climate drivers (Davis et al., 2022).

Instead of being passive, foresters are actively engaged in adaptation. Assuming all flux, or even just the emissions from harvesting, is passive is a mistake.

The US has a large forest carbon sink. The suggestion of Allen et al. is that this carbon is accumulating without our help. Using data collected by the US Forest Service, however, it is clear that the accumulation of carbon has strengthened on private land while it has dissipated on public land (Figure 2). Over the same time period forest management has all but disappeared on federal lands in the US.

In a world with climate change, it is increasingly difficult to assume away human adaptation to the forces of change affecting forests. Although it may be convenient to assume that all of the observed carbon accumulation in forests lacks a human origin, that’s clearly not the case.

Figure 2: US forest CO2 flux on public and private land. Data obtained from Domke et al. (2022)

Why the WRI No Harvest Counterfactual is the wrong approach for carbon accounting in forests.

by

By Brent Sohngen (sohngen.1@osu.edu)

Have you ever wondered who owns the trees? Historically, of course, most observers would agree the trees belong to the landowner who can do with them what they wish. However, because 50% of the wood dry matter in trees is composed of carbon, the answer to this question is set to become less clear under some new rules proposed by the World Resources Institute for their Greenhouse Gas Measurement Protocol.

This new protocol will govern how companies and people who own forestland count the carbon in their forests. WRI is proposing to use an approach that only counts carbon when a stand is cut, ignoring any carbon changes that happen in forests owned by people who do not cut some stands, but instead let them grow. This approach uses what is called a stand-by-stand no-harvesting counterfactual to measure the effect of wood harvesting.

Figure 1 illustrates how the approach works. A growing timber stand is accumulating carbon from time-period 1 to 15 and is then cut. There is a harvest emission, shown by the immediate reduction in carbon. The forest begins to regrow either through planting or natural regeneration. At some point down the road, depending on the species, the new stand will have as much carbon as the original stand would have had if left alone.

The idea of the no-harvest counterfactual is to assume the level of carbon in an unharvested forest follows the hatched line and calculate the emissions associated with the difference between harvesting and not harvesting that stand.

Figure 1: Stand-by-stand no-harvest counterfactual

There are many problems with this no-harvest counterfactual approach. Mostly it is wrong because all wood harvesting is done in mature stands, which only become mature because someone left them alone long enough to get older. The WRI GHG protocol ignores any of this growth (what happens before t=15 in the figure). It focuses solely on the harvesting event, failing to account for the observed fact that landowners hold many trees they do not harvest, plant trees on old farms, let trees regenerate naturally on pastures, hold trees instead of growing crops, grazing, or building subdivisions and box stores. There are real opportunity costs with holding trees, but the WRI GHG Protocol assumes all these activities, and the associated costs, are irrelevant.

So why do they propose this scientifically challenged approach? Here’s my take.

First, WRI doesn’t want to provide carbon-based incentives for companies to grow and cut trees, especially in planted stands. This was clear in an article WRI scientists published in Nature in 2023 and the opinion piece by two of the nation’s most esteemed ecologists in the same issue. The general idea of this approach is that any tree harvesting is bad, including the millions of cubic meters harvested every year for fuelwood uses by folks in developing countries.

This approach – sometimes known as Proforestation – misses fundamental economic realities associated not just with supply and demand for wood, but also with emerging carbon markets. Economic studies have shown that wood market incentives increase carbon in forests (Tian et al. 2018). They have also shown that efficient carbon policies would incentivize lots of avoided deforestation, lots of reforestation and afforestation, lots of avoided old growth harvesting, and lots of improved forest management often coming in the form of extended timber rotations (Sohngen and Mendelsohn, 2003; Austin et al, 2020; Favero et al., 2020).

In these studies, more wood harvesting happens with carbon incentives over time because more carbon in forests means there is also more wood to harvest. Increased supply in turn lowers wood prices. Intensive plantations make up only about 10-15% of the new area in forests, even with high carbon prices. The rest of the carbon gains are predicted to happen in natural forests, where compensation levels for carbon would be large enough to support significant efforts to ward off fires.

Second, WRI has an additionality problem. For decades, people in carbon markets have argued for a strong additionality test, whereby tons of carbon sequestered by companies in the timber business cannot be used to offset fossil carbon emissions because the trees were grown for timber not carbon. The additionality problem arises because WRI and others cannot reconcile an accounting standard that would let a company use tons generated on its own forest as an offset against its own emissions with an approach that does not allow those tons to be sold in carbon markets due to additionality.

Concern about additionality is understandable, but plenty of approaches have been developed to handle it, including following the advice of van Kooten et al. (1995).

Third, WRI is worried that more scientifically appropriate approaches – such as the standard of measuring changes in stocks over time like the US Forest Service does for the US as a whole – will confer benefits on landowners for carbon fertilization and climate change, which have elevated the stock of trees (Davis et al., 2023). It is completely accurate that measuring carbon gains with stock changes over the area owned will credit landowners for carbon gains that are partly attributed to carbon fertilization or climate change. This means that people who hold forests could receive carbon benefits 15-25% larger than otherwise because of carbon fertilization and climate change.

Far from being a liability, as WRI claims, this is exactly what we should want because it means landowners are adapting to climate and market incentives.

Paying for the benefits of carbon fertilization, or in the case of the GHG Protocol, including them in insets generated from a land-based inventory, is the correct approach precisely because it encourages efficient behavior with respect to the atmosphere by landowners.

Rather than leading to more emissions, incentives that embody carbon fertilization values would reduce deforestation, increase afforestation, increase reforestation, reduce fire risks, and increase forest rotation ages. A recent US EPA report found that carbon sequestration would be 28% greater under policy incentives when carbon fertilization benefits are part of the incentives rather than ignored (USEPA, 2024).

In conclusion, WRI’s proposed GHG Protocol approach is the wrong policy approach. If the approach were correct, it could be extended to all forests for carbon accounting, but it makes no more sense in aggregate than it does as applied to a specific forest operation. Ultimately, it aims to reduce the value of carbon embedded in forests, constituting a legal “taking” of a resource in the United States that is worth billions.

WRI seems to have concocted this no-harvest counterfactual approach simply to limit how forest-owning companies count the carbon gains they provide. But it doesn’t work. In contrast, the economic literature illustrates that carbon incentives based on IPCC carbon accounting will lead to more of all forests. WRI should use this far more efficient, and environmentally sound, approach.