China’s annual forest age dataset at 30-m spatial resolution from 1986 to 2022

This dataset presents China’s Annual Forest Age (CAFA) at 30-m resolution from 1986 to 2022 (Version 2.0). It was derived by merging forest disturbance detection using Landsat data and age mapping of undisturbed forests using machine learning methods based on forest height, climate, terrain, and Landsat data (Shang et al., 2024; Shang et al., 2023; Lin et al., 2023). The forest extent was determined by the CLCD forest cover dataset (Yang et al. 2021). Areas classified as non-forest are set to -1, and the forest age for the year in which a disturbance occurred is set to 0. We welcome any feedback on our data for future updates!<b>Update History</b>Version 1.0: Initial release of China's forest age in 2019, as reported by Shang et al., 2023.Version 1.1: Update the forest mask using the CLCD forest cover dataset.Version 1.2: Improved forest age retrieval of undisturbed forests using machine learning methods with optimized model inputs, especially for the northeast and southwest of China.Version 1.3: Improved forest disturbance detection by integrating spatial information.Version 1.4: Expansion to include annual forest age covering the period from 1986 to 2022 based on enhanced forest disturbance detection.Version 1.5: Improved forest disturbance detection by integrating bidirectional monitoring.Version 2.0: Update the forest ages before their first forest disturbance, as reported by Shang et al., 2025.<b>Notice</b>Due to storage limitations, only the forest age data in 2019 is uploaded here. For access to the data from other years, please click Google Drive for downloading.<b>Emails: </b>Rong Shang ([e-mail], https://www.researchgate.net/profile/Rong-Shang), Jing M. Chen ([e-mail]).<b>Citations</b>Shang R., Lin, X. Chen J.M., et al.,(2025), China's annual forest age dataset at 30 m spatial resolution from 1986 to 2022. <i>Earth System Science Data</i>. In press. [Link]Shang R., Chen J.M., Xu M., et al.,(2023), China's current forest age structure will lead to weakened carbon sinks in the near future. <i>The Innovation</i> 4(6),100515. [Link]Lin, X., Shang, R., Chen, J.M., et al.,(2023), High-resolution forest age mapping based on forest height maps derived from GEDI and ICESat-2 space-borne lidar data. <i>Agricultural and Forest Meteorology</i> 339, 109592. [Link]<br>

本数据集提供了1986年至2022年分辨率为30米的中国年度森林年龄（China’s Annual Forest Age, CAFA）数据集（版本2.0）。本数据集通过融合基于陆地卫星（Landsat）数据的森林扰动检测结果，以及基于森林高度、气候、地形和陆地卫星数据、采用机器学习方法生成的未受干扰森林年龄制图得到（Shang等，2024；Shang等，2023；Lin等，2023）。森林范围由CLCD森林覆盖数据集（CLCD Forest Cover Dataset，Yang等，2021）确定。被分类为非森林的区域赋值为-1，发生森林扰动年份的森林年龄赋值为0。欢迎各界对本数据集提出反馈意见，以便后续更新！

更新历史
版本1.0：首次发布2019年版中国森林年龄数据集，相关成果见Shang等，2023。
版本1.1：采用CLCD森林覆盖数据集更新森林掩膜。
版本1.2：优化机器学习模型输入参数，改进未受干扰森林的年龄反演方法，尤其针对中国东北和西南地区。
版本1.3：整合空间信息，提升森林扰动检测精度。
版本1.4：基于改进的森林扰动检测方法，将数据集扩展至覆盖1986-2022年的年度森林年龄数据。
版本1.5：引入双向监测机制，进一步优化森林扰动检测效果。
版本2.0：更新首次森林扰动前的森林年龄数据，相关成果见Shang等，2025。

注意事项
受存储容量限制，本次仅上传2019年的森林年龄数据。如需获取其他年份的数据集，请点击谷歌网盘（Google Drive）下载。

联系方式
尚蓉（[e-mail]，https://www.researchgate.net/profile/Rong-Shang），陈镜明（[e-mail]）

引用文献
1. Shang R., Lin X., Chen J.M. 等(2025). 1986-2022年30米分辨率中国年度森林年龄数据集. 《地球系统科学数据》(Earth System Science Data), 已录用（In press）. [链接]
2. Shang R., Chen J.M., Xu M. 等(2023). 中国当前森林年龄结构将导致近期碳汇能力减弱. 《创新》(The Innovation) 4(6), 100515. [链接]
3. Lin X., Shang R., Chen J.M. 等(2023). 基于GEDI和ICESat-2星载激光雷达数据获取的森林高度图实现高分辨率森林年龄制图. 《农业与森林气象学》(Agricultural and Forest Meteorology) 339, 109592. [链接]

搜集汇总

背景与挑战

背景概述

该数据集提供了中国1986年至2022年年度森林年龄数据，空间分辨率为30米，基于Landsat数据和机器学习方法生成，涵盖森林干扰检测和未受干扰森林的年龄映射。数据集不断更新改进，当前版本为2.0，但受存储限制，页面仅包含2019年数据示例，完整年度数据需通过外部链接获取。