Forest Classification And AGB Estimation Based On Multi-source Data In A Typical Subtropical Region

Species-rich subtropical forests have high carbon sequestration capacity and play important roles in regional and global carbon regulation and climate changes.A timely investigation of the spatial distribution characteristics of subtropical forest classification and aboveground biomass(AGB)is essential as the forest resource management refinement.This research selects Gaofeng Forest Farm as a study area.Multiple data sources(high spatial resolution satellite images such as ZY-3 and Sentinel-2 multispectral images,Lidar and AGB samples)are used to conduct forest classification.This research aims to improve forest AGB estimation through exploring suitable stratification approaches based on lidar and field survey data.Different stratification schemes including non-stratification and stratifications based on forest types and forest stand structures were examined.The AGB estimation models were developed using linear regression(LR)and random forest(RF)approaches.Based on SBFT5 and SBFSS hierarchical model,the spatial distribution pattern of forest biomass in the study area was simulated with the forest classification results and forest phase maps,and the influence of forest misclassification on biomass estimation was revealed.The results indicate the following:(1)The forest classification overall accuracy and kappa coefficient reached 81.6% and 0.79,respectively based on multiple source data.The complex and abundant tree species in the subtropical ecosystem pose a challenge to forest classification,and the spectral and spatial characteristics generated by optical remote sensing are lacking in subdivision of similar features.Stand structure variables play an important role in the classification of subtropical forest tree species,especially for plantations.(2)Proper stratifications improved AGB estimation and reduced the effect of under-and overestimation problems.AGB estimation based on stratification of forest stand structures was similar to that based on five forest types,implying that proper stratification reduces the number of sample plots needed.The RF algorithm provided better AGB estimation for non-stratification than the LR algorithm,but the LR approach provided better estimation with stratification.The optimal AGB estimation model and stratification scheme varied,depending on forest types.(3)Forest misclassification greatly affected the estimation of biomass,and the difference between the biomass of easily confused tree species in different layers was large.