基于地统计学和CFI样地的浙江省森林碳空间分布研究

基于浙江省2009年CFI固定样地数据、森林资源规划设计调查林相图,利用地统计学方法对浙江省森林碳空间分布进行了模拟分析。结果表明,CFI固定样地数据用于省域范围的森林碳汇空间特征研究是合适的。数据显示,浙江森林植被平均碳密度为22.07Mg/hm2;与四川、福建、海南等地相比,平均碳密度较低。受人类活动、自然环境等因素影响,浙江省森林碳分布主要表现为:总体上森林碳密度空间变化趋势自西向东逐渐降低,与自然空间(海拔、地势等)趋势一致。基于地统计学和CFI固定样地,对省域范围的森林资源空间分布的研究,可以为省域森林碳汇管理提供依据,为我国特别是亚热带南方集体林区利用国家CFI数据进行大区域同类研究提供借鉴。

Spatial distribution of forest carbon in Zhejiang Province with geostatistics based on CFI sample plots

Research on forest carbon spatial distribution is an important topic related to global carbon cycling. Much previous research for forest carbon estimation is based on remote sensing techniques or ecosystem models. Use of pure sample plots for mapping forest carbon distribution has not been paid much attention because of the difficulty in collecting a large number of sample plots and the lack of suitable techniques and methods to accurately interpolate the results from sample plots to all areas without sample plots. Geostatistics has been regarded as a powerful tool for spatial data analysis. Its prerequisite is to have grid-based samples with sufficiently high density. The continuous forest inventory (CFI) system conducted in China can meet the need of Geostatistical analysis. The CFI system takes a province as a population, which the systematic sampling technique is used to allocate permanent sample plots on the ground with sufficient number and density. The sample plots are re-inventoried at every 5-year interval in the past 30 years, providing a sound foundation for analyzing forest carbon dynamic change. However, the CFI sample plots have not been effectively used to estimate forest carbon stocks in previous research. Therefore, this paper employed CFI dataset collected in 2009 and the forest distribution maps to simulate spatial distribution of forest carbon in Zhejiang province with the Geostatistical techniques.There are 4252 permanent sample plots in Zhejiang province. According to the definition of forest, 2528 plots belong to forest and these sample plots are used to calculate forest cover percentage. The rest plots belong to non-forest sample plots, including agriculture, water bodies, infrastructure, non-forest lands, and others. The plot size is 800 m² with interval distances of 6 km in east-west direction and 4 km in south-north direction between nearby sample plots. In a forest sample plot, trees, bamboo, shrub, and grass are used to calculate carbon stock. In a non-forest sample plot, only tree and bamboo around the fields or residential areas are used to calculate carbon stock. The vegetation carbon stock includes aboveground and underground carbon.The analysis of spatial autocorrelation of forest carbon density in Zhejiang (Global Moran's I_g=0.2342) indicates that the spatial variation of carbon density appears medium autocorrelation with partially patchy distribution. The results indicate that the Geostatistical technique can be effectively used to examine the spatial variability of forest carbon in Zhejiang province based on CFI data. This research implies the significance for applying CFI data with the Geostatistical methods to examining forest carbon distribution in subtropical regions, especially in the southern collective forest regions of China with fragmented forest stands.The average carbon density in Zhejiang province is 22.07Mg/hm², slightly lower than the carbon density in Sichuan, Fujian, and Hainan. The overall trend of carbon density is high in western part, and gradually decreases in eastern part, similar as the topographic change. The highest carbon density mainly distributes in the southwestern mountainous regions, followed by northwestern and western mountainous regions; and the lowest carbon density distributes in central hills and basins, coastal regions in eastern and southeastern parts, and the plains in northeastern regions. This situation implies the impacts of human-induced factors on forest carbon sequestration. In mountainous regions with relatively high elevation, forest lands have higher carbon density due to high forest cover rates and volumes, while in lower elevation regions, human activities significantly result in forest disturbance. The young plantations often have relatively low carbon density. Therefore, it is important to improve the forest management, especially in the regions with high human activities, in order to increase carbon density and keep forest sustainability.