广西猫儿山森林植被碳密度空间变异特征分析

Spatial variation of forest carbon density in Mao’er mountain,Guangxi

【目的】猫儿山森林类型多样,森林植被空间变异性较大,准确估算该地区森林植被碳密度,探明其空间变异及分布特征对区域生态系统的碳汇管理具有重要意义。【方法】以广西猫儿山的森林植被为研究对象,基于GIS系统抽样布点,布设森林植被临时样地328个,利用地统计学、Moran’s I、基于半方差函数模型的克里格空间插值相结合的方法对猫儿山森林植被的碳密度空间变异及分布格局进行了研究。【结果】1)广西猫儿山森林植被碳密度平均值为46.82 t/hm^(2),森林植被碳密度数据服从正态分布;2)广西猫儿山森林植被碳密度全局Moran’s I值为0.366,森林植被碳密度存在正的空间自相关性(P<0.01),在空间分布上呈聚集分布;3)通过样本数据的统计和分析,选取球状模型(Spherical)、指数模型(Exponential)、高斯模型(Gaussian)、圆形模型(Circular)理论模型对猫儿山森林植被碳密度进行了拟合,指数模型的平均标准差最接近于0,标准均方根预测误差最接近于1,表明指数模型为最优模型。猫儿山森林植被碳密度指数模型的块基比(块金值与基台值)为52.78%,表明森林植被碳密度具有中等空间自相关性,其受随机性和结构性的综合影响。【结论】猫儿山森林植被碳密度空间分布规律与其地形、地貌趋于一致,从低海拔到高海拔,森林植被碳密度呈现先增加后降低的趋势。

【Objective】There are various forest types in Mao’er Mountain,and the spatial variability of forest vegetation is large.It is of great significance to accurately estimate the carbon density of forest vegetation in this region,and explore its spatial variation and distribution characteristics for regional ecological carbon sequestration.【Method】Taking the forest vegetation in Mao’er Mountain of Guangxi as the research object,328 temporary forest vegetation plots were set up based on GIS system sampling.The spatial variation and distribution pattern of forest carbon density in Mao’er Mountain were studied by using the geostatistics method,Moran’s I and Kriging spatial interpolation based on semi-variance function models.【Result】1)The average carbon density of the forest vegetation in Mao’er Mountain was 46.82t/hm^(2),and the data of forest carbon density showed normal distribution.2)The global Moran’s I value of forest carbon density was 0.366,indicating a positive spatial autocorrelation(P<0.01),and the spatial distribution of forest carbon density was clustered.3)The Spherical model,Exponential model,Gaussian model and Circular model were selected to fit the forest carbon density in Mao’er Mountain,and the average standard deviation of the Exponential model was closest to 0 and the root mean square prediction error was closest to 1,indicating that the Exponential model was the best fitting model for the estimation of forest carbon density.The ratio of the nugget block basis ratio was 52.78%,indicating a moderate spatial autocorrelation of caron density,which was affected by randomness and structures.【Conclusion】The spatial distribution of forest carbon density in Mao’er Mountain is consistent with its characteristics of topography and landform.From low altitude to high altitude,forest carbon density increases first and then decreases.