Tropical forest restoration under future climate change

Tropical forest restoration is a promising method for rapidly removing atmospheric CO2. However, the long-term effectiveness of this strategy is threatened by future climate change, which could lead to increased heat, drought, wildfires, and insect outbreaks, jeopardizing the permanence of stored carbon. Current estimates of tropical forest carbon sequestration potential often neglect these future climate impacts and rising CO2 concentrations. Previous research primarily focuses on present-day climates, or merely scales carbon uptake based on temperature and CO2, failing to account for the complex interplay between these factors and the risk of disturbances. This study addresses this critical gap by assessing the combined impacts of future climate change, atmospheric CO2 concentrations, and lightning-caused wildfires on the success and permanence of tropical forest restoration efforts. The successful restoration of tropical forests offers not only a climate mitigation benefit—a carbon benefit less attenuated by climate responses to increased tree cover (albedo warming) compared to high latitudes—but also substantial biodiversity benefits and positive impacts for local communities. The economic and social consequences of failed restoration projects are significant, including potential habitat loss, financial losses, and the release of previously sequestered carbon, particularly if the project was involved in an emission offsetting scheme. The resilience of old-growth humid tropical forests to disturbance contrasts sharply with the greater vulnerability of seasonally arid tropical forests, which face increased risks with decreased turnover times due to disturbances. A recent qualitative assessment highlights the projected future increase in fire and drought across various tropical regions, emphasizing the urgent need for a quantitative investigation of these risks.

The literature review section highlights the existing research on tropical forest restoration, carbon sequestration, and the impact of climate change on these processes. Studies mentioned in this paper include those by Griscom et al. (2017, 2020), Busch et al. (2019), Cook-Patton et al. (2020), Bastin et al. (2020), and Lewis et al. (2019), among others. These studies provide various estimates of the potential for carbon sequestration through forest restoration, but many lack consideration of the influence of future climate change and other factors. The review also underscores previous work indicating the sensitivity of seasonally dry tropical forests to increased disturbances like fire and drought (Pugh et al., 2019). The authors point to the need for a more comprehensive model that incorporates the interplay of climate, CO2, and vegetation productivity and disturbance to accurately assess the long-term effectiveness of restoration efforts. The lack of such a comprehensive analysis serves as the primary motivation for the current research.

This study uses the LPJ-LMfire dynamic global vegetation model (DGVM) to simulate above- and below-ground biomass accumulation from restoring non-crop-producing agricultural lands. The model is driven by a range of future climate scenarios from 13 CMIP6 general circulation models (GCMs) representing the Shared Socioeconomic Pathways (SSPs): SSP1-26, SSP2-45, SSP3-70, and SSP5-85. Two CO2 experiments are conducted: one with CO2 concentrations following the SSP trajectory (CO2free), and another where CO2 concentrations are fixed at 2014 levels (CO22014). These experiments allow for the isolation of climate change effects from CO2 fertilization effects. The impact of wildfires, initiated by lightning, is also included and subsequently assessed by comparing results with simulations where wildfire effects are absent. The study incorporates a published restoration opportunity map to identify potential restoration areas. Given the unlikelihood of restoring the entire potential area, scenarios were modeled restoring only half the available area, prioritizing either high carbon uptake potential, low opportunity cost (waived agricultural revenue), or a combination of both. Additionally, a climate change resilience approach is implemented, where site selection is based on 2100 carbon uptake potential. The model's performance is evaluated by comparing it against observations and an ensemble of other DGVMs from TRENDYv9. The authors acknowledge that the use of a single DGVM (LPJ-LMfire) limits the study, and future research should consider a multi-model approach to capture the variability among different DGVMs.

Restoring 11% (128 Mha) of tropical non-crop-producing agricultural areas to forest resulted in a cumulative carbon uptake of 24.1–39.6 Pg C between 2020 and 2100. The default case (no climate change, CO2 fertilization, or wildfire) showed 28.5 PgC accumulation by 2100. Carbon accumulation increased under all climate change scenarios, regardless of CO2 fertilization inclusion, with continued accumulation until 2100 in 92% of the simulations. The impact of CO2 fertilization was most pronounced under high-CO2 scenarios (SSP5-85), while climate change had a larger effect under low-CO2 scenarios (SSP1-26). Wildfires contributed significantly to carbon loss, particularly under high-CO2 scenarios. Restricting restoration to only half the potential area (64 Mha), the prioritization strategy affected carbon storage. Prioritizing high carbon uptake potential yielded the highest storage (68.9% of full restoration), while minimizing opportunity cost yielded the lowest (56.4%). Including climate change impacts in prioritization increased carbon storage across all strategies. A 'Restoration Opportunity Index', combining climate change, wildfire threats, and opportunity cost, identified regions with high cost-effective, long-term (2030-2100) carbon storage potential, mainly in northwestern South America, West and Central Africa, the maritime continent, and parts of Southeast Asia. The default case estimate of above- and below-ground carbon uptake over 2020-2050 is at the higher end of other independent estimates, with simulated uptake rates higher in northern South America and south China and lower in southeastern Brazil compared to observation-based estimates. The prioritization strategy increased carbon storage by 18%, and accounting for climate change impacts in prioritization mitigated only up to 5% of the carbon storage reductions due to climate change.

The findings demonstrate that even under future climate change scenarios, tropical forest restoration offers significant carbon sequestration potential. While climate change impacts and wildfires reduce this potential, restoration remains a valuable climate mitigation strategy. The model results suggest that the benefits of CO2 fertilization can outweigh negative climate impacts, particularly in marginal lands. The study highlights the importance of strategic prioritization to maximize carbon storage and minimize opportunity costs. The Restoration Opportunity Index provides a valuable tool for identifying suitable locations for future restoration efforts. The higher-than-observed carbon accumulation rates in some regions highlight a limitation of the study and emphasizes the need for further model validation and refinement. The exclusion of anthropogenic fires and biogeophysical feedbacks further limits the study, but these represent conservative estimations of potential carbon sequestration capacity. This study’s finding demonstrates the potential contribution of tropical forest restoration to climate change mitigation alongside other strategies.

Tropical forest restoration shows potential for substantial and largely permanent carbon sequestration, even considering future climate change. While climate impacts reduce this potential, strategic site selection and best-practice restoration can mitigate these effects. Future research should focus on multi-model approaches, incorporating biogeophysical feedbacks, and assessing the combined impacts on biodiversity and socioeconomic factors to enhance the accuracy and generalizability of the findings.

The study's primary limitation is the use of a single DGVM (LPJ-LMfire). This limits the representation of the diversity of responses to climate change and CO2 among different models. The exclusion of anthropogenic fires potentially underestimates fire impacts, and the lack of coupling between the land cover and atmospheric models neglects climate feedbacks. The assumption of successful recruitment in restoration projects and the omission of specific management practices aimed at increasing resilience to climate change (e.g., planting drought-tolerant species) may lead to conservative estimates of carbon uptake potential. The study also does not consider the potential decline of carbon sinks in intact tropical forests under future climate change, which may counteract some of the gains from restoration. Finally, the analysis focuses on carbon uptake, not considering impacts on biodiversity and other factors.