Abstract: Stand age is a key ecological and management indicator for rubber plantations, because it controls plantation productivity, renewal cycles, carbon accumulation, and tropical land-use legacies. However, spatially explicit long-term information on rubber plantation age remains limited in China. In this study, we reconstructed annual 30 m stand-age maps of rubber plantations in China from 1987 to 2022 by integrating Landsat time-series analysis, an improved Bi-directional Continuous Change Detection and Classification (BCDC) algorithm, handheld laser scanning (HLS)-derived plot height, and a four-parameter logistic height-age model. The resulting maps revealed a transition from a juvenile-dominated plantation system in 1987 (≤5 yr, 91.08%) to a multi-cohort structure by 2022, when juvenile (≤5 yr, 10.77%), middle-aged (6-15 yr, 30.02%), mature (16-30 yr, 24.66%), old (31-36 yr, 26.07%), and long-established plantations (37-48 yr, 8.48%) jointly shaped the national age composition. Provincial patterns diverged, with a broad multi-age structure in Yunnan, mature-old dominance in Hainan, and a small fragmented plantation area in Guangdong. Stand age varied along latitude, elevation, and slope gradients, indicating the combined influence of climatic suitability, terrain constraints, and regional plantation histories. Methodological validation supported map reliability (RMSE = 2.95 yr; rRMSE = 23.64%) and showed improved performance relative to standard CCDC and LandTrendr. The results reveal the evolution and regional differentiation of China’s rubber plantation age structure and provide a data basis for rubber plantation renewal planning, latex-production potential assessment, and carbon accounting in tropical rubber plantation ecosystems.

Key words: Google Earth Engine, Handheld laser scanning, Rubber plantations, Stand age, Temporal dynamics, Bi-directional CCDC