Future global land datasets with a 1-km resolution based on the SSP-RCP scenarios

The continuous strengthening of human activities and intensification of climate change have jointly affected global land cover changes. Scenario-based land change simulation has been widely used as a powerful tool that can assess the impact of potential land changes. However, the current future land projection products at a global scale generally suffer from deficiencies such as low resolution, unsuitable scenarios, and few types of land. Therefore, we propose a series global land projection datasets with a resolution of 1-km from 2015 to 2100, which uses the latest IPCC coupling scenarios of socioeconomic and climate change, SSP-RCP.<br> Actually, there are two subdatasets: 1) the 1-km global LULC dataset containing seven land types; 2) the 1-km global PFT-based land dataset containing 20 land types. The production of the LULC dataset combines the top-down land demand constraints provided by the official dataset of CMIP6 and the bottom-up spatial simulation executed by cellular automata (CA). The land demand from official data sources ensures the authority of the land trajectories under different SSP-RCP scenarios. The CA model is executed separately in different regions of the world to reflect spatial heterogeneity. Based on the LULC dataset, we further subdivide the its land types into 20 types according to climate data. The results show that our global land simulation achieves a satisfactory accuracy. The spatial change of various land types in each scenario appropriately reflects the storyline of the corresponding scenario. We believe that our dataset can provide stronger data support for climate research, including climate models, due to its advantages in resolution, scenarios, and land types. Please cite the corresponding paper: Chen, G., Li, X. &amp; Liu, X. Global land projection based on plant functional types with a 1-km resolution under socio-climatic scenarios. <em>Sci Data</em> 9, 125 (2022).人类活动的持续增强与气候变化的加剧共同影响了全球土地覆盖变化。基于情景的土地变化模拟作为评估潜在土地变化影响的有力工具，已得到广泛应用。然而，当前全球尺度的未来土地预测产品普遍存在分辨率偏低、情景适配性不足、土地类型较少等缺陷。为此，本研究构建了一系列2015年至2100年、分辨率为1千米的全球土地预测数据集，采用了最新的政府间气候变化专门委员会（Intergovernmental Panel on Climate Change, IPCC）提出的社会经济与气候变化耦合情景——共享社会经济路径-典型浓度路径（SSP-RCP）。 该数据集包含两个子数据集：1）包含7种土地类型的1千米全球土地利用与土地覆盖（Land Use and Land Cover, LULC）数据集；2）包含20种土地类型的1千米全球基于植物功能型（Plant Functional Type, PFT）的土地数据集。 LULC数据集的制作结合了耦合模式比较计划第六阶段（CMIP6）官方数据集提供的自上而下土地需求约束，以及元胞自动机（Cellular Automata, CA）执行的自下而上空间模拟。官方数据源提供的土地需求保障了不同SSP-RCP情景下土地变化轨迹的权威性。研究针对全球不同区域分别执行元胞自动机模型，以体现空间异质性。 基于LULC数据集，本研究进一步结合气候数据将其土地类型细分为20类。结果表明，本研究的全球土地模拟精度令人满意。各情景下各类土地类型的空间变化恰当地反映了对应情景的叙事逻辑。我们认为，本数据集在分辨率、情景设置与土地类型覆盖度上均具备优势，可为包括气候模式在内的气候研究提供更有力的数据支撑。 请引用相关论文：Chen G, Li X, Liu X. 社会-气候情景下基于植物功能型的1千米分辨率全球土地预测. Sci Data, 9, 125 (2022).