Afforestation on Abandoned Croplands in China Has the Potential to Increase Carbon Sequestration by half

Afforestation on Abandoned Croplands in China Has the Potential to Increase Carbon Sequestration by half | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Afforestation on Abandoned Croplands in China Has the Potential to Increase Carbon Sequestration by half Le Yu, Tao Liu, Ying Tu, Xin Chen, Zhenrong Du, Hui Wu, Shijun Zheng, and 5 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6119575/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Afforestation of abandoned cropland represents a promising strategy for land-based climate change mitigation, particularly in regions where land resources for additional afforestation are limited. However, the carbon sequestration potential of such land remains largely unknown. Here, we assess the spatial distribution of abandoned cropland in China and its carbon sequestration potential through afforestation incentives, using 10,818 carbon empirical data derived from 298 peer-reviewed articles, multisource remote sensing data, and machine learning models. We identify 6.03 Mha of abandoned cropland in China that have been undergoing natural regeneration since the early 21st century. This land has the potential to sequester an additional 215.12–218.94 Tg of biomass carbon and 15.87–17.64 Tg of soil organic carbon (SOC) through afforestation by 2060, representing a 51.95–53.94% increase compared to natural regeneration alone. Our results further show that the carbon benefits from afforestation could offset 47.71–49.57% of government investments (approximately USD 16.254 billion) in abandoned cropland. Our findings highlight the significant potential of afforestation on abandoned cropland to support China’s carbon neutrality goals, while also offering a cost-benefit framework to guide land policy decisions. Earth and environmental sciences/Ecology/Ecosystem services Earth and environmental sciences/Environmental sciences/Environmental impact Earth and environmental sciences/Ecology/Ecosystem ecology Earth and environmental sciences/Ecology/Restoration ecology Abandoned cropland afforestation carbon sequestration investment and carbon benefits China Figures Figure 1 Figure 2 Figure 3 Figure 4 Main Afforestation is widely recognized as one of the most cost-effective nature-based solutions for climate change mitigation, capturing atmospheric carbon while simultaneously contributing to human well-being 1 , 2 . Since the early 21st century, China has implemented a series of afforestation initiatives aimed at enhance ecosystem services, such as soil and water conservation, sand control, carbon sequestration, and biodiversity 3 – 5 . These efforts expanded China’s forest cover by 42.98 Mha (forest cover rate increased from 18.21–23.04%) from 2000 to 2020 in China, sequestrated extra woody biomass carbon of 208.6 ± 51.8 TgC yr − 1 6 . China’s Grain for Green program, one of the largest ongoing restoration initiatives, has delivered substantial ecological and socio-economic benefits, with its impacts assessed as being 4.5 times the government’s investment 7 , 8 . The carbon sequestration potential of this ambitious project has been capable of offsetting 3–5% of China’s annual carbon emissions 3 . However, the pace of forest expansion has slowed in recent years, and the area of land suitable for further afforestation has become increasingly saturated 9 . This presenting a substantial challenge to achieving ambitious China’s carbon neutrality goal by 2060 and the mid-term target of 30% forest cover by 2035. Abandoned cropland with natural regeneration has been proven globally to have significant carbon sequestration potential 10 – 15 . Cropland abandonment is primarily driven by a combination of socio-economic and environmental factors, such as the potential hazards of climate change, disasters, and the low productivity efficiency of marginalized cropland 16 – 18 . The ecological processes of cropland abandonment are typically accompanied by spontaneous vegetation recovery 19 , 20 . The process of abandonment typically requires considerable time to reach a climax community, a process that takes even longer in high-latitude regions 3 . In arid and semi-arid regions, cropland abandonment often leads to a decline in surface SOC during the early stages, and this loss can take a long time to recover 21 , 22 . The lack of vegetation recovery control on abandoned cropland may pose a risk of accelerated decomposition of fast-growing plants (e.g., straw, litter, and residues), potentially leading to increased CO₂ emissions 23 , 24 . Recent studies and land management projects emphasize that alternative methods, such as afforestation, rewilding, and bioenergy crop cultivation, can effectively enhance the carbon capture and storage potential of abandoned cropland 11 , 13 , 25 – 27 . These strategies can establish a closed “biochar-soil” system that continuously removes CO₂ and stores it in soil and plants 28 . By optimizing management practices, abandoned cropland can simultaneously restore soil fertility 23 , promote biodiversity 29 , and enhance other ecosystem services 30 . Afforestation on abandoned cropland provides new insights into close China’s ambitious afforestation and carbon-neutral targets 14 . However, the role of afforestation of abandoned cropland in climate change mitigation remains unclear, as the potential ceiling for carbon sequestration through afforestation, compared to natural regeneration, remains be further confirmed. Understanding spatial prioritization strategies for abandoned cropland restoration, based on carbon sequestration potential, is critical for helping land-managers and stakeholders identify the most effective solutions 11 . Furthermore, China’s government guidelines for abandoned cropland emphasize flexible strategies, such as encouraging farmers to plant trees; however, there is currently a lack of actionable policies to guide and implement restoration efforts 31 . This gap is largely due to the absence of robust scientific assessments regarding the investments against potential benefits. Overall, there is an urgent need for comprehensive quantitative studies to assess the carbon sequestration potential and cost-benefit trade-offs associated with afforestation on abandoned cropland. Here, we evaluate the carbon sequestration potential of biomass carbon (above- and below-ground biomass) and SOC (0–30 cm) through afforestation and natural regeneration scenarios on abandoned cropland across China at a 30m resolution, using 10,818 empirical carbon data (3502 SOC and 7316 biomass carbon data) derived from 298 peer-reviewed articles, combined with multisource remote sensing data, machine learning, and scenario modeling methods (see Methods). Additionally, we assess the government investment on afforestation of abandoned cropland and the potential benefits from the carbon market by 2060 to estimate the cost-benefits of our proposed incentives. Our study try to addresses three key research questions: (1) How much abandoned cropland in China has undergone natural regeneration since the 21st century? (2) What is the carbon sequestration potential by 2060 under scenarios of abandoned cropland either through natural regeneration or afforestation? (3) Considering funding costs and potential carbon benefits, what is the feasibility of afforestation of abandoned cropland? Collectively, our analysis provides a timely and comprehensive estimate of the carbon sequestration potential of abandoned cropland in China, offering valuable insights for policymakers to support the carbon neutrality target by 2060, while considering both investment and return factors. Results Abandoned cropland patterns in China Based on land cover change trajectories and vegetation change trends (see Methods), we generated three sets of 30m resolution maps of abandoned cropland using China’s Land Cover Dataset (CLCD), China’s Land-Use/cover Datasets-A (CLUD-A), and China’s Land Cover Dataset (CACD), respectively. Following established methods for assessing land cover change mapping accuracy, we report the accuracy assessment parameters for abandoned cropland (Fig. S1). The results show that the abandoned cropland map generated from the CACD exhibited the highest accuracy, with an overall accuracy of 82.17%, user accuracy of 86.57%, producer accuracy of 76.16%, and an F1 score of 0.811 (Fig. S1). Compared with other coarser-resolution maps of abandoned cropland in China 14,32,33 , our abandoned cropland product shows superior performance, particularly in the agro-pastoral transition zones and the karst regions of southwest China. We identify that 17.86 Mha of cropland has been converted to natural vegetation from 2000 to 2017 (Fig. 1a), with 66.23% of this area allocated for afforestation (e.g., the Green for Grain Program) (Fig. 1d). A total of 6.03 Mha (33.77%) of cropland has been abandoned with natural regeneration (Fig. 1c), corresponding to an annual abandonment rate of 1.60±0.13% of active cropland (Fig. 1b), which is equivalent to 1.7 times the area of Hainan province. This finding confirms the widespread occurrence of cropland abandonment in China, particularly in farming-pastoral ecotone of northern China (1.67 Mha, 27.77%), the depopulated areas of Northeastern China (0.70 Mha, 11.64%), the Karst regions of Southwestern China (1.06 Mha, 17.62%), and the Southern hilly areas (0.99 Mha, 16.37%). This phenomenon is influenced by a combination of factors, including climate change, natural disasters, limitations in agricultural infrastructure, and agricultural labor shortage 34 . According to the classification of mountainous and hilly counties defined by the Chinese government 35 , we find that 70.63% of the abandoned cropland occurs in these areas (Fig. 1e). Furthermore, by overlaying the China’s terraced map 36 , we further find that 5.06±1.21% of the terraced land has undergo abandonment (Fig. 1f). Carbon accumulation after cropland abandonment We use literature-derived data to assess the carbon accumulation over a 70-year period under two abandonment (natural regeneration) and afforestation (Fig. 2a–b) strategies. Linear regression analysis reveals differences in carbon accumulation between the two strategies (Fig. 2a–b). Specifically, compared with SOC, carbon accumulation in afforested cropland is primarily driven by increase in biomass carbon (Fig. 2b). The carbon accumulation rates in biomass carbon differ significantly, with Slopes of 3.2132 Mg C ha -1 yr -1 for afforestation and 0.0835 Mg C ha -1 yr -1 for abandonment. In contrast, the impacts of afforestation ( Slope = 0.087 Mg C ha -1 yr -1 ) and abandonment ( Slope = 0.021 Mg C ha -1 yr -1 ) on SOC are more limited (Fig. 2a), underscoring the well-known challenges in generating SOC, such as carbon instability driven by microbial activity 37,38 . We generate a binary map of the carbon accumulation rates for abandoned croplands in China (Fig. 2c), combining several covariates (Table S1) and machine learning models (see Methods). By the end of 2023, China’s naturally regenerated abandoned land has sequestered 250.01 Tg of biomass carbon and 14.47 Tg of SOC (Fig. 2c). Considering abandonment duration (Fig. S2), we find that the average carbon accumulation rates for biomass carbon and SOC in abandoned croplands are 1.783 Mg C ha⁻¹ yr⁻¹ and 0.285 Mg C ha⁻¹ yr⁻¹, respectively (Fig. 2c). The spatial results show that in arid/semi-arid regions, such as the Loess Plateau, Inner Mongolia grasslands, and Xinjiang, natural regeneration of post-agricultural vegetation leads to a more pronounced increase in SOC than in biomass carbon during natural regeneration (Fig. 2c). In the Northeastern regions, SOC is more susceptible to long-term decline risks, during the early stages of cropland abandonment, being particularly vulnerable to the influences of climate change, soil physicochemical properties, and shifts in microbial community structure 21 . We further find that the biomass carbon sequestration in the first decade after abandonment (23.05 Mg C ha⁻¹) was higher than in the second decade (12.37 Mg C ha⁻¹) (Fig. 2d). In contrast, the SOC accumulation was more pronounced in the second decade (3.27 Mg C ha⁻¹), about 1.4 times higher than first decade (Fig. 2d). Across different geographic regions, the carbon sequestration on abandoned cropland was most significant in the South-central (2.80 Mg C ha⁻¹ yr⁻¹) (Fig. 2j) and Southwest regions (3.35 Mg C ha⁻¹ yr⁻¹) (Fig. 2i). We also observed SOC loss (−0.34 Mg C ha⁻¹) in the Northeast region during the first five years of cropland abandonment (Fig. 2e). These results indicate significant spatial variations in SOC and biomass carbon changes due to cropland abandonment, particularly in areas experiencing SOC loss. Carbon sequestration potential on abandoned lands by 2060 Based on applicable afforestation strategies, including species selection and planting density (see Methods), our results show that afforestation on abandoned cropland in China could result in a total carbon sequestration (including biomass carbon and SOC from 0–30 cm) of 900.73 – 926.83 Tg C by 2060, with 301.49 – 302.55 Tg of SOC and 598.17 – 624.07 Tg of biomass carbon (Fig. 3a–b). Afforestation sequesters 215.12 – 218.94 Tg more biomass carbon and 15.87 – 17.12 Tg more SOC than the natural regeneration of abandoned cropland, resulting in a 51.95 – 53.94% increase in carbon sequestration (Fig. 3a–b). This result demonstrates the significant carbon sequestration potential of afforestation on abandoned cropland. During the first decade of afforestation, the carbon accumulation rate in SOC (0.182 – 0.193 Mg C ha −1 yr −1 ) is lower than that of natural regeneration (0.251 – 0.297 Mg C ha −1 yr −1 ), with this trend being more pronounced in China’s arid and semi-arid regions (Fig. 3a and 3c). However, this trend reverses after 2040, with afforestation’s SOC accumulation rate surpassing that of abandoned cropland (afforestation: 0.264 – 0.272 Mg C ha −1 yr −1 ; abandoned cropland: 0.109 – 0.148 Mg C ha −1 yr −1 ) (Fig. 3a), driven by the accumulation of leaf litter, humus, and root decomposition from woody plants 38 . In contrast, the gap in biomass carbon accumulation rate between afforestation and cropland abandonment widens over time, with afforestation sequestering 1.35 – 1.51 Mg C ha −1 yr −1 (Fig. 3b). As trees approach maturity, the growth rate stabilizes and even slows, causing biomass carbon accumulation rate approaching saturation (0.363 – 0.755 Mg C ha −1 yr −1 ) in afforested areas (Fig. 3b). Compared to cropland abandonment, the afforestation scenario will sequester additional substantial SOC in Central and Southern China, with net increases ranging from 7.54 – 8.55 Tg C by 2060 (Fig. 3c). The Southwest region also shows notable SOC sequestration gains, with net increases ranging from 7.11 – 7.73 Tg C, suggesting that prioritizing afforestation in these areas would yield greater SOC sequestration benefits (Fig. 3c). In contrast, in northern China, natural regeneration may be more advantageous for SOC sequestration approximately 32.38% (Fig. 3c). This may be the rapid growth of herbaceous vegetation promotes microbial activity, thereby accelerating SOC accumulation 40 . In term of biomass carbon, the southern regions (including East China, Southwest China, and Central South China) will sequester additional biomass carbon with 143.30 – 143.93 Tg C (20.58 – 21.76 Mg C ha⁻¹), which is 2.95 – 3.16 times higher than in the northern regions (Northeast China, North China, and Northwest China) by 2060 (Fig. 3d). Additionally, the results highlight local hotspots of biomass carbon gain potential in the Northeast plain, which may be attributed to region-specific factors such as precipitation, soil nutrients, and physicochemical properties (Fig. 3d) 41,42 . Investment and carbon benefits Based on the subsidy schemes of China’s Grain for Green program, and regional afforestation invest list (see Methods), we estimate that implementing afforestation on abandoned cropland will require an investment of approximately $16.254 billion, which includes about $11.378 billion for land subsidies and $4.876 billion for seedling subsidies (Fig. 4). This amount represents 22.3% of the total investment in the two phases of China’s Grain for Green program (2000 – 2020) 8 . By 2060, the carbon sequestration benefits of afforestation will reach USD 7.75 – 8.06 billion (Fig. 4). This includes revenues of USD 0.74 – 0.75 billion from SOC and USD 7.01 – 7.31 billion from biomass carbon. These benefits are an additional USD 2.706 – 2.755 billion compared to natural regeneration of abandoned cropland (Fig. 4). This is a conservative estimate, as China’s carbon equivalent value is expected to rise in the future 43 . Nevertheless, our results show that the carbon benefits from transforming abandoned cropland can cover 47.71% to 49.57% of the investment (Fig. 4). In comparison, the national report on the comprehensive benefits of the Grain for Green program from 2000 to 2020 indicates that the carbon benefits of the program can cover 43.10% of the total investment, which includes USD 72.873 billion in funding and USD 31.411 billion in carbon sequestration return 8 . We present a theoretical framework for implementing afforestation on abandoned cropland, which not only highlights the cascading effects of government funding inputs and carbon benefits but also emphasizes the spillover effects throughout the process (Fig. 4, blue background). Specifically, we focus on the ecosystem service functions provided by afforestation, such as water conservation and environmental purification (Fig. 4). Reports on the Grain for Green program indicate that the ecosystem service values of these two services are 2.08 and 1.39 times the carbon sequestration value, respectively 8 . Based on the ecological benefits ratio from the Grain for Green program, the ecosystem benefits of afforestation on abandoned cropland could fully offset the government’s financial investment (Fig. 4). From a socio-economic perspective, government investment in afforestation significantly enhances poverty alleviation outcomes, mitigates urban-rural inequalities, and provides employment opportunities for local farmers (Fig. 4). Afforestation on abandoned cropland can also serve as an ecological engineering approach with poverty alleviation potential, directly or indirectly contributing to the achievement of the SDGs, particularly SDGs: 1, 2, and 3, while enhancing the coupling between people and the environment (Fig. 4). Discussion Identifying the spatiotemporal distribution of abandoned cropland is a critical prerequisite for unlocking additional abandoned land resources 11 . Our results show that the annual abandonment rate in China is 1.60±0.13%, and terraced cropland are abandoned at a rate 2.62 – 3.62 times higher than the average (Fig. 1). We surprisingly reveal that 82.89% of abandoned cropland patches are smaller than 1 ha, which highlights the high exposure of widely environmentally disadvantaged cropland, such as fragmented and steep-slope cropland, to abandonment in China. Those posing significant potential challenges for sustainable land resource use 44–46 . More importantly, there is no consensus on the established methods for investigating abandoned cropland in China, leading to significant variation in abandonment rates (see Table S4). Some studies also classify land converted to artificial forests, orchards, and afforestation as abandoned cropland 33,47 , leading to an overestimated cropland annual abandonment rate in China 48 . Therefore, the development of coherent and