Spatial heterogeneity of forest carbon stocks in the Xiangjiang river Basin urban agglomeration: analysis and assessment based on the multiscale geographically weighted regression (MGWR) model
Yingfeng Kuang, Xiaolong Chen
Abstract
Purpose Urban forests play a key role in mitigating global warming and achieving carbon neutrality. This study aims to analyze and evaluate the spatial heterogeneity of forest carbon stocks in the Xiangjiang River Basin urban agglomeration. By constructing and comparing the ordinary least squares model (OLS) and four geographically weighted regression (GWR) models, it is hoped to provide a more reliable method for accurately estimating the spatial distribution of large-scale forest carbon stocks and provide a scientific basis for the construction of the Xiangjiang River Basin forest urban agglomeration. Method Based on the data of the 10th continuous forest resource inventory and climate data in Hunan Province, this study identified five key variables, including average breast diameter of the stand, stand density, average age of the stand, average tree height of the stand, and average annual precipitation. Utilizing SPSS V27 software and MGWR 2.2 software, the OLS model and four GWR models were constructed. By comparing the model fit and the results of the independent samples test, the optimal model, the MGWR (Gaussian) model, was selected to estimate the spatial distribution of forest carbon stocks in the Xiangjiang River Basin. Results The results show that the four GWR models outperform the OLS model in terms of model fit and independent samples test, particularly in estimating the spatial distribution of forest carbon stocks. The results of the spatial non-stationarity test indicate that the MGWR model better captures the spatial heterogeneity of variables. The estimated carbon stock per unit area of forest in the Xiangjiang River Basin using the MGWR (Gaussian) model is 31.162 t/hm 2 , exhibiting an overall pattern of high central values and low peripheral values. This finding provides a crucial scientific basis for the management and ecological protection of forest resources in the Xiangjiang River Basin. Conclusion This study effectively reflects the spatial relationship between forest carbon stocks and variables through the geographically weighted regression method and the selection of appropriate spatial kernel functions, enhances the estimation accuracy of the spatial distribution of large-scale forest carbon stocks, and accurately reveals the spatial distribution pattern of forest carbon stocks in the Xiangjiang River Basin. The study on forest carbon stocks in the urban agglomeration of the Xiangjiang River Basin holds significant implications for regional ecological security and climate change mitigation. It offers a scientific basis for the management and ecological protection of regional forest resources.