Biomass Spatial Pattern and Driving Factors of Different Vegetation Types of Public Welfare Forests in Hunan Province

An ecological public welfare forest is an important basis for the construction of national ecological security. This study took public welfare forests at the provincial level or above in Hunan Province as the research object. Based on the in situ monitoring data and remote sensing data, we constructed a random forest (RF) model for inversing the biomass of public welfare forests with different types. Then, based on the inversion results, we investigated the biomass spatial pattern. Combined with topographical and socio-economic factors, we constructed a geographically weighted regression (GWR) model to analyze the biomass driving factors of different vegetation types in public forests. The results showed the following: (1) The biomass of public welfare forests in Hunan Province presented a strip distribution pattern that gradually increases from the central to the southwest and northeast. The total biomass of public welfare forests in Hunan Province was 338.13 million tons, with an average biomass of 68.31 t center dot hm(-2). In the different types of public welfare forests, the mean biomass of the types were as follows: shrub (4.65 t center dot hm(-2)) < broadleaf forest (59.27 t center dot hm(-2)) < conifer-broadleaf mixed forest (62.44 t center dot hm(-2)) < bamboo forest (71.33 t center dot hm(-2)) < coniferous forest (100.33 t center dot hm(-2)). (2) Topographic and socio-economic factors have a significant impact on the spatial pattern of biomass in public welfare forests. Slope had the greatest effect on coniferous forest, conifer-broadleaf mixed forest, and shrub forest, while POP had the greatest effect on broadleaf forest and bamboo forest. This study investigates the spatial patterns and driving factors of biomass in public welfare forests at the provincial level, filling the gap in forest biomass monitoring in public welfare forests in Hunan Province. It provides a new method to improve the accuracy of forest biomass estimation and data support for the sustainable management of public welfare forests.