standardized procedures are necessary to conduct top-down surveys of abandoned cropland and ensure the effective reuse of abandoned land resources in China. Since 1980, forest expansion has been the dominant driver of the China’s land-based carbon sequestration 49 . The saturation of forest growth and competition for land from agricultural production present significant challenges to these carbon sequestration services 50 . Given the limitations of availability of additional land for afforestation in the future 51 , if strategic afforestation is implemented on abandoned cropland, it could make a considerable potential contribution to China’s climate goals. This contribution primarily comes from the Southern regions and biomass carbon 52 , with its carbon sequestration potential being 2.95 – 3.16 times greater than that of the Northern regions (Fig. 3). Afforestation in the Northern China has a relatively weaker contribution to climate change mitigation 40 . In drier ecosystems, such as alpine meadows and high-altitude grasslands, plants tend to allocate more biomass underground, thereby promoting the long-term accumulation of SOC 53 . This provides valuable and reliable spatial insights for implementing large-scale afforestation efforts on abandoned cropland. To fully harness the potential of abandoned cropland, further examination is needed of these alternative uses, particularly with regard to their ecological impacts and effects on the atmospheric coupling of ecosystems, such as growing bioenergy crops 54,55 , grassland restoration, and rewilding 25 . Afforestation can also provide a range of co-benefits and ecosystem services, including water filtration, air purification, and soil quality improvement 56,57 . For example, globally, abandoned cropland can serve as habitats for a majority of bird species (62.7%) and mammal species (77.7%) 29 . In addition, transforming abandoned cropland and unlocking its potential for diverse uses, such as landscape conservation, flood prevention, urban cooling, and cultural ecosystem services, is receiving increasing attention 58–60 . The release of abandoned cropland resources should consider the synergies and trade-offs among multiple functions, maximizing its potential to enhance environmental quality and human well-being. Our calculations show that the cost of transforming abandoned cropland for carbon sequestration is approximately $23.62 – 24.42 per ton of CO₂, which is slightly higher than the average carbon sequestration cost of previous ecological projects in China (approximately $21.18 per ton of CO₂) 9 (Fig. 4). As a carbon return, by 2060, without accounting for inflation, the carbon benefits from afforestation schemes are projected to offset 47.71 – 49.57% of government funding inputs (Fig. 4). This estimate does not include additional ecological benefits, such as water conservation, air purification, and biodiversity, nor the broader socio-economic benefits 8 . Moreover, evidence from farmer surveys indicates that the project could improve rural livelihoods in China by increasing income, providing employment opportunities, and enhancing residents’ well-being 61,62 . These robust findings suggest that afforestation on abandoned land could generate significant environmental and human welfare benefits. Afforestation on abandoned cropland has the potential to unlock significant carbon sequestration, but further efforts are needed to translate this potential into tangible benefits 11 . Various factors, including but not limited to labor supply, afforestation activities, and irrigation, may need improvement to enhance the afforestation adaptability of abandoned cropland. The prioritization of these factors should incorporate local knowledge and context-specific action plans 63 . A more sustainable approach may involve integrating abandoned cropland with agricultural land substitution strategies, such as land swapping or land consolidation 64,65 . Additionally, ensuring the long-term persistence of climate change mitigation potential requires cross-sectoral policies and incentives 11 . Without such efforts, afforestation risks becoming a short-term activity, thereby reducing the achievable potential of abandoned cropland 15 . Establishing and expanding reliable carbon markets is critical for ensuring cross-sectoral coordination in China and securing carbon offsets from reforesting abandoned cropland. However, there are several limitations to our findings. Identifying abandoned cropland in China involves overcoming challenges related to data accuracy, which is influenced by remote sensing, biases in identifying abandoned cropland, and variations in land policies 37 . To address these challenges, we employed cross-comparisons of multiple data products and widely accepted approaches for identifying abandoned cropland. These limitations do not diminish the utility of our research; rather, they highlight the importance of cross-scale information 11 . National-level surveys of abandoned cropland are essential for enhancing policymakers’ understanding of its potential and for informing land policy development, particularly in relation to China’s carbon neutrality and food security goals 66 . It is crucial to integrate local-scale field surveys with national survey results to ensure that knowledge of abandoned cropland is translated into actionable strategies that can also benefit local communities 11 . This process must take into account factors such as land tenure, cultural considerations, costs, and labor availability. While we have considered incorporating permanently abandoned cropland from the 21st century into afforestation efforts, a significant portion of this land remains suitable for recultivation (approximately 82.54% in China 67 ), aligning with China’s strategic food security needs through land improvement and rehabilitation 11 . Furthermore, implementing local afforestation projects on abandoned cropland must safeguard the rights of rural and indigenous communities, leveraging their knowledge to foster success, enhance resilience, and avoid negative social and ethical consequences 63 . Finally, afforestation should be carried out cautiously to prevent adverse environmental impacts. Large-scale afforestation in hotspots, particularly in arid and semi-arid regions such as the Loess Plateau and agro-pastoral transition zones, increases the demand for water resources, potentially leading to water shortages and low tree survival rates 68,69 . Our research demonstrates that unlocking the carbon potential of abandoned cropland and offsetting carbon emissions provides substantial benefits. This represents a crucial pathway for China’s dual carbon targets, the United Nations Decade on Ecosystem Restoration, the Sustainable Development Goals, and the next phase of the Grain-for-Green Program. Our simulation results offer valuable insights for policymakers and land-use planners in China, identifying key strategies to maximize the carbon sequestration potential of abandoned cropland. Methods We analyzed the spatial distribution of abandoned cropland in China, along with the carbon sequestration potential and cost-benefit considerations under afforestation scenarios. This analysis integrated empirical data, machine learning models, and scenario analysis methods. The following sections outline the datasets and methodologies employed in this study. Mapping abandoned cropland To accurately capture the fragmented and geographically marginalized abandoned cropland in China, we utilized three sets of annual land products: China’s Land-Use/cover Datasets-A 70 (CLUD-A), China’s Land Cover Dataset 71 (CLCD), and China’s Annual Cropland Dataset 72 (CACD). These are the only three available datasets providing annual land data with 30m resolution (1 arc-second) for China. The first two are annual land cover products that include various land cover types, such as cropland, forests, grasslands, and impervious surfaces, while the latter provides the annual spatiotemporal distribution of cropland from 1986 to 2021. The CACD product integrates time-series Landsat satellite images, automated training sample generation methods, machine learning algorithms, and LandTrendr techniques, resulting in high user accuracy (0.93), producer accuracy (0.79), and overall accuracy (0.79) 72 . Cropland abandonment, as defined by the FAO, refers to cropland that has not been cultivated for at least five years 73 . To mitigate the uncertainty caused by unstable cropland, we ensured that the cropland remained stable for at least five years prior to abandonment 45 . Specifically, a pixel was classified as cropland for the five years preceding abandonment and subsequently transitioned to abandoned cropland category for at least five consecutive years (excluding built-up areas and water bodies). We constructed a decade-long temporal trajectory window in Google Earth Engine, analyzing the data pixel by pixel to identify waste cropland. Pixels where cropland was converted to settlements, water bodies, or reclaimed for at least one year were excluded 11 . This approach helps prevent overestimation of uncultivated pixels due to short-term fallows and removes temporarily non-cultivated cropland, ensuring a more reliable estimate of abandoned cropland. We also did not consider abandoned cropland prior to 2000, as its reliability has been reported to be lower 11 . Importantly, China’s Grain-for-Green program, which began in 2000, plays a pivotal role in land restoration efforts on abandoned cropland 21 . Here, we adopt a stricter definition of cropland abandonment. In China, it is defined as the natural regeneration of vegetation rather than afforestation in post-agriculture period 19,20,74,75 . Vegetation growth trends offer new insights for identifying abandoned cropland, distinguishing it from afforestation efforts 20 . In this study, we constructed NDVI max datasets with 30m resolution for China’s growing season (April to October, 1995 – 2022) using the maximum value composite method applied to Landsat satellite data, with a temporal resolution of 16 days. NDVI max is widely used as a remote sensing proxy to detect changes in terrestrial vegetation activity 76 . We applied quality mask information to eliminate the influence of clouds and their shadows in the Landsat imagery, and used MOD13Q1 data to fill the gaps. Our preliminary experiments have shown that the NDVI max tends to decrease significantly in the first year of cropland abandonment, whereas for afforestation, the NDVI max shows a clear increase, indicating vegetation growth due to tree planting 19 . Based on a systematic review of methods for distinguishing between abandoned cropland and afforestation (Table S3) and multiple preliminary experiments (Fig. 1g), we propose a combined algorithm based on land cover change trajectories and vegetation index trends. Land change trajectories were used to identify the year of cropland abandonment. Subsequently, we applied a double ordinary least squares model to fit the NDVI max trends for the five years prior to abandonment (cropland) and the five years following abandonment, and compared the NDVI max performances of the abandonment year. If a significant decrease (Mutation point, shown in Fig. 1g) in NDVI max was observed in the abandonment year, we identified it as abandoned cropland. In contrast, if the NDVI max showed an increase, it was recognized as afforestation (Fig. 1g). This integrated method effectively distinguishes between abandonment and afforestation in the post-agriculture period, yielding convincing results in both subtropical regions and arid/semi-arid areas (Fig. S1). Based on the above criteria, we generate three sets of abandoned cropland datasets, with cropland converted to natural vegetation, from the CULD-A, CLCD, and CACD datasets, respectively. Notably, for the CACD dataset, we incorporated urban impervious surface data from the GALA dataset 77 and permanent surface water data 78 to ensure the consistency of the output results. We employ a sampling-based method to assess the accuracy of abandoned cropland maps derived from three land data sources (Fig. S1). To determine the sample size for validation, we applied a disproportionate stratified sampling strategy, which is well-suited for evaluating the classification accuracy of small categories, such as abandoned cropland 79 . For both abandoned and non-abandoned cropland categories, we randomly selected 1,080 samples from each category, totaling 2,160 pixels (see sample locations in Fig. S1). These reference samples were visually interpreted using very-high-resolution (VHR) historical images from Google Earth, with time-series NDVI max data from Google Earth Engine assisting in sample identification 11 . We compared the overall accuracy of the three abandoned cropland maps and ultimately selected the map with the highest accuracy, derived from the CACD dataset 72 . Assembling a carbon database and standardizing data We systematically reviewed literature from the Web of Science and China National Knowledge Infrastructure databases (28 July 2024), using the keywords: (abandoned cropland OR cropland abandonment OR land-cover change) AND (carbon sequestration OR biomass carbon OR SOC OR carbon accumulation rate) AND (China) AND (afforestation OR natural regeneration OR secondary succession OR plantation). We included “Grain-for-Green” as a keyword because this program represents a large-scale, purpose-driven transition of cropland to forests in China 21 . In term of SOC or biomass carbon changes induced by land-cover change, we considered only post-agricultural land transitions. Here, we initially retrieved 3,426 peer-reviewed studies, later expanding the total to 3,713 entries. We reviewed all abstracts to identify accessible studies that described pathways for transitioning cropland to natural vegetation, including abandonment and afforestation, and identified research that quantified soil organic carbon (SOC) or biomass carbon stocks. To avoid duplicate measurements, we prioritized original studies, reviews 12,39,80–82 and the Forest Carbon Database (ForC) 83 . For these, we accessed the original studies to verify figures, correct errors, and extract additional variables. Notably, this study focuses exclusively on vegetation growth following cropland transitions. Natural regeneration after forest fires, logging, or pest outbreaks was not considered 39 . To be included in our database, studies needed to meet the following criteria: (1) The prior land cover of the area undergoing abandonment (natural regeneration) or afforestation must have been cropland; (2) Geographic coordinates (latitude and longitude) or identifiable locations (e.g., recognizable place names) must be provided; (3) Empirical measurements of carbon (or biomass) in aboveground or belowground vegetation, litter, coarse woody debris, and/or soil pools must be reported. In addition, the database includes critical information such as the duration of abandonment (years), types of carbon pools, and estimates of carbon or biomass stocks (Mg ha⁻¹). The resulting dataset includes 7,316 empirical measurements of carbon stocks in aboveground and belowground biomass, soil, litter, and coarse woody debris. For studies reporting biomass only, we converted the values to carbon (Mg C ha⁻¹) using default conversion factors: 0.47 for aboveground and belowground carbon pools (in addition to the available and reference conversion rate from original paper), 0.37 for litter biomass, and 0.50 for coarse woody debris biomass 39 . These conversion ratios were not applied to grassland-stage biomass estimations, as research indicates a higher proportion of belowground biomass in abandoned cropland, particularly in arid and semi-arid regions. Consequently, we excluded cases where only aboveground biomass was considered for abandoned croplands. This resulted in 684 independent plot measurements for total vegetation carbon. For dead carbon pools (litter and coarse woody debris), measurements often included additional pools, but we did not attempt to disaggregate litter or coarse woody debris from these combined measurements due to their high variability and site specificity 39 . Most SOC data (79.7%; N=2,793/3,502) were originally reported in Mg C ha⁻¹ per centimeter of depth. For the remaining cases, we converted SOC from concentrations (per 1,000 g of soil) or soil organic matter (SOM). For SOM concentration data (N = 709), we used conversion factors provided in the literature whenever available; otherwise, we estimated SOC concentrations as SOM/1.724 81 . SOC concentrations were converted to Mg C ha⁻¹ per centimeter of depth using either empirically reported bulk density data (N = 735) or depth-specific bulk density data from SoilGrids 84 . SoilGrids provides bulk density models at 15 cm, 30 cm, and 60 cm depths, and we used the value closest to the depth of SOC concentration measurements. For soil data, we standardized measurements to a depth of 30 cm, as the effects of cropland abandonment on SOC are most pronounced in the surface layer 21 . For plots with multiple depth measurements, we used the slope of log-log curves fitted to cumulative SOC stocks as a function of depth to estimate SOC at the standard depth 85 . For plots without multiple depth measurements, biome-specific slope coefficients were applied 39 . For control plots, SOC stocks for cropland at the initial year of abandonment were recorded (N = 512). Throughout this analysis, we focused on the 2000 – 2060 timeframe for cropland abandonment, as this period represents a critical biophysical and policy-relevant window for achieving net-zero emissions and mitigating the most severe impacts of global warming. The database also includes other valuable information, such as tree species and soil types. Spatial prediction and modeling of carbon sequestration To develop spatial prediction models for carbon sequestration, we selected representative environmental covariate layers based on the conceptual framework of previous carbon accumulation models 39 . These layers included variables related to topography, climate, soil nutrients, soil chemistry, soil physical properties, and downscaled CO₂ and nitrogen deposition data 6,86 (Table S1). We first sampled these environmental covariate stacks at the locations of each point in the literature-derived dataset. Variables representing current vegetation conditions, such as leaf area index or forest cover, as well as satellite-derived indices like the normalized difference vegetation index (NDVI), were excluded, as they do not reflect the fundamental biophysical controls on future carbon accumulation rates in plant biomass 39 . The covariate map layers were resampled and reprojected to a uniform pixel grid at a resolution of 1 km (30 arc-seconds) using the World Geodetic System 1984 (EPSG: 4326). High-resolution data were downsampled using an average aggregation method, while low-resolution data were upsampled using a simple nearest-neighbor approach (i.e., without interpolation). This resolution was chosen to balance pixel-level uncertainty (which increases proportionally at smaller pixel sizes) with the finest resolution available for most covariates 87 . For multi-year covariate layers under non-future scenarios, we used mean values from 2000 – 2020 to capture average conditions under current and historical climates since the 21st century. Temporal information on cropland abandonment was also included as a key variable for evaluating future carbon accumulation (replacing time with space). Additionally, management practices for abandoned cropland, including abandonment and afforestation—were