Mapping forest stand age in China using remotely sensed forest height and observation data

Abstract Forest stand age plays a crucial role in determining the terrestrial carbon source or sink strength and reflects major disturbance information. Forests in China have changed drastically in recent decades, but quantification of spatially explicit forest age at national level has been lacking to date. This study generated a national map of forest age at 1 km spatial resolution using the remotely sensed forest height and forest type data in 2005, as well as relationships between age and height retrieved from field observations. These relationships include biomass as an intermediate parameter for major forest types in different regions of China. Biomass‐height and age‐biomass relationships were well fitted using field observations, with respective R 2 values greater than 0.60 and 0.71 ( P < 0.01), indicating the viability of age‐height relationships developed for age estimation in China. The resulting map was evaluated by comparison with national, provincial, and county forest inventories. The validation had high regional ( R 2 = 0.87, 2–8 years errors in six regions), provincial ( R 2 = 0.53, errors less than 10 years and consistent age structure in most provinces), and plot ( R 2 values of 0.16−0.32, P < 0.01) agreement between map values and inventory‐based estimates. This confirms the reliability and applicability of the age‐height approach demonstrated in this study for quantifying forest age over large regions. The map reveals a large spatial heterogeneity of forest age in China: old in southwestern, northwestern, and northeastern areas, and young in southern and eastern regions. Key Points A statistically viable approach to map forest stand age was presented A 1 km national forest age map was produced using observations and height data Main potential uncertainties for broad‐scale applications were discussed