The historical reliance on sinks

Since the negotiation and adoption of the Kyoto Protocol, land-based solutions, particularly tree planting and forest conservation, have received sustained attention in climate change discussions as potentially large and cheap solutions. The scientific community in the Intergovernmental Panel on Climate Change (IPCC) multiple assessment reports has continuously indicated that anthropogenic land-based CO₂ fluxes are critical to balance anthropogenic CO₂ emissions and removals for two reasons [3]: first, halting emissions from deforestation and forest degradation; and second, enhancing CO₂ removals through afforestation, reforestation, and improved forest management, which are needed to offset residual GHG emissions. It is well known that CO₂ emissions will continue to drive warming until they are fully offset by active anthropogenic removals of CO₂, hence the early inclusion of afforestation and reforestation and other land-based activities as carbon sinks in the United Nations Framework Convention on Climate Change (UNFCCC) negotiations, the heated political and academic debates on reducing emissions from deforestation, forest degradation, sustainable forest management, enhancement of sinks and conservation (known as REDD+), and the branding of forestry as part of the broader promise of “natural climate solutions” few years ago [4]. As a result, many countries are planning large-scale land-based CO₂ removals to meet their climate targets, including an aggregate area of afforestation and tree planting that would amount to almost one-third of the current permanent global cropland area, showing an extensive reliance on reforestation in climate mitigation pledges [5].

Are the expectations realistic?

The most recent global carbon budget suggests that the increasing trend of the land sink may have slowed and that it has declined each decade since the 1990s, particularly in the last decade (-20%) [2]. This could simply be natural variability, but it could also signal the beginning of a “saturation point” in the Earth’s system. The declining trends in global GPP indicate that vegetation GPP may stagnate in the future, resulting in the saturation of land carbon sequestration [6]. This independent evidence suggests that a natural solution based on terrestrial ecosystems as carbon sinks, which may reach saturation, will not be sufficient and that greater efforts to reduce anthropogenic CO₂ emissions will be needed in the future if we are to achieve a climate-neutral goal. Extreme climate events can also affect sink capacity, as seen in 2023, where land regions exposed to extreme heat in 2023 contributed a gross carbon loss of 1.73 GtC yr-1, indicating that the record warming in 2023 had a strong negative impact on the ability of terrestrial ecosystems to mitigate climate change [7].

It is evident that ecosystems are undergoing major transformations due to anthropogenic activities that have led to drastic changes in land [8], weather [9], and climate [2]. In recent decades, the scientific community has failed to fully characterize the complexity behind these changes and the contributions of land to observed climate change and its interactions and feedback with atmospheric processes. More recently, the climate change modelling community has recognized that these effects are difficult to quantify specifically because observations and models do not represent the relevant processes in sufficient detail to capture all potential feedback [10].

Is there a risk of a lack of common understanding among different knowledge communities?

Two risks associated with the role of carbon sinks in the net zero approach can be identified [11]: First, Earth system feedbacks such as carbon release from thawing permafrost, drying of some wetlands, or increased wildfire activity [12] could affect the net size of biosphere carbon sinks and weaken passive uptake; The second arises from policy choices, as systems based on NGHGIs reported under the UNFCCC classify passive uptake that occurs on “managed land” as anthropogenic greenhouse gas removal while the modelling community does not [13]. IPCC developed a practical approach to estimate anthropogenic emissions and removals in the national greenhouse gas inventories to allow countries to report anthropogenic land-based fluxes regularly in a consistent manner, the so-called “managed land proxy” that differs from how scientific models estimate it [14]. The definitions used by the two communities serve different purposes, reflecting the data and methodologies used by each community. The policy implications are profound. Global models are used to determine emissions pathways consistent with the goals of the Paris Agreement, such as limiting warming to “well below 2°C.” National greenhouse gas inventories assess compliance with the Paris Agreement by tracking each country’s progress toward its climate commitments and collective progress through the Global Stocktake, toward the Agreement’s goals. In the medium to long term, ensuring full methodological consistency between the IPCC emission pathways and the mitigation processes of the Paris Agreement will require the gradual integration of the global modelling, Earth observation, and national inventory communities and their estimates. Several improvements are expected from the global modelling community, including improved accessibility to methods and data, more consistent estimates of direct, indirect, and natural effects, improved representation of forest demographics, and more accurate protocols for translation into the inventory approach. This integration will increase confidence in land-use CO2 flux estimates and enable a more accurate assessment of alternative land-based policy options, which will become even more critical as we move toward the second GST to properly assess the role of land in meeting the goals of the Paris Agreement.

Reflections on the role of land-based carbon sink contributions to the Paris Agreement

Overreliance on models, which often oversimplify the complexity of Earth systems and their interconnections, leads to the failure to predict some observed impacts of climate change. Therefore, to propose land-based contributions to mitigation pathways that are difficult to implement and are unlikely to consider the impacts and feedbacks of climate change in land systems could undermine needs to achieve net zero. There is a need to strengthen observational systems and process understanding.

Overall, the uncertainties surrounding natural carbon sinks pose a major concern. Scientists and policymakers need to be alert to potential problems: plans to avoid exceeding the Paris Agreement limits on global warming rely on current model projections of sink capacity. If the models overestimate sinking, the true carbon budget could be even smaller, and current pathways to net-zero emissions would be insufficient to meet the targets. To reduce uncertainties and avoid overreliance on natural carbon, we need reliable and robust sink quantification. For example, the problem is well-illustrated with the example of fire: while fire is a key driver of change that will increase in the future, fire processes (along with other natural disturbances) are not fully incorporated in current assessments of the future carbon sequestration potential of forests. Similarly, the carbon absorption gains from re/afforestation in some regions would be largely offset by counteracting albedo effects, especially in the short term. However, forest conservation and re-afforestation are prominent in many countries as a mitigation strategy.

Land is limited; we rely on it for food, nature, biodiversity, leisure, and water storage. It cannot offset more than a portion of fossil emissions, even now, probably less in the future with worsening pressures on the biosphere, such as population increase, fires, and drought. Although natural carbon sinks cannot solve global warming alone, their protection is essential. Forest conservation, sustainable land management initiatives ensure that these ecosystems continue to absorb CO₂ and moderate climate impacts.

Acknowledgments

The author would like to thank Jim Penman, Donna Lee, Sandro Federici and Giacomo Grassi por the initial discussions on this topic held along the years.