incorporated as categorical covariates in the models. We constructed separate models (model pipelines) for SOC and biomass carbon. After testing several models, including SVR and XGBoost, the Random Forest algorithm with high performance, was selected to estimate temporal trends in future carbon sequestration. This choice was due to its robustness in extrapolating empirical data to large-scale datasets, insensitivity to feature scales, and reduced susceptibility to overfitting 39 . Using Python’s scikit-learn package, we defined models and pipelines for both SOC and biomass carbon. The total dataset was randomly split into training (70%) and testing (30%) subsets, with separate pipelines for each. We determined the optimal number of decision trees in increments of 50 and selected the best feature selection method and machine learning algorithm with defined hyperparameters using triple cross-validation with root mean square error (RMSE) as the performance metric (Fig. S5). Cross-validation creates pseudo-training datasets to estimate out-of-sample error, mitigating overfitting to the training data 88 . In triple cross-validation, the training dataset was randomly divided into three equally sized subsets. Two subsets were combined to form a new training set, while the remaining subset served as a validation set to assess model performance. The final results indicated that the optimal number of trees for the SOC and biomass carbon random forest models was 650 and 850, respectively, achieving the best fit accuracy (Fig. S5). The SOC and biomass carbon model pipelines achieved R² values of 0.76 (RMSE = 11.00 Mg ha⁻¹) and 0.72 (RMSE = 20.97 Mg ha⁻¹), respectively, providing a robust foundation for further research (Fig. S6). Simulation Design Here, we consider two main pathways for abandoned land management: natural regeneration and afforestation. We do not account for additional scenarios that are spatially difficult to control, such as re-cultivation of fallow land or newly abandoned land, as their areas are sufficiently small in the short term to be neglected. The first pathway (scenario) involves allowing natural vegetation regeneration on abandoned land until 2060 without human intervention. In ours model, we adjusted the duration covariate to predict the future carbon sequestration potential. Note that we assume all biomass from abandoned cropland is removed initially 11 . The future changes in SOC are assessed based on the SOC density of nearby long-term cultivated lands. The second pathway (scenario) considers implementing afforestation programs on abandoned cropland. In this land management practice, we modeled afforestation efforts in alignment with the Grain-for-Green program, one of China’s major ecological restoration initiatives 5 . Specifically, the empirical samples of carbon changes for afforestation on cropland are primarily driven by this project. The spatial variability in carbon sequestration associated with these tree species can be addressed using empirical data, which includes species tailored to local conditions and indigenous knowledge, such as sea buckthorn, poplars, and walnut trees in arid and semi-arid regions 89 . To minimize estimation errors in carbon sequestration potential under afforestation scenarios, we calculated the average carbon sequestration for regions where multiple tree species were reported in the carbon database. This approach helps reduce uncertainty caused by variability among tree species. We used SOC stocks from nearby long-term croplands (N = 512) as a baseline and estimated future SOC stocks based on annual SOC change rates 39 . In the future simulation design, we incorporated climate data from the four Shared Socioeconomic Pathways (SSPs) of CMIP6 (SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5), along with CO₂ and nitrogen deposition data. Numerous studies have highlighted the critical importance of these variables for ecosystem carbon sinks 90 . We focused on carbon sequestration levels through 2060, as this year marks a key milestone in China’s commitment to carbon neutrality. Additionally, empirical data models are more robust within a 60-year timeframe. Accounting for inputs and carbon returns after afforestation in abandoned cropland According to the Chinese government’s funding allocation plan for the new round of the Grain-for-Green program in 2024, a subsidy of 1,200 RMB (USD 169.00) per mu (1 mu = 1/15 ha) is provided for reforested cropland 91 . This subsidy includes 900 RMB (USD 126.76) for land compensation and 300 RMB (USD 42.24) for sapling support, with funding from the central government. For local government subsidies, detailed afforestation data were calculated at the provincial level for two periods (2000 – 2010 and 2010 – 2020) across 21 provinces in China. This data includes afforestation area, total investment, and provincial government contributions. By integrating investments from both the central and local governments and adjusting for inflation using the World Bank’s RMB inflation rate, we estimated the per-unit area investment for afforestation at the provincial level (Table S2). For provinces without quantified investment information, we used the average values from comparable geographic regions as a reference, given the similarities in the geographic and physical factors affecting land restoration. To estimate the carbon sequestration benefits, we calculated the incremental revenue from afforestation on abandoned cropland based on the carbon trading value in China’s carbon market. Since 2021, China has established nine carbon trading markets 92 . Using daily trading data up to May 2024, including average transaction prices, trading volumes, and total amounts, we computed an average carbon trading revenue of USD 11.71 per ton. This represents a conservative estimate of carbon value, which may increase in the future 43 . In estimating the carbon sequestration of biomass and SOC under the afforestation scenario, we projected potential carbon revenue from abandoned cropland. We also referred to reports on the socio-economic and environmental benefits of the Grain-for-Green program over the past 20 years 8 . These reports provided comparative insights to support our carbon sequestration findings and other ecosystem service values, such as soil conservation and sand fixation. Uncertainty analysis To evaluate the spatial uncertainty of the model regarding biomass carbon and SOC carbon sequestration, we applied a random 70% training subset of the best-performing step length to a new ensemble of 50 random forest models, trained on all covariates and carbon empirical points. Each random forest model was trained on independent pipeline, allowing us to estimate the model uncertainty for each pixel by calculating the standard deviation of predictions calculation 39 . The mean standard deviations of model uncertainty for simulated SOC sequestration under abandonment and afforestation scenarios were 3.03 Mg C ha⁻¹ and 2.95 Mg C ha⁻¹, respectively. Similarly, the mean standard deviations of model uncertainty for biomass carbon sequestration were 4.65 Mg C ha⁻¹ and 4.7 Mg C ha⁻¹, respectively. We observed significant spatial differences in the uncertainties of biomass carbon and SOC sequestration. The highest uncertainties for biomass carbon were concentrated in the southwestern regions, while uncertainties for SOC were predominantly observed in northeastern China, particularly in the Greater Khingan Mountains region (Fig. S7). Declarations Data availability The annual CLCD 71 , CLUD-A 70 , and CACD 72 maps of China were obtained from previous studies. The derived 30m resolution abandoned cropland map for China is available via 10.6084/m9.figshare.28386476. Provincial-scale data on afforestation areas and investment (including both national and local investments) in China from 2003 to 2020 can be accessed via 10.6084/m9.figshare.28386476. The daily transaction records from China's nine carbon emission trading exchanges, including transaction prices, volumes, and price fluctuations (as of May 2024), can be freely accessed via 10.6084/m9.figshare.28386476. The raw carbon datasets for abandoned cropland are available via 10.6084/m9.figshare.28386476. Detailed descriptions of the environmental covariates can be found in Table S1. Code availability The code, including JavaScript code for identifying abandoned land and Python code for machine learning, related to the key methods of this work, is available at 10.6084/m9.figshare.28386476. Acknowledgement This work is supported by the National Key R&D Program of China (grant number: 2024YFF1307600) References Zeng, Y. et al. Economic and social constraints on reforestation for climate mitigation in Southeast Asia. Nat. Clim. Chang. 10 , 842–844 (2020). Austin, K. G. et al. The economic costs of planting, preserving, and managing the world’s forests to mitigate climate change. Nat Commun 11 , 5946 (2020). Ma, Y., Huang, L., Li, J., Cao, W. & Cai, Y. Carbon potential of China’s Grain to Green Program and its contribution to the carbon target. Resources, Conservation and Recycling 200 , 107272 (2024). Deng, L., Liu, G. & Shangguan, Z. Land‐use conversion and changing soil carbon stocks in C hina’s ‘Grain‐for‐Green’ Program: a synthesis. Global Change Biology 20 , 3544–3556 (2014). Persson, M., Moberg, J., Ostwald, M. & Xu, J. The Chinese Grain for Green Programme: Assessing the carbon sequestered via land reform. Journal of Environmental Management 126 , 142–146 (2013). Liao, Z. et al. Growing biomass carbon stock in China driven by expansion and conservation of woody areas. Nat. Geosci. 17 , 1127–1134 (2024). Deng, L. et al. Past and future carbon sequestration benefits of China’s grain for green program. Global Environmental Change 47 , 13–20 (2017). GOV. Comprehensive Benefit Assessment Report of China’s Grain for Green Project. (2022). Yao, L. et al. Carbon sequestration potential of tree planting in China. Nat Commun 15 , 8398 (2024). Schierhorn, F. et al. Post‐Soviet cropland abandonment and carbon sequestration in European Russia, Ukraine, and Belarus. Global Biogeochemical Cycles 27 , 1175–1185 (2013). Zheng, Q. et al. The neglected role of abandoned cropland in supporting both food security and climate change mitigation. Nat Commun 14 , 6083 (2023). Tian, D. et al. Cropland abandonment enhances soil inorganic nitrogen retention and carbon stock in China : A meta‐analysis. Land Degrad Dev 29 , 3898–3906 (2018). Gvein, M. H. et al. Potential of land-based climate change mitigation strategies on abandoned cropland. Commun Earth Environ 4 , 39 (2023). Li, M., Cui, Y., Dong, J. & Qin, Y. Abandoned cropland compensates the decrease in net ecosystem productivity of impervious surface expansion in China. Environmental Impact Assessment Review 104 , 107363 (2024). Crawford, C. L., Yin, H., Radeloff, V. C. & Wilcove, D. S. Rural land abandonment is too ephemeral to provide major benefits for biodiversity and climate. Sci. Adv. 8 , eabm8999 (2022). Leal Filho, W., Mandel, M., Al-Amin, A. Q., Feher, A. & Chiappetta Jabbour, C. J. An assessment of the causes and consequences of agricultural land abandonment in Europe. International Journal of Sustainable Development & World Ecology 24 , 554–560 (2017). Zhou, T., Koomen, E. & Ke, X. Determinants of Farmland Abandonment on the Urban–Rural Fringe. Environmental Management 65 , 369–384 (2020). Ustaoglu, E. & J. Collier, M. Farmland abandonment in Europe: an overview of drivers, consequences, and assessment of the sustainability implications. Environmental Reviews 26 , 396–416 (2018). Hong, C., Prishchepov, A. V., Jin, X. & Zhou, Y. Mapping cropland abandonment and distinguishing from intentional afforestation with Landsat time series. International Journal of Applied Earth Observation and Geoinformation 127 , 103693 (2024). Du, Z., Yang, J., Ou, C. & Zhang, T. Agricultural Land Abandonment and Retirement Mapping in the Northern China Crop-Pasture Band Using Temporal Consistency Check and Trajectory-Based Change Detection Approach. IEEE Trans. Geosci. Remote Sensing 60 , 1–12 (2022). Deng, L., Liu, G. & Shangguan, Z. Land-use conversion and changing soil carbon stocks in China’s ‘Grain-for-Green’ Program: a synthesis. Global Change Biology 20 , 3544–3556 (2014). Cramer, V., Hobbs, R. & Standish, R. What’s new about old fields? Land abandonment and ecosystem assembly. Trends in Ecology & Evolution 23 , 104–112 (2008). Bell, S. M., Barriocanal, C., Terrer, C. & Rosell-Melé, A. Management opportunities for soil carbon sequestration following agricultural land abandonment. Environmental Science & Policy 108 , 104–111 (2020). Lasanta, T., Nadal-Romero, E. & Arnáez, J. Managing abandoned farmland to control the impact of re-vegetation on the environment. The state of the art in Europe. Environmental Science & Policy 52 , 99–109 (2015). Wang, L., Pedersen, P. B. M. & Svenning, J.-C. Rewilding abandoned farmland has greater sustainability benefits than afforestation. npj biodivers 2 , 5 (2023). Janus, J. & Bozek, P. Using ALS data to estimate afforestation and secondary forest succession on agricultural areas: An approach to improve the understanding of land abandonment causes. Applied Geography 97 , 128–141 (2018). Yang, Y. et al. Restoring Abandoned Farmland to Mitigate Climate Change on a Full Earth. One Earth 3 , 176–186 (2020). Yang, Y. et al. Restoring Abandoned Farmland to Mitigate Climate Change on a Full Earth. One Earth 3 , 176–186 (2020). Crawford, C. L., Wiebe, R. A., Yin, H., Radeloff, V. C. & Wilcove, D. S. Biodiversity consequences of cropland abandonment. Nat Sustain 7 , 1596–1607 (2024). Anpilogova, D. & Pakina, A. Assessing ecosystem services of abandoned agricultural lands: a case study in the forested zone of European Russia. OE 7 , e77969 (2022). MARAPPR. Guidance on the integrated utilization of abandoned cropland for the development of agricultural production. (2021). Zhu, X., Xiao, G., Zhang, D. & Guo, L. Mapping abandoned farmland in China using time series MODIS NDVI. Science of The Total Environment 755 , 142651 (2021). Wang, H. et al. Is Abandoned Cropland Continuously Growing in China? Quantitative Evidence and Enlightenment from Landsat-Derived Annual China Land Cover Dataset. Land 13 , 45 (2023). Shi, T., Li, X., Xin, L. & Xu, X. The spatial distribution of farmland abandonment and its influential factors at the township level: A case study in the mountainous area of China. Land Use Policy 70 , 510–520 (2018). MARA. List of mountainous and hilly counties in China. (2019). Cao, B. et al. A 30 m terrace mapping in China using Landsat 8 imagery and digital elevation model based on the Google Earth Engine. Earth Syst. Sci. Data 13 , 2437–2456 (2021). Hong, S. et al. Divergent responses of soil organic carbon to afforestation. Nat Sustain 3 , 694–700 (2020). Vesterdal, L., Ritter, E. & Gundersen, P. Change in soil organic carbon following afforestation of former arable land. Forest Ecology and Management 169 , 137–147 (2002). Cook-Patton, S. C. et al. Mapping carbon accumulation potential from global natural forest regrowth. Nature 585 , 545–550 (2020). Kristensen, J. Å. et al. Tree planting is no climate solution at northern high latitudes. Nat. Geosci. 17 , 1087–1092 (2024). Liu, T. et al. Environmental laws and ecological restoration projects enhancing ecosystem services in China: A meta-analysis. Journal of Environmental Management 327 , 116810 (2023). Chen, X. et al. Unraveling carbon stock dynamics and their determinants in China’s Loess Plateau over the past 40 years. Ecological Indicators 159 , 111760 (2024). Qi, S., Cheng, S., Tan, X., Feng, S. & Zhou, Q. Predicting China’s carbon price based on a multi-scale integrated model. Applied Energy 324 , 119784 (2022). Zhao, N., Chen, K., Wu, X., Zhang, L. & Wang, W. Cropland fragmentation change across China over the last two decades. Agricultural Systems 218 , 104010 (2024). Deng, O. et al. Managing fragmented croplands for environmental and economic benefits in China. Nat Food 5 , 230–240 (2024). Zhang, Q. et al. Outlook of China’s agriculture transforming from smallholder operation to sustainable production. Global Food Security 26 , 100444 (2020). Zhang, M. et al. Reveal the severe spatial and temporal patterns of abandoned cropland in China over the past 30 years. Science of The Total Environment 857 , 159591 (2023). He, T. et al. Quantitative analysis of abandonment and grain production loss under armed conflict in Ukraine. Journal of Cleaner Production 412 , 137367 (2023). Yu, Z. et al. Forest expansion dominates China’s land carbon sink since 1980. Nat Commun 13 , 5374 (2022). Tong, X. et al. Forest management in southern China generates short term extensive carbon sequestration. Nat Commun 11 , 129 (2020). Cheng, K. et al. Carbon storage through China’s planted forest expansion. Nat Commun 15 , 4106 (2024). Chen, Y. et al. Reforestation makes a minor contribution to soil carbon accumulation in the short term: Evidence from four subtropical plantations. Forest Ecology and Management 384 , 400–405 (2017). Yun, H. et al. Changes in above- versus belowground biomass distribution in permafrost regions in response to climate warming. Proc. Natl. Acad. Sci. U.S.A. 121 , e2314036121 (2024). Gvein, M. H. Potential of land-based climate change mitigation strategies on abandoned cropland. Næss, J. S., Cavalett, O. & Cherubini, F. The land–energy–water nexus of global bioenergy potentials from abandoned cropland. Nat Sustain 4 , 525–536 (2021). Griscom, B. W. et al. Natural climate solutions. Proc. Natl. Acad. Sci. U.S.A. 114 , 11645–11650 (2017). Meyfroidt, P., Schierhorn, F., Prishchepov, A. V., Müller, D. & Kuemmerle, T. Drivers, constraints and trade-offs associated with recultivating abandoned cropland in Russia, Ukraine and Kazakhstan. Global Environmental Change 37 , 1–15 (2016). Affek, A. N., Wolski, J., Latocha, A., Zachwatowicz, M. & Wieczorek, M. The use of LiDAR in reconstructing the pre‐World War II landscapes of abandoned mountain villages in southern Poland. Archaeological Prospection 29 , 157–173 (2022). Nguyen, H. D. et al. Cropland abandonment and flood risks: Spatial analysis of a case in North Central Vietnam. Anthropocene 38 , 100341 (2022). García-Ruiz, J. M. & Lana-Renault, N. Hydrological and erosive consequences of farmland abandonment in Europe, with special reference to the Mediterranean region – A review. Agriculture, Ecosystems & Environment 140 , 317–338 (2011). Liang, Y., Li, S., Feldman, M. W. & Daily, G. C. Does household composition matter? The impact of the Grain for Green Program on rural livelihoods in China. Ecological Economics 75 , 152–160 (2012). Wang, K. et al. Impact of the Grain for Green Project on the Well-Being of Farmer Households: A Case Study of the Mountainous Areas of Northern Hebei Province, China. Land 12 , 1257 (2023). Strassburg, B. B. N. et al. Global priority areas for ecosystem restoration. Nature 586 , 724–729 (2020). Duan, J. et al. Consolidation of agricultural land can contribute to agricultural sustainability in China. Nat Food 2 , 1014–1022 (2021). Ren, C. et al. Ageing threatens sustainability of smallholder farming in China. Nature 616 , 96–103 (2023). Doelman, J. C., Stehfest, E., Tabeau, A. & Van Meijl, H. Making the Paris agreement climate targets consistent with food security objectives. Global Food Security 23 , 93–103 (2019). Wu, X. et al. The potential role of abandoned cropland for food security in China. Resources, Conservation and Recycling 212 , 108004 (2025). Peng, S. et al. Carbon restoration potential on global land under water resource constraints. Nat Water 2 , 1071–1081 (2024). Zan, B. et al. Spatiotemporal inequality in land water availability amplified by global tree restoration. Nat Water 2 , 863–874 (2024). Xu, Y. et al. Annual 30-m land use/land cover maps of China for 1980–2015 from the integration of AVHRR, MODIS and Landsat data using the BFAST algorithm. Sci. China Earth Sci. 63 , 1390–1407 (2020). Yang, J. & Huang, X. The 30 m annual land cover dataset and its dynamics in China from 1990 to 2019. Earth Syst. Sci. Data 13 , 3907–3925 (2021). Tu, Y. et al. A 30 m annual cropland dataset of China from 1986 to 2021. Earth Syst. Sci. Data 16 , 2297–2316 (2024). FAO. Methods & Standards. (2016). Deng, L., Liu, G. & Shangguan, Z. Land-use conversion and changing soil carbon stocks in China’s ‘Grain-for-Green’ Program: a synthesis. Global Change Biology 20 , 3544–3556 (2014). GOV. Comprehensive Benefit Assessment Report of China’s Grain for Green Project. (2022). Yan, Y. et al. Climate-induced tree-mortality pulses are obscured by broad-scale and long-term greening. Nat Ecol Evol 8 , 912–923 (2024). Li, X. et al. Mapping global urban boundaries from the global artificial impervious area (GAIA) data. Environ. Res. Lett. 15 , 094044 (2020). Pekel, J.