Carbon sequestration potential of tree planting in China

Tree planting is a cost-effective nature-based solution for climate change mitigation. China, a global leader in afforestation, has implemented ambitious tree-planting programs since the 1970s, significantly increasing forest cover. However, the rate of forest expansion has slowed, necessitating intensified efforts to meet carbon neutrality goals by 2060. The success of future tree-planting projects hinges on strategic location and implementation, as past projects highlight the negative impacts of planting trees in unsuitable locations. Unsuccessful afforestation not only undermines carbon benefits but can also lead to adverse environmental effects, such as drought and soil erosion. Existing Chinese forests are often sparse, limiting carbon sequestration capacity. The remaining area suitable for planting is limited by existing croplands essential for food security. This study addresses this challenge by providing a comprehensive methodological framework to assess the potential for tree planting and carbon sequestration in China, focusing on both afforestation and densification strategies.

The literature extensively discusses the role of tree planting in carbon sequestration and climate change mitigation (Lewis et al., 2019; Griscom et al., 2017). China's efforts in afforestation are well-documented (Zhang et al., 2000; Wang et al., 2020; FAO, 2020; National Forestry and Grassland Administration, 2019; Yue et al., 2021; Lu et al., 2018), showcasing significant increases in forest cover but also revealing challenges in sustainability and effectiveness (Abbasi et al., 2023). Studies on unsuccessful tree-planting projects in China, particularly in arid and semi-arid regions (Cao et al., 2011; Cao, 2011; Xu, 2011; Cao, 2008; Li et al., 2021; National Forestry and Grassland Administration, 2017; Zhang & Song, 2006; Cao et al., 2009), highlight the importance of site suitability assessment. The limitations of planted forests compared to natural forests in terms of carbon sequestration and water consumption are also well-established (Yu et al., 2019; Feng et al., 2016; Lu et al., 2018). Prior research indicates the need for rigorous quantitative studies to evaluate the optimal location and methods for tree planting, considering economic costs and potential carbon benefits (Ahrends et al., 2017; Crouzeilles et al., 2017).

This study employs a comprehensive machine learning framework to assess China's tree-planting potential for carbon sequestration. It integrates diverse environmental variables (climate, soil, topography) with historical tree cover data (Hansen Global Forest Change dataset) to generate tree growth suitability (TGS) scores. A neural network model (two hidden layers, 64 and 32 neurons, trained using stochastic gradient descent) classifies areas as suitable or unsuitable for tree growth, with the confidence level representing the TGS score. The framework incorporates tree density data to analyze the relationship between TGS scores and tree density, differentiating between afforestation (planting in areas without trees) and densification (increasing density in existing forests). Three scenarios (25%, 50%, 75% quantiles of tree density) are evaluated for both afforestation and densification. The relationship between tree density and aboveground/belowground biomass carbon stocks is established using pixel-scale data to estimate carbon sequestration potential. The national zoning program of ecological functions (NZPEF) is incorporated to create a realistic tree planting scenario, considering constraints like water conservation and avoidance of carbon stock reduction. Economic costs are estimated based on historical data from previous ecological restoration projects in China. The robustness of the method is validated using different datasets for forest biomass carbon.

The spatial pattern of TGS scores aligns with existing tree density, decreasing from southeast to northwest. Higher TGS scores generally correspond to higher tree densities, but a single TGS score can support various densities, indicating afforestation and densification potential. The relationships between TGS scores and tree densities differ across four forest types (deciduous broadleaf, deciduous needleleaf, evergreen broadleaf, evergreen needleleaf). Under the ideal scenarios (25%, 50%, 75% quantiles), afforestation could plant 10.1, 17.2, and 25.2 billion trees, respectively, while densification could plant 7.8, 21.1, and 37.5 billion trees. Forest stock would increase substantially under these scenarios. Afforestation and densification show varying impacts on aboveground and belowground forest biomass carbon stocks, with densification showing a greater potential increase. In the 75% quantile scenario, afforestation could sequester 1866.9 ± 110.1 TgC (aboveground) and 576.8 ± 125.7 TgC (belowground), while densification could sequester 3604.6 ± 286.4 TgC (aboveground) and 1029.8 ± 43.0 TgC (belowground). A realistic scenario, constrained by the NZPEF, estimates planting 18.3 billion trees (3.1 ± 0.4 billion m³ increase in forest stock), with densification (26.4 billion trees, 6.7 ± 0.6 billion m³ increase) exceeding afforestation. In the realistic scenario, the total carbon sink potential is approximately 5.9 ± 0.5 PgC, double China's 2020 industrial CO2 emissions. The estimated cost per tonne of CO2 is significantly lower than in past projects, especially in regions with favorable conditions. The study also explores the ambiguity in defining 'densification' and 'afforestation' based on tree cover thresholds and shows that densification consistently outperforms afforestation, except when using a high threshold (above 30%).

The findings highlight the significant potential of tree densification for achieving China's carbon neutrality goals, particularly in areas with favorable environmental conditions. Densification, compared to afforestation, offers greater carbon sequestration potential while preserving land use for agriculture and other purposes. The realistic scenario, incorporating policy constraints, demonstrates the feasibility of large-scale tree planting as a climate mitigation strategy in China. However, it's crucial to acknowledge the limitations of the 1-km spatial resolution, which may not capture local variability in climate and tree distribution. Potential ecological risks associated with increased tree density (e.g., wildfire vulnerability) require further investigation. The economic assessments should consider dynamic factors like inflation and fluctuating costs. Despite these limitations, the study provides valuable insights into the large-scale potential for carbon sequestration through afforestation and densification in China.

This study demonstrates the substantial potential of both afforestation and, especially, densification for enhancing carbon sequestration in China. The findings underscore the importance of strategic planning that considers ecological and economic factors. Future research should focus on incorporating higher-resolution data and dynamic modeling to improve the accuracy of predictions, as well as integrating risk assessment models to better understand potential ecological consequences. These findings will aid in the development of more informed and effective forest management policies and practices to achieve China's carbon neutrality goals.

The study's limitations include the 1-km spatial resolution, which may not adequately capture microclimate heterogeneity and precise tree distribution. The analysis does not account for ecological risks associated with increased tree density, such as wildfire vulnerability or changes to biodiversity. Dynamic economic factors, including inflation and cost fluctuations, are not fully considered. The methodology relies on statistical relationships between tree density and carbon sequestration, potentially overlooking factors like species diversity and tree maturity.