-F., Cottam, A., Gorelick, N. & Belward, A. S. High-resolution mapping of global surface water and its long-term changes. Nature 540 , 418–422 (2016). Olofsson, P. et al. Good practices for estimating area and assessing accuracy of land change. Remote Sensing of Environment 148 , 42–57 (2014). Song, X., Peng, C., Zhou, G., Jiang, H. & Wang, W. Chinese Grain for Green Program led to highly increased soil organic carbon levels: A meta-analysis. Sci Rep 4 , 4460 (2014). Chen, S. et al. GSOCS-LULCC: the Global Soil Organic Carbon Stock dataset after Land Use and Land Cover Change. Preprint at https://doi.org/10.5194/essd-2024-373 (2024). Deng, L., Liu, G. & Shangguan, Z. Carbon storage in successional and plantation forest soils: a tropical analysis. Global Change Biology 20 , 3544–3556 (2014). Marín‐Spiotta, E. & Sharma, S. Carbon storage in successional and plantation forest soils: a tropical analysis. Global Ecology and Biogeography 22 , 105–117 (2013). Hengl, T. et al. SoilGrids250m: Global gridded soil information based on machine learning. PLoS ONE 12 , e0169748 (2017). Wang, S. et al. Vertical Distribution of Soil Organic Carbon in China. Environmental Management 33 , (2004). Yang, X. et al. Nitrogen addition promotes soil carbon accumulation globally. Sci. China Life Sci. (2024) doi:10.1007/s11427-024-2752-2. Liu, T. et al. Utilizing nighttime light datasets to uncover the spatial patterns of county-level relative poverty-returning risk in China and its alleviating factors. Journal of Cleaner Production 448 , 141682 (2024). Yates, L. A., Aandahl, Z., Richards, S. A. & Brook, B. W. Cross validation for model selection: A review with examples from ecology. Ecological Monographs 93 , e1557 (2023). Yang, K. et al. Seabuckthorn (Hippophae rhamnoides L.) plantation degradation aggravates microbial metabolic C and P limitations on the Northern Loess Plateau in China. Science of The Total Environment 945 , 174088 (2024). Wang, J. et al. Large Chinese land carbon sink estimated from atmospheric carbon dioxide data. Nature 586 , 720–723 (2020). GOV. Management of Forestry and Grassland Reform and Development Funds. (2024). Weng, Q. & Xu, H. A review of China’s carbon trading market. Renewable and Sustainable Energy Reviews 91 , 613–619 (2018). Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryInformation.docx Afforestation on Abandoned Croplands in China Has the Potential to Increase Carbon Sequestration by half Abandonedlandaerialphoto1.jpg Abandonedlandaerialphoto2.jpg Abandonedlandaerialphoto3.jpg Abandonedlandaerialphoto4.jpg Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6119575","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":423485517,"identity":"8fbc336b-26d9-487a-a4b3-3b281aa24cc0","order_by":0,"name":"Le Yu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA50lEQVRIiWNgGAWjYBACPhDxgYGBB4yAgLGBkBY2kKoZCaRqYeZJADGJ1sLeYyZt+6NOxpx/7TGJHww2shsOMD97gFcLzxkz6ZyEwzyWM96lSfYwpBlvOMBmboBXi0QOSMsBHoMbZ8wkeBgOJ244wMMmgVeL/BszaYuEOrAWyT8M/4nQIsFjJs2QwMxjcB7oKR6GA0Ro4UkrtuxJOwy0hcfYWsYg2XjmYTYzvFr42Q9vvPHDps7e4PwZw5tvKuxk+443P8OrhYGBAxo8EglAAsRmxq8eCNgfQO07QFDpKBgFo2AUjFAAAAbHQChm2nAgAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0003-3115-2042","institution":"Department of Earth System Science, Ministry of Education Key Laboratory for Earth System Modeling, Institute for Global Change Studies, Tsinghua University","correspondingAuthor":true,"prefix":"","firstName":"Le","middleName":"","lastName":"Yu","suffix":""},{"id":423485518,"identity":"394b3bc0-5ada-418d-b4d1-dff285309a1d","order_by":1,"name":"Tao Liu","email":"","orcid":"","institution":"Ministry of Education Key Laboratory for Earth System Modeling, Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Tao","middleName":"","lastName":"Liu","suffix":""},{"id":423485519,"identity":"b9a8503c-5105-4547-bf15-ce2f914d9c4d","order_by":2,"name":"Ying Tu","email":"","orcid":"https://orcid.org/0000-0002-2240-5389","institution":"Cornell University","correspondingAuthor":false,"prefix":"","firstName":"Ying","middleName":"","lastName":"Tu","suffix":""},{"id":423485520,"identity":"0ed748c4-91d3-4a4c-aa0c-ac453ce62f5c","order_by":3,"name":"Xin Chen","email":"","orcid":"","institution":"Institute of Loess Plateau, Shanxi University","correspondingAuthor":false,"prefix":"","firstName":"Xin","middleName":"","lastName":"Chen","suffix":""},{"id":423485521,"identity":"21027f1e-9b4e-41ef-8ccf-394c0d13b115","order_by":4,"name":"Zhenrong Du","email":"","orcid":"","institution":"Dalian University of Technology","correspondingAuthor":false,"prefix":"","firstName":"Zhenrong","middleName":"","lastName":"Du","suffix":""},{"id":423485522,"identity":"fcd5b11c-19aa-49a7-a4ff-1eed3316a76e","order_by":5,"name":"Hui Wu","email":"","orcid":"","institution":"Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Hui","middleName":"","lastName":"Wu","suffix":""},{"id":423485523,"identity":"8d371189-5606-4bb7-bdc3-2528e4dc1808","order_by":6,"name":"Shijun Zheng","email":"","orcid":"","institution":"Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Shijun","middleName":"","lastName":"Zheng","suffix":""},{"id":423485524,"identity":"9e4075df-d621-4d29-8b09-03dcd275446a","order_by":7,"name":"Minxuan Sun","email":"","orcid":"https://orcid.org/0009-0005-1670-3485","institution":"Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Minxuan","middleName":"","lastName":"Sun","suffix":""},{"id":423485525,"identity":"24848ecf-b679-4b8a-a61d-eabfadce26b9","order_by":8,"name":"Yixuan Li","email":"","orcid":"","institution":"Centre for Environmental Policy, Imperial College London","correspondingAuthor":false,"prefix":"","firstName":"Yixuan","middleName":"","lastName":"Li","suffix":""},{"id":423485526,"identity":"5fc2ac14-444e-41e4-9a67-836cebf99fd2","order_by":9,"name":"Dailiang Peng","email":"","orcid":"https://orcid.org/0000-0003-1159-0723","institution":"Key Laboratory of Digital Earth Science, Aerospace Information Research Institute, Chinese Academy of Sciences","correspondingAuthor":false,"prefix":"","firstName":"Dailiang","middleName":"","lastName":"Peng","suffix":""},{"id":423485527,"identity":"7dfd2bb0-35fd-473a-b3ac-51dac4887647","order_by":10,"name":"Chao Wu","email":"","orcid":"https://orcid.org/0000-0003-3233-856X","institution":"Department of Earth System Science, Ministry of Education Key Laboratory for Earth System Modeling, Institute for Global Change Studies, Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Chao","middleName":"","lastName":"Wu","suffix":""},{"id":423485528,"identity":"ecb64739-5eb6-4f92-b657-4205ef893a23","order_by":11,"name":"Yuyu Zhou","email":"","orcid":"https://orcid.org/0000-0003-1765-6789","institution":"The University of Hong Kong","correspondingAuthor":false,"prefix":"","firstName":"Yuyu","middleName":"","lastName":"Zhou","suffix":""}],"badges":[],"createdAt":"2025-02-27 09:37:17","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6119575/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6119575/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":77687699,"identity":"eb599414-21ed-4511-93ee-2451f11626f9","added_by":"auto","created_at":"2025-03-04 09:15:48","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":9421783,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAbandoned cropland in China: extent and intensity. (\u003c/strong\u003eMaps showing cropland conversion to natural vegetation (\u003cstrong\u003ea\u003c/strong\u003e), including afforestation (\u003cstrong\u003ed\u003c/strong\u003e) and abandonment (\u003cstrong\u003ec\u003c/strong\u003e), with subplot bar charts displaying the average area of abandoned cropland within a 5 km²patches. Annual area of cropland abandonment and its rate from 2000 to 2015 (\u003cstrong\u003eb\u003c/strong\u003e). Proportion of cropland abandonment in mountainous and hilly counties (green color) versus other counties (yellow color) (\u003cstrong\u003ee\u003c/strong\u003e). Annual abandonment rate in terraced cropland and its uncertainty (\u003cstrong\u003ef\u003c/strong\u003e). Vegetation index trends distinguishing between afforestation and abandonment (\u003cstrong\u003eg\u003c/strong\u003e). The color scale is classified into deciles (\u003cstrong\u003ea\u003c/strong\u003e), and the maps are aggregated from 30m (1 arc-second) to 5 km resolution for improved visualization.\u003c/p\u003e","description":"","filename":"Fig1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6119575/v1/aa58808af76717a759527013.jpg"},{"id":77687700,"identity":"5fa9ee55-591c-4619-84a8-026ebd3666bf","added_by":"auto","created_at":"2025-03-04 09:15:48","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":6837172,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCarbon accumulation rates after cropland abandonment. \u003c/strong\u003e(Linear relationship between stand age and SOC (\u003cstrong\u003ea\u003c/strong\u003e) and biomass carbon (\u003cstrong\u003eb\u003c/strong\u003e) density in abandoned cropland, derived from the literature. Bivariate map of SOC and biomass carbon accumulation rates in abandoned cropland (\u003cstrong\u003ec\u003c/strong\u003e). Five-years changes in SOC and biomass carbon density across six major geographic regions (\u003cstrong\u003ed−j\u003c/strong\u003e), with the dark sections of the pie chart indicating the proportion of abandoned cropland. For SOC, a threshold of 0 Mg C ha⁻¹ yr⁻¹ is used; for biomass carbon, the threshold accumulation rates of 1.68 Mg C ha⁻¹ yr⁻¹ is referenced from the average carbon accumulation rate of natural forest regeneration in China\u003csup\u003e39\u003c/sup\u003e.)\u003c/p\u003e","description":"","filename":"Fig2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6119575/v1/8ee501c6832c59ef5f2ebb60.jpg"},{"id":77689672,"identity":"a45fa2b0-9550-4e9b-82b2-36b3cb0ab5e7","added_by":"auto","created_at":"2025-03-04 09:31:49","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":4218477,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCarbon sequestration potential of afforestation under SSP scenarios. \u003c/strong\u003e(Carbon sequestration potential of SOC (\u003cstrong\u003ea\u003c/strong\u003e) and biomass carbon (\u003cstrong\u003eb\u003c/strong\u003e) in afforestation of abandoned croplands under four representative SSP scenarios. The purple and green lines represent the carbon sequestration trends for natural regeneration and afforestation, respectively, with 60% transparency indicating the error margin. The yellow fill represents the carbon sequestration gap, comparing afforestation with abandoned cropland. The spatially explicit map shows the gains in SOC (\u003cstrong\u003ec\u003c/strong\u003e) and biomass carbon (\u003cstrong\u003ed\u003c/strong\u003e) from afforestation carbon sequestration relative to natural regeneration by 2060.)\u003c/p\u003e","description":"","filename":"Fig3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6119575/v1/76e2ffae1ca1fb808b5af4b9.jpg"},{"id":77688744,"identity":"5c5e7b1e-9e21-41df-ae64-0747380337b8","added_by":"auto","created_at":"2025-03-04 09:23:49","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":3354995,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eA theoretical framework for afforestation on abandoned cropland and spillover effects. \u003c/strong\u003e(The theoretical framework is divided into two parts based on the background colors: blue and green. Part I (blue background) describes the closed-loop process of government investment in afforestation on abandoned cropland and the carbon sequestration benefits derived from vegetation, along with the spillover effects. On the one hand, afforestation generates additional ecosystem services, such as habitat quality, soil and water conservation, and cultural services, all of which contribute positively to the environmental pillar of the SDGs. On the other hand, this section highlights subsidies related to the transformation of abandoned cropland, such as land subsidies and seedling subsidies, which not only increase local farmers’ income and reduce rural-urban inequalities but also provide employment opportunities, such as forest management positions. In Part II (green background), we emphasize the causes of abandoned cropland (lower-left corner), the natural succession process, and the dynamics of biomass carbon and SOC density (lower-right corner). We also present four aerial images of abandoned cropland, obtained through field investigations in various regions of China.)\u003c/p\u003e","description":"","filename":"Fig4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6119575/v1/d7bdb834eb143ed2e422d0ba.jpg"},{"id":80752418,"identity":"f9c3e08a-26b8-4a13-8b34-5eac442579f8","added_by":"auto","created_at":"2025-04-16 16:49:47","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":24885303,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6119575/v1/fcfa7378-1acb-43a3-a30a-4acdcff2a29a.pdf"},{"id":77688748,"identity":"41319f04-443c-45b2-bcd4-9f340388cda8","added_by":"auto","created_at":"2025-03-04 09:23:49","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":3254932,"visible":true,"origin":"","legend":"Afforestation on Abandoned Croplands in China Has the Potential to Increase Carbon Sequestration by half","description":"","filename":"SupplementaryInformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-6119575/v1/d974416269e4cce86ccf9659.docx"},{"id":77687732,"identity":"205997f2-9cb9-436a-8c0d-d434166920c1","added_by":"auto","created_at":"2025-03-04 09:15:49","extension":"jpg","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":4191432,"visible":true,"origin":"","legend":"","description":"","filename":"Abandonedlandaerialphoto1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6119575/v1/1246d5c1a8d735f6ac30ca68.jpg"},{"id":77689667,"identity":"5e0ece69-197c-4854-8542-914f7ea8641b","added_by":"auto","created_at":"2025-03-04 09:31:49","extension":"jpg","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":3788166,"visible":true,"origin":"","legend":"","description":"","filename":"Abandonedlandaerialphoto2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6119575/v1/cfb40a132afe80bfd80c7578.jpg"},{"id":77688755,"identity":"9f603748-140e-4d49-8da4-56a74cf5b227","added_by":"auto","created_at":"2025-03-04 09:23:49","extension":"jpg","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":3795129,"visible":true,"origin":"","legend":"","description":"","filename":"Abandonedlandaerialphoto3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6119575/v1/fec5ec5983e8c99e0544be32.jpg"},{"id":77687717,"identity":"3cbbd1c5-f7b7-4515-b6de-3717eb97a14f","added_by":"auto","created_at":"2025-03-04 09:15:49","extension":"jpg","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":2331994,"visible":true,"origin":"","legend":"","description":"","filename":"Abandonedlandaerialphoto4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6119575/v1/808e539123202dcb83fccbbb.jpg"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Afforestation on Abandoned Croplands in China Has the Potential to Increase Carbon Sequestration by half","fulltext":[{"header":"Main","content":"\u003cp\u003eAfforestation is widely recognized as one of the most cost-effective nature-based solutions for climate change mitigation, capturing atmospheric carbon while simultaneously contributing to human well-being\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. Since the early 21st century, China has implemented a series of afforestation initiatives aimed at enhance ecosystem services, such as soil and water conservation, sand control, carbon sequestration, and biodiversity\u003csup\u003e\u003cspan additionalcitationids=\"CR4\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. These efforts expanded China\u0026rsquo;s forest cover by 42.98 Mha (forest cover rate increased from 18.21\u0026ndash;23.04%) from 2000 to 2020 in China, sequestrated extra woody biomass carbon of 208.6\u0026thinsp;\u0026plusmn;\u0026thinsp;51.8 TgC yr\u003csup\u003e\u0026minus;\u0026thinsp;1 6\u003c/sup\u003e. China\u0026rsquo;s Grain for Green program, one of the largest ongoing restoration initiatives, has delivered substantial ecological and socio-economic benefits, with its impacts assessed as being 4.5 times the government\u0026rsquo;s investment\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. The carbon sequestration potential of this ambitious project has been capable of offsetting 3\u0026ndash;5% of China\u0026rsquo;s annual carbon emissions\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. However, the pace of forest expansion has slowed in recent years, and the area of land suitable for further afforestation has become increasingly saturated\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. This presenting a substantial challenge to achieving ambitious China\u0026rsquo;s carbon neutrality goal by 2060 and the mid-term target of 30% forest cover by 2035.\u003c/p\u003e \u003cp\u003eAbandoned cropland with natural regeneration has been proven globally to have significant carbon sequestration potential\u003csup\u003e\u003cspan additionalcitationids=\"CR11 CR12 CR13 CR14\" citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. Cropland abandonment is primarily driven by a combination of socio-economic and environmental factors, such as the potential hazards of climate change, disasters, and the low productivity efficiency of marginalized cropland\u003csup\u003e\u003cspan additionalcitationids=\"CR17\" citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. The ecological processes of cropland abandonment are typically accompanied by spontaneous vegetation recovery\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. The process of abandonment typically requires considerable time to reach a climax community, a process that takes even longer in high-latitude regions\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. In arid and semi-arid regions, cropland abandonment often leads to a decline in surface SOC during the early stages, and this loss can take a long time to recover\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e,\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e. The lack of vegetation recovery control on abandoned cropland may pose a risk of accelerated decomposition of fast-growing plants (e.g., straw, litter, and residues), potentially leading to increased CO₂ emissions\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e,\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. Recent studies and land management projects emphasize that alternative methods, such as afforestation, rewilding, and bioenergy crop cultivation, can effectively enhance the carbon capture and storage potential of abandoned cropland\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan additionalcitationids=\"CR26\" citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. These strategies can establish a closed \u0026ldquo;biochar-soil\u0026rdquo; system that continuously removes CO₂ and stores it in soil and plants\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e. By optimizing management practices, abandoned cropland can simultaneously restore soil fertility\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e, promote biodiversity\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e, and enhance other ecosystem services\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eAfforestation on abandoned cropland provides new insights into close China\u0026rsquo;s ambitious afforestation and carbon-neutral targets\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. However, the role of afforestation of abandoned cropland in climate change mitigation remains unclear, as the potential ceiling for carbon sequestration through afforestation, compared to natural regeneration, remains be further confirmed. Understanding spatial prioritization strategies for abandoned cropland restoration, based on carbon sequestration potential, is critical for helping land-managers and stakeholders identify the most effective solutions\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. Furthermore, China\u0026rsquo;s government guidelines for abandoned cropland emphasize flexible strategies, such as encouraging farmers to plant trees; however, there is currently a lack of actionable policies to guide and implement restoration efforts\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. This gap is largely due to the absence of robust scientific assessments regarding the investments against potential benefits. Overall, there is an urgent need for comprehensive quantitative studies to assess the carbon sequestration potential and cost-benefit trade-offs associated with afforestation on abandoned cropland.\u003c/p\u003e \u003cp\u003eHere, we evaluate the carbon sequestration potential of biomass carbon (above- and below-ground biomass) and SOC (0\u0026ndash;30 cm) through afforestation and natural regeneration scenarios on abandoned cropland across China at a 30m resolution, using 10,818 empirical carbon data (3502 SOC and 7316 biomass carbon data) derived from 298 peer-reviewed articles, combined with multisource remote sensing data, machine learning, and scenario modeling methods (see Methods). Additionally, we assess the government investment on afforestation of abandoned cropland and the potential benefits from the carbon market by 2060 to estimate the cost-benefits of our proposed incentives. Our study try to addresses three key research questions: (1) How much abandoned cropland in China has undergone natural regeneration since the 21st century? (2) What is the carbon sequestration potential by 2060 under scenarios of abandoned cropland either through natural regeneration or afforestation? (3) Considering funding costs and potential carbon benefits, what is the feasibility of afforestation of abandoned cropland? Collectively, our analysis provides a timely and comprehensive estimate of the carbon sequestration potential of abandoned cropland in China, offering valuable insights for policymakers to support the carbon neutrality target by 2060, while considering both investment and return factors.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003eAbandoned cropland patterns in China\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBased on land cover change trajectories and vegetation change trends (see Methods), we generated three sets of 30m resolution maps of abandoned cropland using China\u0026rsquo;s Land Cover Dataset (CLCD), China\u0026rsquo;s Land-Use/cover Datasets-A (CLUD-A), and China\u0026rsquo;s Land Cover Dataset (CACD), respectively. Following established methods for assessing land cover change mapping accuracy, we report the accuracy assessment parameters for abandoned cropland (Fig. S1). The results show that the abandoned cropland map generated from the CACD exhibited the highest accuracy, with an overall accuracy of 82.17%, user accuracy of 86.57%, producer accuracy of 76.16%, and an F1 score of 0.811 (Fig. S1). Compared with other coarser-resolution maps of abandoned cropland in China\u003csup\u003e14,32,33\u003c/sup\u003e, our abandoned cropland product shows superior performance, particularly in the agro-pastoral transition zones and the karst regions of southwest China.\u003c/p\u003e\n\u003cp\u003eWe identify that 17.86 Mha of cropland has been converted to natural vegetation from 2000 to 2017 (Fig. 1a), with 66.23% of this area allocated for afforestation (e.g., the Green for Grain Program) (Fig. 1d). A total of 6.03 Mha (33.77%) of cropland has been abandoned with natural regeneration (Fig. 1c), corresponding to an annual abandonment rate of 1.60\u0026plusmn;0.13% of active cropland (Fig. 1b), which is equivalent to 1.7 times the area of Hainan province. This finding confirms the widespread occurrence of cropland abandonment in China, particularly in farming-pastoral ecotone of northern China (1.67 Mha, 27.77%), the depopulated areas of Northeastern China (0.70 Mha, 11.64%), the Karst regions of Southwestern China (1.06 Mha, 17.62%), and the Southern hilly areas (0.99 Mha, 16.37%). This phenomenon is influenced by a combination of factors, including climate change, natural disasters, limitations in agricultural infrastructure, and\u003csup\u003e\u0026nbsp;\u003c/sup\u003eagricultural labor shortage\u003csup\u003e34\u003c/sup\u003e. According to the classification of mountainous and hilly counties defined by the Chinese government\u003csup\u003e35\u003c/sup\u003e, we find that 70.63% of the abandoned cropland occurs in these areas (Fig. 1e). Furthermore, by overlaying the China\u0026rsquo;s terraced map\u003csup\u003e36\u003c/sup\u003e, we further find that 5.06\u0026plusmn;1.21% of the terraced land has undergo abandonment (Fig. 1f).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCarbon accumulation after cropland\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eabandonment\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe use literature-derived data to assess the carbon accumulation over a 70-year period under two abandonment (natural regeneration) and afforestation (Fig. 2a\u0026ndash;b) strategies. Linear regression analysis reveals differences in carbon accumulation between the two strategies (Fig. 2a\u0026ndash;b). Specifically, compared with SOC, carbon accumulation in afforested cropland is primarily driven by increase in biomass carbon (Fig. 2b). The carbon accumulation rates in biomass carbon differ significantly, with \u003cem\u003eSlopes\u003c/em\u003e of 3.2132 Mg C ha\u003csup\u003e-1\u003c/sup\u003e yr\u003csup\u003e-1\u003c/sup\u003e for afforestation and 0.0835 Mg C ha\u003csup\u003e-1\u003c/sup\u003e yr\u003csup\u003e-1\u003c/sup\u003e for abandonment. In contrast, the impacts of afforestation (\u003cem\u003eSlope\u003c/em\u003e = 0.087 Mg C ha\u003csup\u003e-1\u003c/sup\u003e yr\u003csup\u003e-1\u003c/sup\u003e) and abandonment (\u003cem\u003eSlope\u003c/em\u003e = 0.021 Mg C ha\u003csup\u003e-1\u003c/sup\u003e yr\u003csup\u003e-1\u003c/sup\u003e) on SOC are more limited (Fig. 2a), underscoring the well-known challenges in generating SOC, such as carbon instability driven by microbial activity\u003csup\u003e37,38\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eWe generate a binary map of the carbon accumulation rates for abandoned croplands in China (Fig. 2c), combining several covariates (Table S1) and machine learning models (see Methods). By the end of 2023, China\u0026rsquo;s naturally regenerated abandoned land has sequestered 250.01 Tg of biomass carbon and 14.47 Tg of SOC (Fig. 2c). Considering abandonment duration (Fig. S2), we find that the average carbon accumulation rates for biomass carbon and SOC in abandoned croplands are 1.783 Mg C ha⁻\u0026sup1; yr⁻\u0026sup1; and 0.285 Mg C ha⁻\u0026sup1; yr⁻\u0026sup1;, respectively (Fig. 2c). The spatial results show that in arid/semi-arid regions, such as the Loess Plateau, Inner Mongolia grasslands, and Xinjiang, natural regeneration of post-agricultural vegetation leads to a more pronounced increase in SOC than in biomass carbon during natural regeneration (Fig. 2c). In the Northeastern regions, SOC is more susceptible to long-term decline risks, during the early stages of cropland abandonment, being particularly vulnerable to the influences of climate change, soil physicochemical properties, and shifts in microbial community structure\u003csup\u003e21\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eWe further find that the biomass carbon sequestration in the first decade after abandonment (23.05 Mg C ha⁻\u0026sup1;) was higher than in the second decade (12.37 Mg C ha⁻\u0026sup1;) (Fig. 2d). In contrast, the SOC accumulation was more pronounced in the second decade (3.27 Mg C ha⁻\u0026sup1;), about 1.4 times higher than first decade (Fig. 2d). Across different geographic regions, the carbon sequestration on abandoned cropland was most significant in the South-central (2.80 Mg C ha⁻\u0026sup1; yr⁻\u0026sup1;) (Fig. 2j) and Southwest regions (3.35 Mg C ha⁻\u0026sup1; yr⁻\u0026sup1;) (Fig. 2i). We also observed SOC loss (\u0026minus;0.34 Mg C ha⁻\u0026sup1;) in the Northeast region during the first five years of cropland abandonment (Fig. 2e). These results indicate significant spatial variations in SOC and biomass carbon changes due to cropland abandonment, particularly in areas experiencing SOC loss.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCarbon sequestration potential on abandoned lands by 2060\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBased on applicable afforestation strategies, including species selection and planting density (see Methods), our results show that afforestation on abandoned cropland in China could result in a total carbon sequestration (including biomass carbon and SOC from 0\u0026ndash;30 cm) of 900.73 \u0026ndash; 926.83 Tg C by 2060, with 301.49 \u0026ndash; 302.55 Tg of SOC and 598.17 \u0026ndash; 624.07 Tg of biomass carbon (Fig. 3a\u0026ndash;b). Afforestation sequesters 215.12 \u0026ndash; 218.94 Tg more biomass carbon and 15.87 \u0026ndash; 17.12 Tg more SOC than the natural regeneration of abandoned cropland, resulting in a 51.95 \u0026ndash; 53.94% increase in carbon sequestration (Fig. 3a\u0026ndash;b). This result demonstrates the significant carbon sequestration potential of afforestation on abandoned cropland.\u003c/p\u003e\n\u003cp\u003eDuring the first decade of afforestation, the carbon accumulation rate in SOC (0.182 \u0026ndash; 0.193 Mg C ha\u003csup\u003e\u0026minus;1\u003c/sup\u003e yr\u003csup\u003e\u0026minus;1\u003c/sup\u003e) is lower than that of natural regeneration (0.251 \u0026ndash; 0.297 Mg C ha\u003csup\u003e\u0026minus;1\u003c/sup\u003e yr\u003csup\u003e\u0026minus;1\u003c/sup\u003e), with this trend being more pronounced in China\u0026rsquo;s arid and semi-arid regions (Fig. 3a and 3c). However, this trend reverses after 2040, with afforestation\u0026rsquo;s SOC accumulation rate surpassing that of abandoned cropland (afforestation: 0.264 \u0026ndash; 0.272 Mg C ha\u003csup\u003e\u0026minus;1\u003c/sup\u003e yr\u003csup\u003e\u0026minus;1\u003c/sup\u003e; abandoned cropland: 0.109 \u0026ndash; 0.148 Mg C ha\u003csup\u003e\u0026minus;1\u003c/sup\u003e yr\u003csup\u003e\u0026minus;1\u003c/sup\u003e) (Fig. 3a), driven by the accumulation of leaf litter, humus, and root decomposition from woody plants\u003csup\u003e38\u003c/sup\u003e. In contrast, the gap in biomass carbon accumulation rate between afforestation and cropland abandonment widens over time, with afforestation sequestering 1.35 \u0026ndash; 1.51 Mg C ha\u003csup\u003e\u0026minus;1\u003c/sup\u003e yr\u003csup\u003e\u0026minus;1\u0026nbsp;\u003c/sup\u003e(Fig. 3b). As trees approach maturity, the growth rate stabilizes and even slows, causing biomass carbon accumulation rate approaching saturation (0.363 \u0026ndash; 0.755 Mg C ha\u003csup\u003e\u0026minus;1\u003c/sup\u003e yr\u003csup\u003e\u0026minus;1\u003c/sup\u003e) in afforested areas (Fig. 3b).\u003c/p\u003e\n\u003cp\u003eCompared to cropland abandonment, the afforestation scenario will sequester additional substantial SOC in Central and Southern China, with net increases ranging from 7.54 \u0026ndash; 8.55 Tg C by 2060 (Fig. 3c). The Southwest region also shows notable SOC sequestration gains, with net increases ranging from 7.11 \u0026ndash; 7.73 Tg C, suggesting that prioritizing afforestation in these areas would yield greater SOC sequestration benefits (Fig. 3c). In contrast, in northern China, natural regeneration may be more advantageous for SOC sequestration approximately 32.38% (Fig. 3c). This may be the rapid growth of herbaceous vegetation promotes microbial activity, thereby accelerating SOC accumulation\u003csup\u003e40\u003c/sup\u003e. In term of biomass carbon, the southern regions (including East China, Southwest China, and Central South China) will sequester additional biomass carbon with 143.30 \u0026ndash; 143.93 Tg C (20.58 \u0026ndash; 21.76 Mg C ha⁻\u0026sup1;), which is 2.95 \u0026ndash; 3.16 times higher than in the northern regions (Northeast China, North China, and Northwest China) by 2060 (Fig. 3d). Additionally, the results highlight local hotspots of biomass carbon gain potential in the Northeast plain, which may be attributed to region-specific factors such as precipitation, soil nutrients, and physicochemical properties (Fig. 3d) \u003csup\u003e41,42\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eInvestment and carbon benefits\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBased on the subsidy schemes of China\u0026rsquo;s Grain for Green program, and regional afforestation invest list (see Methods), we estimate that implementing afforestation on abandoned cropland will require an investment of approximately $16.254 billion, which includes about $11.378 billion for land subsidies and $4.876 billion for seedling subsidies (Fig. 4). This amount represents 22.3% of the total investment in the two phases of China\u0026rsquo;s Grain for Green program (2000 \u0026ndash; 2020)\u003csup\u003e8\u003c/sup\u003e. By 2060, the carbon sequestration benefits of afforestation will reach USD 7.75 \u0026ndash; 8.06 billion (Fig. 4). This includes revenues of USD 0.74 \u0026ndash; 0.75 billion from SOC and USD 7.01 \u0026ndash; 7.31 billion from biomass carbon. These benefits are an additional USD 2.706 \u0026ndash; 2.755 billion compared to natural regeneration of abandoned cropland (Fig. 4). This is a conservative estimate, as China\u0026rsquo;s carbon equivalent value is expected to rise in the future\u003csup\u003e43\u003c/sup\u003e. Nevertheless, our results show that the carbon benefits from transforming abandoned cropland can cover 47.71% to 49.57% of the investment (Fig. 4). In comparison, the national report on the comprehensive benefits of the Grain for Green program from 2000 to 2020 indicates that the carbon benefits of the program can cover 43.10% of the total investment, which includes USD 72.873 billion in funding and USD 31.411 billion in carbon sequestration return\u003csup\u003e8\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eWe present a theoretical framework for implementing afforestation on abandoned cropland, which not only highlights the cascading effects of government funding inputs and carbon benefits but also emphasizes the spillover effects throughout the process (Fig. 4, blue background). Specifically, we focus on the ecosystem service functions provided by afforestation, such as water conservation and environmental purification (Fig. 4). Reports on the Grain for Green program indicate that the ecosystem service values of these two services are 2.08 and 1.39 times the carbon sequestration value, respectively\u003csup\u003e8\u003c/sup\u003e. Based on the ecological benefits ratio from the Grain for Green program, the ecosystem benefits of afforestation on abandoned cropland could fully offset the government\u0026rsquo;s financial investment (Fig. 4). From a socio-economic perspective, government investment in afforestation significantly enhances poverty alleviation outcomes, mitigates urban-rural inequalities, and provides employment opportunities for local farmers (Fig. 4). Afforestation on abandoned cropland can also serve as an ecological engineering approach with poverty alleviation potential, directly or indirectly contributing to the achievement of the SDGs, particularly SDGs: 1, 2, and 3, while enhancing the coupling between people and the environment (Fig. 4).\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIdentifying the spatiotemporal distribution of abandoned cropland is a critical prerequisite for unlocking additional abandoned land resources\u003csup\u003e11\u003c/sup\u003e. Our results show that the annual abandonment rate in China is 1.60\u0026plusmn;0.13%, and terraced cropland are abandoned at a rate 2.62 \u0026ndash; 3.62 times higher than the average (Fig. 1). We surprisingly reveal that 82.89% of abandoned cropland patches are smaller than 1 ha, which highlights the high exposure of widely environmentally disadvantaged cropland, such as fragmented and steep-slope cropland, to abandonment in China. Those posing significant potential challenges for sustainable land resource use\u003csup\u003e44\u0026ndash;46\u003c/sup\u003e. More importantly, there is no consensus on the established methods for investigating abandoned cropland in China, leading to significant variation in abandonment rates (see Table S4). Some studies also classify land converted to artificial forests, orchards, and afforestation as abandoned cropland\u003csup\u003e33,47\u003c/sup\u003e, leading to an overestimated cropland annual abandonment rate in China\u003csup\u003e48\u003c/sup\u003e. Therefore, the development of coherent and standardized procedures are necessary to conduct top-down surveys of abandoned cropland and ensure the effective reuse of abandoned land resources in China.\u003c/p\u003e\n\u003cp\u003eSince 1980, forest expansion has been the dominant driver of the China\u0026rsquo;s land-based carbon sequestration\u003csup\u003e49\u003c/sup\u003e. The saturation of forest growth and competition for land from agricultural production present significant challenges to these carbon sequestration services\u003csup\u003e50\u003c/sup\u003e. Given the limitations of availability of additional land for afforestation in the future\u003csup\u003e51\u003c/sup\u003e, if strategic afforestation is implemented on abandoned cropland, it could make a considerable potential contribution to China\u0026rsquo;s climate goals. This contribution primarily comes from the Southern regions and biomass carbon\u003csup\u003e52\u003c/sup\u003e, with its carbon sequestration potential being 2.95 \u0026ndash; 3.16 times greater than that of the Northern regions (Fig. 3). Afforestation in the Northern China has a relatively weaker contribution to climate change mitigation\u003csup\u003e40\u003c/sup\u003e. In drier ecosystems, such as alpine meadows and high-altitude grasslands, plants tend to allocate more biomass underground, thereby promoting the long-term accumulation of SOC\u003csup\u003e53\u003c/sup\u003e. This provides valuable and reliable spatial insights for implementing large-scale afforestation efforts on abandoned cropland.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo fully harness the potential of abandoned cropland, further examination is needed of these alternative uses, particularly with regard to their ecological impacts and effects on the atmospheric coupling of ecosystems, such as growing bioenergy crops\u003csup\u003e54,55\u003c/sup\u003e, grassland restoration, and rewilding\u003csup\u003e25\u003c/sup\u003e. Afforestation can also provide a range of co-benefits and ecosystem services, including water filtration, air purification, and soil quality improvement\u003csup\u003e56,57\u003c/sup\u003e. For example, globally, abandoned cropland can serve as habitats for a majority of bird species (62.7%) and mammal species (77.7%)\u003csup\u003e29\u003c/sup\u003e. In addition, transforming abandoned cropland and unlocking its potential for diverse uses, such as landscape conservation, flood prevention, urban cooling, and cultural ecosystem services, is receiving increasing attention\u003csup\u003e58\u0026ndash;60\u003c/sup\u003e. The release of abandoned cropland resources should consider the synergies and trade-offs among multiple functions, maximizing its potential to enhance environmental quality and human well-being.\u003c/p\u003e\n\u003cp\u003eOur calculations show that the cost of transforming abandoned cropland for carbon sequestration is approximately $23.62 \u0026ndash; 24.42 per ton of CO₂, which is slightly higher than the average carbon sequestration cost of previous ecological projects in China (approximately $21.18 per ton of CO₂)\u003csup\u003e9\u003c/sup\u003e (Fig. 4). As a carbon return, by 2060, without accounting for inflation, the carbon benefits from afforestation schemes are projected to offset 47.71 \u0026ndash; 49.57% of government funding inputs (Fig. 4). This estimate does not include additional ecological benefits, such as water conservation, air purification, and biodiversity, nor the broader socio-economic benefits\u003csup\u003e8\u003c/sup\u003e. Moreover, evidence from farmer surveys indicates that the project could improve rural livelihoods in China by increasing income, providing employment opportunities, and enhancing residents\u0026rsquo; well-being\u003csup\u003e61,62\u003c/sup\u003e. These robust findings suggest that afforestation on abandoned land could generate significant environmental and human welfare benefits.\u003c/p\u003e\n\u003cp\u003eAfforestation on abandoned cropland has the potential to unlock significant carbon sequestration, but further efforts are needed to translate this potential into tangible benefits\u003csup\u003e11\u003c/sup\u003e. Various factors, including but not limited to labor supply, afforestation activities, and irrigation, may need improvement to enhance the afforestation adaptability of abandoned cropland. The prioritization of these factors should incorporate local knowledge and context-specific action plans\u003csup\u003e63\u003c/sup\u003e. A more sustainable approach may involve integrating abandoned cropland with agricultural land substitution strategies, such as land swapping or land consolidation\u003csup\u003e64,65\u003c/sup\u003e. Additionally, ensuring the long-term persistence of climate change mitigation potential requires cross-sectoral policies and incentives\u003csup\u003e11\u003c/sup\u003e. Without such efforts, afforestation risks becoming a short-term activity, thereby reducing the achievable potential of abandoned cropland\u003csup\u003e15\u003c/sup\u003e. Establishing and expanding reliable carbon markets is critical for ensuring cross-sectoral coordination in China and securing carbon offsets from reforesting abandoned cropland.\u003c/p\u003e\n\u003cp\u003eHowever, there are several limitations to our findings. Identifying abandoned cropland in China involves overcoming challenges related to data accuracy, which is influenced by remote sensing, biases in identifying abandoned cropland, and variations in land policies\u003csup\u003e37\u003c/sup\u003e. To address these challenges, we employed cross-comparisons of multiple data products and widely accepted approaches for identifying abandoned cropland. These limitations do not diminish the utility of our research; rather, they highlight the importance of cross-scale information\u003csup\u003e11\u003c/sup\u003e. National-level surveys of abandoned cropland are essential for enhancing policymakers\u0026rsquo; understanding of its potential and for informing land policy development, particularly in relation to China\u0026rsquo;s carbon neutrality and food security goals\u003csup\u003e66\u003c/sup\u003e. It is crucial to integrate local-scale field surveys with national survey results to ensure that knowledge of abandoned cropland is translated into actionable strategies that can also benefit local communities\u003csup\u003e11\u003c/sup\u003e. This process must take into account factors such as land tenure, cultural considerations, costs, and labor availability. While we have considered incorporating permanently abandoned cropland from the 21st century into afforestation efforts, a significant portion of this land remains suitable for recultivation (approximately 82.54% in China\u003csup\u003e67\u003c/sup\u003e), aligning with China\u0026rsquo;s strategic food security needs through land improvement and rehabilitation\u003csup\u003e11\u003c/sup\u003e. Furthermore, implementing local afforestation projects on abandoned cropland must safeguard the rights of rural and indigenous communities, leveraging their knowledge to foster success, enhance resilience, and avoid negative social and ethical consequences\u003csup\u003e63\u003c/sup\u003e. Finally, afforestation should be carried out cautiously to prevent adverse environmental impacts. Large-scale afforestation in hotspots, particularly in arid and semi-arid regions such as the Loess Plateau and agro-pastoral transition zones, increases the demand for water resources, potentially leading to water shortages and low tree survival rates\u003csup\u003e68,69\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eOur research demonstrates that unlocking the carbon potential of abandoned cropland and offsetting carbon emissions provides substantial benefits. This represents a crucial pathway for China\u0026rsquo;s dual carbon targets, the United Nations Decade on Ecosystem Restoration, the Sustainable Development Goals, and the next phase of the Grain-for-Green Program. Our simulation results offer valuable insights for policymakers and land-use planners in China, identifying key strategies to maximize the carbon sequestration potential of abandoned cropland.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eWe analyzed the spatial distribution of abandoned cropland in China, along with the carbon sequestration potential and cost-benefit considerations under afforestation scenarios. This analysis integrated empirical data, machine learning models, and scenario analysis methods. The following sections outline the datasets and methodologies employed in this study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMapping abandoned cropland\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo accurately capture the fragmented and geographically marginalized abandoned cropland in China, we utilized three sets of annual land products: China\u0026rsquo;s Land-Use/cover Datasets-A\u003csup\u003e70\u003c/sup\u003e (CLUD-A), China\u0026rsquo;s Land Cover Dataset\u003csup\u003e71\u003c/sup\u003e (CLCD), and China\u0026rsquo;s Annual Cropland Dataset\u003csup\u003e72\u003c/sup\u003e (CACD). These are the only three available datasets providing annual land data with 30m resolution (1 arc-second) for China. The first two are annual land cover products that include various land cover types, such as cropland, forests, grasslands, and impervious surfaces, while the latter provides the annual spatiotemporal distribution of cropland from 1986 to 2021. The CACD product integrates time-series Landsat satellite images, automated training sample generation methods, machine learning algorithms, and LandTrendr techniques, resulting in high user accuracy (0.93), producer accuracy (0.79), and overall accuracy (0.79)\u003csup\u003e72\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eCropland abandonment, as defined by the FAO, refers to cropland that has not been cultivated for at least five years\u003csup\u003e73\u003c/sup\u003e. To mitigate the uncertainty caused by unstable cropland, we ensured that the cropland remained stable for at least five years prior to abandonment\u003csup\u003e45\u003c/sup\u003e. Specifically, a pixel was classified as cropland for the five years preceding abandonment and subsequently transitioned to abandoned cropland category for at least five consecutive years (excluding built-up areas and water bodies). We constructed a decade-long temporal trajectory window in Google Earth Engine, analyzing the data pixel by pixel to identify waste cropland. Pixels where cropland was converted to settlements, water bodies, or reclaimed for at least one year were excluded\u003csup\u003e11\u003c/sup\u003e. This approach helps prevent overestimation of uncultivated pixels due to short-term fallows and removes temporarily non-cultivated cropland, ensuring a more reliable estimate of abandoned cropland. We also did not consider abandoned cropland prior to 2000, as its reliability has been reported to be lower\u003csup\u003e11\u003c/sup\u003e. Importantly, China\u0026rsquo;s Grain-for-Green program, which began in 2000, plays a pivotal role in land restoration efforts on abandoned cropland\u003csup\u003e21\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eHere, we adopt a stricter definition of cropland abandonment. In China, it is defined as the natural regeneration of vegetation rather than afforestation in post-agriculture period\u003csup\u003e19,20,74,75\u003c/sup\u003e. Vegetation growth trends offer new insights for identifying abandoned cropland, distinguishing it from afforestation efforts\u003csup\u003e20\u003c/sup\u003e. In this study, we constructed NDVI\u003csub\u003emax\u003c/sub\u003e datasets with 30m resolution for China\u0026rsquo;s growing season (April to October, 1995 \u0026ndash; 2022) using the maximum value composite method\u0026nbsp;applied to Landsat satellite data, with a temporal resolution of 16 days. NDVI\u003csub\u003emax\u003c/sub\u003e is widely used as a remote sensing proxy to detect changes in terrestrial vegetation activity\u003csup\u003e76\u003c/sup\u003e. We applied quality mask information to eliminate the influence of clouds and their shadows in the Landsat imagery, and used MOD13Q1 data to fill the gaps.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eOur preliminary experiments have shown that the NDVI\u003csub\u003emax\u003c/sub\u003e tends to decrease significantly in the first year of cropland abandonment, whereas for afforestation, the NDVI\u003csub\u003emax\u003c/sub\u003e shows a clear increase, indicating vegetation growth due to tree planting\u003csup\u003e19\u003c/sup\u003e. Based on a systematic review of methods for distinguishing between abandoned cropland and afforestation (Table S3) and multiple preliminary experiments (Fig. 1g), we propose a combined algorithm based on land cover change trajectories and vegetation index trends. Land change trajectories were used to identify the year of cropland abandonment. Subsequently, we applied a double ordinary least squares model to fit the NDVI\u003csub\u003emax\u003c/sub\u003e trends for the five years prior to abandonment (cropland) and the five years following abandonment, and compared the NDVI\u003csub\u003emax\u003c/sub\u003e performances of the abandonment year. If a significant decrease (Mutation point, shown in Fig. 1g) in NDVI\u003csub\u003emax\u003c/sub\u003e was observed in the abandonment year, we identified it as abandoned cropland. In contrast, if the NDVI\u003csub\u003emax\u003c/sub\u003e showed an increase, it was recognized as afforestation (Fig. 1g). This integrated method effectively distinguishes between abandonment and afforestation in the post-agriculture period, yielding convincing results in both subtropical regions and arid/semi-arid areas (Fig. S1).\u003c/p\u003e\n\u003cp\u003eBased on the above criteria, we generate three sets of abandoned cropland datasets, with cropland converted to natural vegetation, from the CULD-A, CLCD, and CACD datasets, respectively. Notably, for the CACD dataset, we incorporated urban impervious surface data from the GALA dataset\u003csup\u003e77\u003c/sup\u003e and permanent surface water data\u003csup\u003e78\u003c/sup\u003e to ensure the consistency of the output results. We employ a sampling-based method to assess the accuracy of abandoned cropland maps derived from three land data sources (Fig. S1). To determine the sample size for validation, we applied a disproportionate stratified sampling strategy, which is well-suited for evaluating the classification accuracy of small categories, such as abandoned cropland\u003csup\u003e79\u003c/sup\u003e. For both abandoned and non-abandoned cropland categories, we randomly selected 1,080 samples from each category, totaling 2,160 pixels (see sample locations in Fig. S1). These reference samples were visually interpreted using very-high-resolution (VHR) historical images from Google Earth, with time-series NDVI\u003csub\u003emax\u003c/sub\u003e data from Google Earth Engine assisting in sample identification\u003csup\u003e11\u003c/sup\u003e. We compared the overall accuracy of the three abandoned cropland maps and ultimately selected the map with the highest accuracy, derived from the CACD dataset\u003csup\u003e72\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAssembling a carbon database and standardizing data\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe systematically reviewed literature from the Web of Science and China National Knowledge Infrastructure databases (28 July 2024), using the keywords: (abandoned cropland OR cropland abandonment OR land-cover change) AND (carbon sequestration OR biomass carbon OR SOC OR carbon accumulation rate) AND (China) AND (afforestation OR natural regeneration OR secondary succession OR plantation). We included \u0026ldquo;Grain-for-Green\u0026rdquo; as a keyword because this program represents a large-scale, purpose-driven transition of cropland to forests in China\u003csup\u003e21\u003c/sup\u003e. In term of SOC or biomass carbon changes induced by land-cover change, we considered only post-agricultural land transitions.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eHere, we initially retrieved 3,426 peer-reviewed studies, later expanding the total to 3,713 entries. We reviewed all abstracts to identify accessible studies that described pathways for transitioning cropland to natural vegetation, including abandonment and afforestation, and identified research that quantified soil organic carbon (SOC) or biomass carbon stocks. To avoid duplicate measurements, we prioritized original studies, reviews\u003csup\u003e12,39,80\u0026ndash;82\u003c/sup\u003e and the Forest Carbon Database (ForC)\u003csup\u003e83\u003c/sup\u003e. For these, we accessed the original studies to verify figures, correct errors, and extract additional variables. Notably, this study focuses exclusively on vegetation growth following cropland transitions. Natural regeneration after forest fires, logging, or pest outbreaks was not considered\u003csup\u003e39\u003c/sup\u003e.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo be included in our database, studies needed to meet the following criteria: (1) The prior land cover of the area undergoing abandonment (natural regeneration) or afforestation must have been cropland; (2) Geographic coordinates (latitude and longitude) or identifiable locations (e.g., recognizable place names) must be provided; (3) Empirical measurements of carbon (or biomass) in aboveground or belowground vegetation, litter, coarse woody debris, and/or soil pools must be reported. In addition, the database includes critical information such as the duration of abandonment (years), types of carbon pools, and estimates of carbon or biomass stocks (Mg ha⁻\u0026sup1;). The resulting dataset includes 7,316 empirical measurements of carbon stocks in aboveground and belowground biomass, soil, litter, and coarse woody debris. For studies reporting biomass only, we converted the values to carbon (Mg C ha⁻\u0026sup1;) using default conversion factors: 0.47 for aboveground and belowground carbon pools (in addition to the available and reference conversion rate from original paper), 0.37 for litter biomass, and 0.50 for coarse woody debris biomass\u003csup\u003e39\u003c/sup\u003e. These conversion ratios were not applied to grassland-stage biomass estimations, as research indicates a higher proportion of belowground biomass in abandoned cropland, particularly in arid and semi-arid regions. Consequently, we excluded cases where only aboveground biomass was considered for abandoned croplands. This resulted in 684 independent plot measurements for total vegetation carbon. For dead carbon pools (litter and coarse woody debris), measurements often included additional pools, but we did not attempt to disaggregate litter or coarse woody debris from these combined measurements due to their high variability and site specificity\u003csup\u003e39\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eMost SOC data (79.7%; N=2,793/3,502) were originally reported in Mg C ha⁻\u0026sup1; per centimeter of depth. For the remaining cases, we converted SOC from concentrations (per 1,000 g of soil) or soil organic matter (SOM). For SOM concentration data (N = 709), we used conversion factors provided in the literature whenever available; otherwise, we estimated SOC concentrations as SOM/1.724\u003csup\u003e81\u003c/sup\u003e. SOC concentrations were converted to Mg C ha⁻\u0026sup1; per centimeter of depth using either empirically reported bulk density data (N = 735) or depth-specific bulk density data from SoilGrids\u003csup\u003e84\u003c/sup\u003e. SoilGrids provides bulk density models at 15 cm, 30 cm, and 60 cm depths, and we used the value closest to the depth of SOC concentration measurements. For soil data, we standardized measurements to a depth of 30 cm, as the effects of cropland abandonment on SOC are most pronounced in the surface layer\u003csup\u003e21\u003c/sup\u003e. For plots with multiple depth measurements, we used the slope of log-log curves fitted to cumulative SOC stocks as a function of depth to estimate SOC at the standard depth\u003csup\u003e85\u003c/sup\u003e. For plots without multiple depth measurements, biome-specific slope coefficients were applied\u003csup\u003e39\u003c/sup\u003e. For control plots, SOC stocks for cropland at the initial year of abandonment were recorded (N = 512). Throughout this analysis, we focused on the 2000 \u0026ndash; 2060 timeframe for cropland abandonment, as this period represents a critical biophysical and policy-relevant window for achieving net-zero emissions and mitigating the most severe impacts of global warming. The database also includes other valuable information, such as tree species and soil types.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSpatial prediction and modeling of carbon sequestration\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo develop spatial prediction models for carbon sequestration, we selected representative environmental covariate layers based on the conceptual framework of previous carbon accumulation models\u003csup\u003e39\u003c/sup\u003e. These layers included variables related to topography, climate, soil nutrients, soil chemistry, soil physical properties, and downscaled CO₂ and nitrogen deposition data\u003csup\u003e6,86\u003c/sup\u003e (Table S1). We first sampled these environmental covariate stacks at the locations of each point in the literature-derived dataset. Variables representing current vegetation conditions, such as leaf area index or forest cover, as well as satellite-derived indices like the normalized difference vegetation index (NDVI), were excluded, as they do not reflect the fundamental biophysical controls on future carbon accumulation rates in plant biomass\u003csup\u003e39\u003c/sup\u003e. The covariate map layers were resampled and reprojected to a uniform pixel grid at a resolution of 1 km (30 arc-seconds) using the World Geodetic System 1984 (EPSG: 4326). High-resolution data were downsampled using an average aggregation method, while low-resolution data were upsampled using a simple nearest-neighbor approach (i.e., without interpolation). This resolution was chosen to balance pixel-level uncertainty (which increases proportionally at smaller pixel sizes) with the finest resolution available for most covariates\u003csup\u003e87\u003c/sup\u003e. For multi-year covariate layers under non-future scenarios, we used mean values from 2000 \u0026ndash; 2020 to capture average conditions under current and historical climates since the 21st century. Temporal information on cropland abandonment was also included as a key variable for evaluating future carbon accumulation (replacing time with space). Additionally, management practices for abandoned cropland, including abandonment and afforestation\u0026mdash;were incorporated as categorical covariates in the models.\u003c/p\u003e\n\u003cp\u003eWe constructed separate models (model pipelines) for SOC and biomass carbon. After testing several models, including SVR and XGBoost, the Random Forest algorithm with high performance, was selected to estimate temporal trends in future carbon sequestration. This choice was due to its robustness in extrapolating empirical data to large-scale datasets, insensitivity to feature scales, and reduced susceptibility to overfitting\u003csup\u003e39\u003c/sup\u003e. Using Python\u0026rsquo;s scikit-learn package, we defined models and pipelines for both SOC and biomass carbon. The total dataset was randomly split into training (70%) and testing (30%) subsets, with separate pipelines for each. We determined the optimal number of decision trees in increments of 50 and selected the best feature selection method and machine learning algorithm with defined hyperparameters using triple cross-validation with root mean square error (RMSE) as the performance metric (Fig. S5). Cross-validation creates pseudo-training datasets to estimate out-of-sample error, mitigating overfitting to the training data\u003csup\u003e88\u003c/sup\u003e. In triple cross-validation, the training dataset was randomly divided into three equally sized subsets. Two subsets were combined to form a new training set, while the remaining subset served as a validation set to assess model performance. The final results indicated that the optimal number of trees for the SOC and biomass carbon random forest models was 650 and 850, respectively, achieving the best fit accuracy (Fig. S5). The SOC and biomass carbon model pipelines achieved R\u0026sup2; values of 0.76 (RMSE = 11.00 Mg ha⁻\u0026sup1;) and 0.72 (RMSE = 20.97 Mg ha⁻\u0026sup1;), respectively, providing a robust foundation for further research (Fig. S6).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSimulation Design\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eHere, we consider two main pathways for abandoned land management: natural regeneration and afforestation. We do not account for additional scenarios that are spatially difficult to control, such as re-cultivation of fallow land or newly abandoned land, as their areas are sufficiently small in the short term to be neglected. The first pathway (scenario) involves allowing natural vegetation regeneration on abandoned land until 2060 without human intervention. In ours model, we adjusted the duration covariate to predict the future carbon sequestration potential. Note that we assume all biomass from abandoned cropland is removed initially\u003csup\u003e11\u003c/sup\u003e. The future changes in SOC are assessed based on the SOC density of nearby long-term cultivated lands.\u003c/p\u003e\n\u003cp\u003eThe second pathway (scenario) considers implementing afforestation programs on abandoned cropland. In this land management practice, we modeled afforestation efforts in alignment with the Grain-for-Green program, one of China\u0026rsquo;s major ecological restoration initiatives\u003csup\u003e5\u003c/sup\u003e. Specifically, the empirical samples of carbon changes for afforestation on cropland are primarily driven by this project. The spatial variability in carbon sequestration associated with these tree species can be addressed using empirical data, which includes species tailored to local conditions and indigenous knowledge, such as sea buckthorn, poplars, and walnut trees in arid and semi-arid regions\u003csup\u003e89\u003c/sup\u003e. To minimize estimation errors in carbon sequestration potential under afforestation scenarios, we calculated the average carbon sequestration for regions where multiple tree species were reported in the carbon database. This approach helps reduce uncertainty caused by variability among tree species. We used SOC stocks from nearby long-term croplands (N = 512) as a baseline and estimated future SOC stocks based on annual SOC change rates\u003csup\u003e39\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eIn the future simulation design, we incorporated climate data from the four Shared Socioeconomic Pathways (SSPs) of CMIP6 (SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5), along with CO₂ and nitrogen deposition data. Numerous studies have highlighted the critical importance of these variables for ecosystem carbon sinks\u003csup\u003e90\u003c/sup\u003e. We focused on carbon sequestration levels through 2060, as this year marks a key milestone in China\u0026rsquo;s commitment to carbon neutrality. Additionally, empirical data models are more robust within a 60-year timeframe.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAccounting for inputs and carbon returns after afforestation in abandoned cropland\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAccording to the Chinese government\u0026rsquo;s funding allocation plan for the new round of the Grain-for-Green program in 2024, a subsidy of 1,200 RMB (USD 169.00) per mu (1 mu = 1/15 ha) is provided for reforested cropland\u003csup\u003e91\u003c/sup\u003e. This subsidy includes 900 RMB (USD 126.76) for land compensation and 300 RMB (USD 42.24) for sapling support, with funding from the central government. For local government subsidies, detailed afforestation data were calculated at the provincial level for two periods (2000 \u0026ndash; 2010 and 2010 \u0026ndash; 2020) across 21 provinces in China. This data includes afforestation area, total investment, and provincial government contributions. By integrating investments from both the central and local governments and adjusting for inflation using the World Bank\u0026rsquo;s RMB inflation rate, we estimated the per-unit area investment for afforestation at the provincial level (Table S2). For provinces without quantified investment information, we used the average values from comparable geographic regions as a reference, given the similarities in the geographic and physical factors affecting land restoration.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo estimate the carbon sequestration benefits, we calculated the incremental revenue from afforestation on abandoned cropland based on the carbon trading value in China\u0026rsquo;s carbon market. Since 2021, China has established nine carbon trading markets\u003csup\u003e92\u003c/sup\u003e. Using daily trading data up to May 2024, including average transaction prices, trading volumes, and total amounts, we computed an average carbon trading revenue of USD 11.71 per ton. This represents a conservative estimate of carbon value, which may increase in the future\u003csup\u003e43\u003c/sup\u003e. In estimating the carbon sequestration of biomass and SOC under the afforestation scenario, we projected potential carbon revenue from abandoned cropland. We also referred to reports on the socio-economic and environmental benefits of the Grain-for-Green program over the past 20 years\u003csup\u003e8\u003c/sup\u003e. These reports provided comparative insights to support our carbon sequestration findings and other ecosystem service values, such as soil conservation and sand fixation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eUncertainty analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo evaluate the spatial uncertainty of the model regarding biomass carbon and SOC carbon sequestration, we applied a random 70% training subset of the best-performing step length to a new ensemble of 50 random forest models, trained on all covariates and carbon empirical points. Each random forest model was trained on independent pipeline, allowing us to estimate the model uncertainty for each pixel by calculating the standard deviation of predictions calculation\u003csup\u003e39\u003c/sup\u003e. The mean standard deviations of model uncertainty for simulated SOC sequestration under abandonment and afforestation scenarios were 3.03 Mg C ha⁻\u0026sup1; and 2.95 Mg C ha⁻\u0026sup1;, respectively. Similarly, the mean standard deviations of model uncertainty for biomass carbon sequestration were 4.65 Mg C ha⁻\u0026sup1; and 4.7 Mg C ha⁻\u0026sup1;, respectively. We observed significant spatial differences in the uncertainties of biomass carbon and SOC sequestration. The highest uncertainties for biomass carbon were concentrated in the southwestern regions, while uncertainties for SOC were predominantly observed in northeastern China, particularly in the Greater Khingan Mountains region (Fig. S7).\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe annual CLCD\u003csup\u003e71\u003c/sup\u003e, CLUD-A\u003csup\u003e70\u003c/sup\u003e, and CACD\u003csup\u003e72\u003c/sup\u003e maps of China were obtained from previous studies. The derived 30m resolution abandoned cropland map for China is available via 10.6084/m9.figshare.28386476. Provincial-scale data on afforestation areas and investment (including both national and local investments) in China from 2003 to 2020 can be accessed via 10.6084/m9.figshare.28386476. The daily transaction records from China\u0026apos;s nine carbon emission trading exchanges, including transaction prices, volumes, and price fluctuations (as of May 2024), can be freely accessed via 10.6084/m9.figshare.28386476. The raw carbon datasets for abandoned cropland are available via 10.6084/m9.figshare.28386476. Detailed descriptions of the environmental covariates can be found in Table S1.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCode availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe code, including JavaScript code for identifying abandoned land and Python code for machine learning, related to the key methods of this work, is available at 10.6084/m9.figshare.28386476.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work is supported by the National Key R\u0026amp;D Program of China (grant number: 2024YFF1307600)\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eZeng, Y. \u003cem\u003eet al.\u003c/em\u003e Economic and social constraints on reforestation for climate mitigation in Southeast Asia. \u003cem\u003eNat. Clim. Chang.\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 842\u0026ndash;844 (2020).\u003c/li\u003e\n\u003cli\u003eAustin, K. G. \u003cem\u003eet al.\u003c/em\u003e The economic costs of planting, preserving, and managing the world\u0026rsquo;s forests to mitigate climate change. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 5946 (2020).\u003c/li\u003e\n\u003cli\u003eMa, Y., Huang, L., Li, J., Cao, W. \u0026amp; Cai, Y. Carbon potential of China\u0026rsquo;s Grain to Green Program and its contribution to the carbon target. \u003cem\u003eResources, Conservation and Recycling\u003c/em\u003e \u003cstrong\u003e200\u003c/strong\u003e, 107272 (2024).\u003c/li\u003e\n\u003cli\u003eDeng, L., Liu, G. \u0026amp; Shangguan, Z. Land‐use conversion and changing soil carbon stocks in C hina\u0026rsquo;s \u0026lsquo;Grain‐for‐Green\u0026rsquo; Program: a synthesis. \u003cem\u003eGlobal Change Biology\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 3544\u0026ndash;3556 (2014).\u003c/li\u003e\n\u003cli\u003ePersson, M., Moberg, J., Ostwald, M. \u0026amp; Xu, J. The Chinese Grain for Green Programme: Assessing the carbon sequestered via land reform. \u003cem\u003eJournal of Environmental Management\u003c/em\u003e \u003cstrong\u003e126\u003c/strong\u003e, 142\u0026ndash;146 (2013).\u003c/li\u003e\n\u003cli\u003eLiao, Z. \u003cem\u003eet al.\u003c/em\u003e Growing biomass carbon stock in China driven by expansion and conservation of woody areas. \u003cem\u003eNat. Geosci.\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 1127\u0026ndash;1134 (2024).\u003c/li\u003e\n\u003cli\u003eDeng, L. \u003cem\u003eet al.\u003c/em\u003e Past and future carbon sequestration benefits of China\u0026rsquo;s grain for green program. \u003cem\u003eGlobal Environmental Change\u003c/em\u003e \u003cstrong\u003e47\u003c/strong\u003e, 13\u0026ndash;20 (2017).\u003c/li\u003e\n\u003cli\u003eGOV. Comprehensive Benefit Assessment Report of China\u0026rsquo;s Grain for Green Project. (2022).\u003c/li\u003e\n\u003cli\u003eYao, L. \u003cem\u003eet al.\u003c/em\u003e Carbon sequestration potential of tree planting in China. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 8398 (2024).\u003c/li\u003e\n\u003cli\u003eSchierhorn, F. \u003cem\u003eet al.\u003c/em\u003e Post‐Soviet cropland abandonment and carbon sequestration in European Russia, Ukraine, and Belarus. \u003cem\u003eGlobal Biogeochemical Cycles\u003c/em\u003e \u003cstrong\u003e27\u003c/strong\u003e, 1175\u0026ndash;1185 (2013).\u003c/li\u003e\n\u003cli\u003eZheng, Q. \u003cem\u003eet al.\u003c/em\u003e The neglected role of abandoned cropland in supporting both food security and climate change mitigation. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 6083 (2023).\u003c/li\u003e\n\u003cli\u003eTian, D. \u003cem\u003eet al.\u003c/em\u003e Cropland abandonment enhances soil inorganic nitrogen retention and carbon stock in China : A meta‐analysis. \u003cem\u003eLand Degrad Dev\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e, 3898\u0026ndash;3906 (2018).\u003c/li\u003e\n\u003cli\u003eGvein, M. H. \u003cem\u003eet al.\u003c/em\u003e Potential of land-based climate change mitigation strategies on abandoned cropland. \u003cem\u003eCommun Earth Environ\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 39 (2023).\u003c/li\u003e\n\u003cli\u003eLi, M., Cui, Y., Dong, J. \u0026amp; Qin, Y. Abandoned cropland compensates the decrease in net ecosystem productivity of impervious surface expansion in China. \u003cem\u003eEnvironmental Impact Assessment Review\u003c/em\u003e \u003cstrong\u003e104\u003c/strong\u003e, 107363 (2024).\u003c/li\u003e\n\u003cli\u003eCrawford, C. L., Yin, H., Radeloff, V. C. \u0026amp; Wilcove, D. S. Rural land abandonment is too ephemeral to provide major benefits for biodiversity and climate. \u003cem\u003eSci. Adv.\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, eabm8999 (2022).\u003c/li\u003e\n\u003cli\u003eLeal Filho, W., Mandel, M., Al-Amin, A. Q., Feher, A. \u0026amp; Chiappetta Jabbour, C. J. An assessment of the causes and consequences of agricultural land abandonment in Europe. \u003cem\u003eInternational Journal of Sustainable Development \u0026amp; World Ecology\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e, 554\u0026ndash;560 (2017).\u003c/li\u003e\n\u003cli\u003eZhou, T., Koomen, E. \u0026amp; Ke, X. Determinants of Farmland Abandonment on the Urban\u0026ndash;Rural Fringe. \u003cem\u003eEnvironmental Management\u003c/em\u003e \u003cstrong\u003e65\u003c/strong\u003e, 369\u0026ndash;384 (2020).\u003c/li\u003e\n\u003cli\u003eUstaoglu, E. \u0026amp; J. Collier, M. Farmland abandonment in Europe: an overview of drivers, consequences, and assessment of the sustainability implications. \u003cem\u003eEnvironmental Reviews\u003c/em\u003e \u003cstrong\u003e26\u003c/strong\u003e, 396\u0026ndash;416 (2018).\u003c/li\u003e\n\u003cli\u003eHong, C., Prishchepov, A. V., Jin, X. \u0026amp; Zhou, Y. Mapping cropland abandonment and distinguishing from intentional afforestation with Landsat time series. \u003cem\u003eInternational Journal of Applied Earth Observation and Geoinformation\u003c/em\u003e \u003cstrong\u003e127\u003c/strong\u003e, 103693 (2024).\u003c/li\u003e\n\u003cli\u003eDu, Z., Yang, J., Ou, C. \u0026amp; Zhang, T. Agricultural Land Abandonment and Retirement Mapping in the Northern China Crop-Pasture Band Using Temporal Consistency Check and Trajectory-Based Change Detection Approach. \u003cem\u003eIEEE Trans. Geosci. Remote Sensing\u003c/em\u003e \u003cstrong\u003e60\u003c/strong\u003e, 1\u0026ndash;12 (2022).\u003c/li\u003e\n\u003cli\u003eDeng, L., Liu, G. \u0026amp; Shangguan, Z. Land-use conversion and changing soil carbon stocks in China\u0026rsquo;s \u0026lsquo;Grain-for-Green\u0026rsquo; Program: a synthesis. \u003cem\u003eGlobal Change Biology\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 3544\u0026ndash;3556 (2014).\u003c/li\u003e\n\u003cli\u003eCramer, V., Hobbs, R. \u0026amp; Standish, R. What\u0026rsquo;s new about old fields? Land abandonment and ecosystem assembly. \u003cem\u003eTrends in Ecology \u0026amp; Evolution\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 104\u0026ndash;112 (2008).\u003c/li\u003e\n\u003cli\u003eBell, S. M., Barriocanal, C., Terrer, C. \u0026amp; Rosell-Mel\u0026eacute;, A. Management opportunities for soil carbon sequestration following agricultural land abandonment. \u003cem\u003eEnvironmental Science \u0026amp; Policy\u003c/em\u003e \u003cstrong\u003e108\u003c/strong\u003e, 104\u0026ndash;111 (2020).\u003c/li\u003e\n\u003cli\u003eLasanta, T., Nadal-Romero, E. \u0026amp; Arn\u0026aacute;ez, J. Managing abandoned farmland to control the impact of re-vegetation on the environment. The state of the art in Europe. \u003cem\u003eEnvironmental Science \u0026amp; Policy\u003c/em\u003e \u003cstrong\u003e52\u003c/strong\u003e, 99\u0026ndash;109 (2015).\u003c/li\u003e\n\u003cli\u003eWang, L., Pedersen, P. B. M. \u0026amp; Svenning, J.-C. Rewilding abandoned farmland has greater sustainability benefits than afforestation. \u003cem\u003enpj biodivers\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 5 (2023).\u003c/li\u003e\n\u003cli\u003eJanus, J. \u0026amp; Bozek, P. Using ALS data to estimate afforestation and secondary forest succession on agricultural areas: An approach to improve the understanding of land abandonment causes. \u003cem\u003eApplied Geography\u003c/em\u003e \u003cstrong\u003e97\u003c/strong\u003e, 128\u0026ndash;141 (2018).\u003c/li\u003e\n\u003cli\u003eYang, Y. \u003cem\u003eet al.\u003c/em\u003e Restoring Abandoned Farmland to Mitigate Climate Change on a Full Earth. \u003cem\u003eOne Earth\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 176\u0026ndash;186 (2020).\u003c/li\u003e\n\u003cli\u003eYang, Y. \u003cem\u003eet al.\u003c/em\u003e Restoring Abandoned Farmland to Mitigate Climate Change on a Full Earth. \u003cem\u003eOne Earth\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 176\u0026ndash;186 (2020).\u003c/li\u003e\n\u003cli\u003eCrawford, C. L., Wiebe, R. A., Yin, H., Radeloff, V. C. \u0026amp; Wilcove, D. S. Biodiversity consequences of cropland abandonment. \u003cem\u003eNat Sustain\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 1596\u0026ndash;1607 (2024).\u003c/li\u003e\n\u003cli\u003eAnpilogova, D. \u0026amp; Pakina, A. Assessing ecosystem services of abandoned agricultural lands: a case study in the forested zone of European Russia. \u003cem\u003eOE\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, e77969 (2022).\u003c/li\u003e\n\u003cli\u003eMARAPPR. Guidance on the integrated utilization of abandoned cropland for the development of agricultural production. (2021).\u003c/li\u003e\n\u003cli\u003eZhu, X., Xiao, G., Zhang, D. \u0026amp; Guo, L. Mapping abandoned farmland in China using time series MODIS NDVI. \u003cem\u003eScience of The Total Environment\u003c/em\u003e \u003cstrong\u003e755\u003c/strong\u003e, 142651 (2021).\u003c/li\u003e\n\u003cli\u003eWang, H. \u003cem\u003eet al.\u003c/em\u003e Is Abandoned Cropland Continuously Growing in China? Quantitative Evidence and Enlightenment from Landsat-Derived Annual China Land Cover Dataset. \u003cem\u003eLand\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 45 (2023).\u003c/li\u003e\n\u003cli\u003eShi, T., Li, X., Xin, L. \u0026amp; Xu, X. The spatial distribution of farmland abandonment and its influential factors at the township level: A case study in the mountainous area of China. \u003cem\u003eLand Use Policy\u003c/em\u003e \u003cstrong\u003e70\u003c/strong\u003e, 510\u0026ndash;520 (2018).\u003c/li\u003e\n\u003cli\u003eMARA. List of mountainous and hilly counties in China. (2019).\u003c/li\u003e\n\u003cli\u003eCao, B. \u003cem\u003eet al.\u003c/em\u003e A 30 m terrace mapping in China using Landsat 8 imagery and digital elevation model based on the Google Earth Engine. \u003cem\u003eEarth Syst. Sci. Data\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 2437\u0026ndash;2456 (2021).\u003c/li\u003e\n\u003cli\u003eHong, S. \u003cem\u003eet al.\u003c/em\u003e Divergent responses of soil organic carbon to afforestation. \u003cem\u003eNat Sustain\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 694\u0026ndash;700 (2020).\u003c/li\u003e\n\u003cli\u003eVesterdal, L., Ritter, E. \u0026amp; Gundersen, P. Change in soil organic carbon following afforestation of former arable land. \u003cem\u003eForest Ecology and Management\u003c/em\u003e \u003cstrong\u003e169\u003c/strong\u003e, 137\u0026ndash;147 (2002).\u003c/li\u003e\n\u003cli\u003eCook-Patton, S. C. \u003cem\u003eet al.\u003c/em\u003e Mapping carbon accumulation potential from global natural forest regrowth. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e585\u003c/strong\u003e, 545\u0026ndash;550 (2020).\u003c/li\u003e\n\u003cli\u003eKristensen, J. \u0026Aring;. \u003cem\u003eet al.\u003c/em\u003e Tree planting is no climate solution at northern high latitudes. \u003cem\u003eNat. Geosci.\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 1087\u0026ndash;1092 (2024).\u003c/li\u003e\n\u003cli\u003eLiu, T. \u003cem\u003eet al.\u003c/em\u003e Environmental laws and ecological restoration projects enhancing ecosystem services in China: A meta-analysis. \u003cem\u003eJournal of Environmental Management\u003c/em\u003e \u003cstrong\u003e327\u003c/strong\u003e, 116810 (2023).\u003c/li\u003e\n\u003cli\u003eChen, X. \u003cem\u003eet al.\u003c/em\u003e Unraveling carbon stock dynamics and their determinants in China\u0026rsquo;s Loess Plateau over the past 40 years. \u003cem\u003eEcological Indicators\u003c/em\u003e \u003cstrong\u003e159\u003c/strong\u003e, 111760 (2024).\u003c/li\u003e\n\u003cli\u003eQi, S., Cheng, S., Tan, X., Feng, S. \u0026amp; Zhou, Q. Predicting China\u0026rsquo;s carbon price based on a multi-scale integrated model. \u003cem\u003eApplied Energy\u003c/em\u003e \u003cstrong\u003e324\u003c/strong\u003e, 119784 (2022).\u003c/li\u003e\n\u003cli\u003eZhao, N., Chen, K., Wu, X., Zhang, L. \u0026amp; Wang, W. Cropland fragmentation change across China over the last two decades. \u003cem\u003eAgricultural Systems\u003c/em\u003e \u003cstrong\u003e218\u003c/strong\u003e, 104010 (2024).\u003c/li\u003e\n\u003cli\u003eDeng, O. \u003cem\u003eet al.\u003c/em\u003e Managing fragmented croplands for environmental and economic benefits in China. \u003cem\u003eNat Food\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 230\u0026ndash;240 (2024).\u003c/li\u003e\n\u003cli\u003eZhang, Q. \u003cem\u003eet al.\u003c/em\u003e Outlook of China\u0026rsquo;s agriculture transforming from smallholder operation to sustainable production. \u003cem\u003eGlobal Food Security\u003c/em\u003e \u003cstrong\u003e26\u003c/strong\u003e, 100444 (2020).\u003c/li\u003e\n\u003cli\u003eZhang, M. \u003cem\u003eet al.\u003c/em\u003e Reveal the severe spatial and temporal patterns of abandoned cropland in China over the past 30 years. \u003cem\u003eScience of The Total Environment\u003c/em\u003e \u003cstrong\u003e857\u003c/strong\u003e, 159591 (2023).\u003c/li\u003e\n\u003cli\u003eHe, T. \u003cem\u003eet al.\u003c/em\u003e Quantitative analysis of abandonment and grain production loss under armed conflict in Ukraine. \u003cem\u003eJournal of Cleaner Production\u003c/em\u003e \u003cstrong\u003e412\u003c/strong\u003e, 137367 (2023).\u003c/li\u003e\n\u003cli\u003eYu, Z. \u003cem\u003eet al.\u003c/em\u003e Forest expansion dominates China\u0026rsquo;s land carbon sink since 1980. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 5374 (2022).\u003c/li\u003e\n\u003cli\u003eTong, X. \u003cem\u003eet al.\u003c/em\u003e Forest management in southern China generates short term extensive carbon sequestration. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 129 (2020).\u003c/li\u003e\n\u003cli\u003eCheng, K. \u003cem\u003eet al.\u003c/em\u003e Carbon storage through China\u0026rsquo;s planted forest expansion. \u003cem\u003eNat Commun\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 4106 (2024).\u003c/li\u003e\n\u003cli\u003eChen, Y. \u003cem\u003eet al.\u003c/em\u003e Reforestation makes a minor contribution to soil carbon accumulation in the short term: Evidence from four subtropical plantations. \u003cem\u003eForest Ecology and Management\u003c/em\u003e \u003cstrong\u003e384\u003c/strong\u003e, 400\u0026ndash;405 (2017).\u003c/li\u003e\n\u003cli\u003eYun, H. \u003cem\u003eet al.\u003c/em\u003e Changes in above- versus belowground biomass distribution in permafrost regions in response to climate warming. \u003cem\u003eProc. Natl. Acad. Sci. U.S.A.\u003c/em\u003e \u003cstrong\u003e121\u003c/strong\u003e, e2314036121 (2024).\u003c/li\u003e\n\u003cli\u003eGvein, M. H. Potential of land-based climate change mitigation strategies on abandoned cropland.\u003c/li\u003e\n\u003cli\u003eN\u0026aelig;ss, J. S., Cavalett, O. \u0026amp; Cherubini, F. The land\u0026ndash;energy\u0026ndash;water nexus of global bioenergy potentials from abandoned cropland. \u003cem\u003eNat Sustain\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 525\u0026ndash;536 (2021).\u003c/li\u003e\n\u003cli\u003eGriscom, B. W. \u003cem\u003eet al.\u003c/em\u003e Natural climate solutions. \u003cem\u003eProc. Natl. Acad. Sci. U.S.A.\u003c/em\u003e \u003cstrong\u003e114\u003c/strong\u003e, 11645\u0026ndash;11650 (2017).\u003c/li\u003e\n\u003cli\u003eMeyfroidt, P., Schierhorn, F., Prishchepov, A. V., M\u0026uuml;ller, D. \u0026amp; Kuemmerle, T. Drivers, constraints and trade-offs associated with recultivating abandoned cropland in Russia, Ukraine and Kazakhstan. \u003cem\u003eGlobal Environmental Change\u003c/em\u003e \u003cstrong\u003e37\u003c/strong\u003e, 1\u0026ndash;15 (2016).\u003c/li\u003e\n\u003cli\u003eAffek, A. N., Wolski, J., Latocha, A., Zachwatowicz, M. \u0026amp; Wieczorek, M. The use of LiDAR in reconstructing the pre‐World War II landscapes of abandoned mountain villages in southern Poland. \u003cem\u003eArchaeological Prospection\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e, 157\u0026ndash;173 (2022).\u003c/li\u003e\n\u003cli\u003eNguyen, H. D. \u003cem\u003eet al.\u003c/em\u003e Cropland abandonment and flood risks: Spatial analysis of a case in North Central Vietnam. \u003cem\u003eAnthropocene\u003c/em\u003e \u003cstrong\u003e38\u003c/strong\u003e, 100341 (2022).\u003c/li\u003e\n\u003cli\u003eGarc\u0026iacute;a-Ruiz, J. M. \u0026amp; Lana-Renault, N. Hydrological and erosive consequences of farmland abandonment in Europe, with special reference to the Mediterranean region \u0026ndash; A review. \u003cem\u003eAgriculture, Ecosystems \u0026amp; Environment\u003c/em\u003e \u003cstrong\u003e140\u003c/strong\u003e, 317\u0026ndash;338 (2011).\u003c/li\u003e\n\u003cli\u003eLiang, Y., Li, S., Feldman, M. W. \u0026amp; Daily, G. C. Does household composition matter? The impact of the Grain for Green Program on rural livelihoods in China. \u003cem\u003eEcological Economics\u003c/em\u003e \u003cstrong\u003e75\u003c/strong\u003e, 152\u0026ndash;160 (2012).\u003c/li\u003e\n\u003cli\u003eWang, K. \u003cem\u003eet al.\u003c/em\u003e Impact of the Grain for Green Project on the Well-Being of Farmer Households: A Case Study of the Mountainous Areas of Northern Hebei Province, China. \u003cem\u003eLand\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 1257 (2023).\u003c/li\u003e\n\u003cli\u003eStrassburg, B. B. N. \u003cem\u003eet al.\u003c/em\u003e Global priority areas for ecosystem restoration. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e586\u003c/strong\u003e, 724\u0026ndash;729 (2020).\u003c/li\u003e\n\u003cli\u003eDuan, J. \u003cem\u003eet al.\u003c/em\u003e Consolidation of agricultural land can contribute to agricultural sustainability in China. \u003cem\u003eNat Food\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 1014\u0026ndash;1022 (2021).\u003c/li\u003e\n\u003cli\u003eRen, C. \u003cem\u003eet al.\u003c/em\u003e Ageing threatens sustainability of smallholder farming in China. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e616\u003c/strong\u003e, 96\u0026ndash;103 (2023).\u003c/li\u003e\n\u003cli\u003eDoelman, J. C., Stehfest, E., Tabeau, A. \u0026amp; Van Meijl, H. Making the Paris agreement climate targets consistent with food security objectives. \u003cem\u003eGlobal Food Security\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 93\u0026ndash;103 (2019).\u003c/li\u003e\n\u003cli\u003eWu, X. \u003cem\u003eet al.\u003c/em\u003e The potential role of abandoned cropland for food security in China. \u003cem\u003eResources, Conservation and Recycling\u003c/em\u003e \u003cstrong\u003e212\u003c/strong\u003e, 108004 (2025).\u003c/li\u003e\n\u003cli\u003ePeng, S. \u003cem\u003eet al.\u003c/em\u003e Carbon restoration potential on global land under water resource constraints. \u003cem\u003eNat Water\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 1071\u0026ndash;1081 (2024).\u003c/li\u003e\n\u003cli\u003eZan, B. \u003cem\u003eet al.\u003c/em\u003e Spatiotemporal inequality in land water availability amplified by global tree restoration. \u003cem\u003eNat Water\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 863\u0026ndash;874 (2024).\u003c/li\u003e\n\u003cli\u003eXu, Y. \u003cem\u003eet al.\u003c/em\u003e Annual 30-m land use/land cover maps of China for 1980\u0026ndash;2015 from the integration of AVHRR, MODIS and Landsat data using the BFAST algorithm. \u003cem\u003eSci. China Earth Sci.\u003c/em\u003e \u003cstrong\u003e63\u003c/strong\u003e, 1390\u0026ndash;1407 (2020).\u003c/li\u003e\n\u003cli\u003eYang, J. \u0026amp; Huang, X. The 30 m annual land cover dataset and its dynamics in China from 1990 to 2019. \u003cem\u003eEarth Syst. Sci. Data\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 3907\u0026ndash;3925 (2021).\u003c/li\u003e\n\u003cli\u003eTu, Y. \u003cem\u003eet al.\u003c/em\u003e A 30 m annual cropland dataset of China from 1986 to 2021. \u003cem\u003eEarth Syst. Sci. Data\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 2297\u0026ndash;2316 (2024).\u003c/li\u003e\n\u003cli\u003eFAO. Methods \u0026amp; Standards. (2016).\u003c/li\u003e\n\u003cli\u003eDeng, L., Liu, G. \u0026amp; Shangguan, Z. Land-use conversion and changing soil carbon stocks in China\u0026rsquo;s \u0026lsquo;Grain-for-Green\u0026rsquo; Program: a synthesis. \u003cem\u003eGlobal Change Biology\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 3544\u0026ndash;3556 (2014).\u003c/li\u003e\n\u003cli\u003eGOV. Comprehensive Benefit Assessment Report of China\u0026rsquo;s Grain for Green Project. (2022).\u003c/li\u003e\n\u003cli\u003eYan, Y. \u003cem\u003eet al.\u003c/em\u003e Climate-induced tree-mortality pulses are obscured by broad-scale and long-term greening. \u003cem\u003eNat Ecol Evol\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 912\u0026ndash;923 (2024).\u003c/li\u003e\n\u003cli\u003eLi, X. \u003cem\u003eet al.\u003c/em\u003e Mapping global urban boundaries from the global artificial impervious area (GAIA) data. \u003cem\u003eEnviron. Res. Lett.\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 094044 (2020).\u003c/li\u003e\n\u003cli\u003ePekel, J.-F., Cottam, A., Gorelick, N. \u0026amp; Belward, A. S. High-resolution mapping of global surface water and its long-term changes. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e540\u003c/strong\u003e, 418\u0026ndash;422 (2016).\u003c/li\u003e\n\u003cli\u003eOlofsson, P. \u003cem\u003eet al.\u003c/em\u003e Good practices for estimating area and assessing accuracy of land change. \u003cem\u003eRemote Sensing of Environment\u003c/em\u003e \u003cstrong\u003e148\u003c/strong\u003e, 42\u0026ndash;57 (2014).\u003c/li\u003e\n\u003cli\u003eSong, X., Peng, C., Zhou, G., Jiang, H. \u0026amp; Wang, W. Chinese Grain for Green Program led to highly increased soil organic carbon levels: A meta-analysis. \u003cem\u003eSci Rep\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 4460 (2014).\u003c/li\u003e\n\u003cli\u003eChen, S. \u003cem\u003eet al.\u003c/em\u003e GSOCS-LULCC: the Global Soil Organic Carbon Stock dataset after Land Use and Land Cover Change. Preprint at https://doi.org/10.5194/essd-2024-373 (2024).\u003c/li\u003e\n\u003cli\u003eDeng, L., Liu, G. \u0026amp; Shangguan, Z. Carbon storage in successional and plantation forest soils: a tropical analysis. \u003cem\u003eGlobal Change Biology\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 3544\u0026ndash;3556 (2014).\u003c/li\u003e\n\u003cli\u003eMar\u0026iacute;n‐Spiotta, E. \u0026amp; Sharma, S. Carbon storage in successional and plantation forest soils: a tropical analysis. \u003cem\u003eGlobal Ecology and Biogeography\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 105\u0026ndash;117 (2013).\u003c/li\u003e\n\u003cli\u003eHengl, T. \u003cem\u003eet al.\u003c/em\u003e SoilGrids250m: Global gridded soil information based on machine learning. \u003cem\u003ePLoS ONE\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, e0169748 (2017).\u003c/li\u003e\n\u003cli\u003eWang, S. \u003cem\u003eet al.\u003c/em\u003e Vertical Distribution of Soil Organic Carbon in China. \u003cem\u003eEnvironmental Management\u003c/em\u003e \u003cstrong\u003e33\u003c/strong\u003e, (2004).\u003c/li\u003e\n\u003cli\u003eYang, X. \u003cem\u003eet al.\u003c/em\u003e Nitrogen addition promotes soil carbon accumulation globally. \u003cem\u003eSci. China Life Sci.\u003c/em\u003e (2024) doi:10.1007/s11427-024-2752-2.\u003c/li\u003e\n\u003cli\u003eLiu, T. \u003cem\u003eet al.\u003c/em\u003e Utilizing nighttime light datasets to uncover the spatial patterns of county-level relative poverty-returning risk in China and its alleviating factors. \u003cem\u003eJournal of Cleaner Production\u003c/em\u003e \u003cstrong\u003e448\u003c/strong\u003e, 141682 (2024).\u003c/li\u003e\n\u003cli\u003eYates, L. A., Aandahl, Z., Richards, S. A. \u0026amp; Brook, B. W. Cross validation for model selection: A review with examples from ecology. \u003cem\u003eEcological Monographs\u003c/em\u003e \u003cstrong\u003e93\u003c/strong\u003e, e1557 (2023).\u003c/li\u003e\n\u003cli\u003eYang, K. \u003cem\u003eet al.\u003c/em\u003e Seabuckthorn (Hippophae rhamnoides L.) plantation degradation aggravates microbial metabolic C and P limitations on the Northern Loess Plateau in China. \u003cem\u003eScience of The Total Environment\u003c/em\u003e \u003cstrong\u003e945\u003c/strong\u003e, 174088 (2024).\u003c/li\u003e\n\u003cli\u003eWang, J. \u003cem\u003eet al.\u003c/em\u003e Large Chinese land carbon sink estimated from atmospheric carbon dioxide data. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e586\u003c/strong\u003e, 720\u0026ndash;723 (2020).\u003c/li\u003e\n\u003cli\u003eGOV. Management of Forestry and Grassland Reform and Development Funds. (2024).\u003c/li\u003e\n\u003cli\u003eWeng, Q. \u0026amp; Xu, H. A review of China\u0026rsquo;s carbon trading market. \u003cem\u003eRenewable and Sustainable Energy Reviews\u003c/em\u003e \u003cstrong\u003e91\u003c/strong\u003e, 613\u0026ndash;619 (2018).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Abandoned cropland, afforestation, carbon sequestration, investment and carbon benefits, China","lastPublishedDoi":"10.21203/rs.3.rs-6119575/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6119575/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eAfforestation of abandoned cropland represents a promising strategy for land-based climate change mitigation, particularly in regions where land resources for additional afforestation are limited. However, the carbon sequestration potential of such land remains largely unknown. Here, we assess the spatial distribution of abandoned cropland in China and its carbon sequestration potential through afforestation incentives, using 10,818 carbon empirical data derived from 298 peer-reviewed articles, multisource remote sensing data, and machine learning models. We identify 6.03 Mha of abandoned cropland in China that have been undergoing natural regeneration since the early 21st century. This land has the potential to sequester an additional 215.12\u0026ndash;218.94 Tg of biomass carbon and 15.87\u0026ndash;17.64 Tg of soil organic carbon (SOC) through afforestation by 2060, representing a 51.95\u0026ndash;53.94% increase compared to natural regeneration alone. Our results further show that the carbon benefits from afforestation could offset 47.71\u0026ndash;49.57% of government investments (approximately USD 16.254\u0026nbsp;billion) in abandoned cropland. Our findings highlight the significant potential of afforestation on abandoned cropland to support China\u0026rsquo;s carbon neutrality goals, while also offering a cost-benefit framework to guide land policy decisions.\u003c/p\u003e","manuscriptTitle":"Afforestation on Abandoned Croplands in China Has the Potential to Increase Carbon Sequestration by half","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-03-04 09:07:43","doi":"10.21203/rs.3.rs-6119575/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"c2f68946-52e4-4ad6-88ec-c8d5ccec89ab","owner":[],"postedDate":"March 4th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":45129512,"name":"Earth and environmental sciences/Ecology/Ecosystem services"},{"id":45129513,"name":"Earth and environmental sciences/Environmental sciences/Environmental impact"},{"id":45129514,"name":"Earth and environmental sciences/Ecology/Ecosystem ecology"},{"id":45129515,"name":"Earth and environmental sciences/Ecology/Restoration ecology"}],"tags":[],"updatedAt":"2025-04-16T16:41:33+00:00","versionOfRecord":[],"versionCreatedAt":"2025-03-04 09:07:43","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6119575","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6119575","identity":"rs-6119575","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}