基于地理加权回归克里格模型的帽儿山地区森林碳储量空间分布

森林碳储量对于全球气候变化具有重要影响,以往的模型估算未考虑到模型残差的空间相关性和碳储量数据的非平稳性,影响模型的预测精度.本研究基于东北林业大学帽儿山实验林场的ETM+遥感影像数据和193块固定样地,利用地理加权克里格回归（GWRK）建立森林碳储量与遥感和地形因子的回归模型,同时对比最小二乘模型（OLS）、地理加权回归模型（GWR）的预测精度.结果表明:对于帽儿山地区的森林碳储量估算,GWRK的平均绝对误差（MAE）、均方根误差（RMSE）低于OLS模型和GWR模型,GWRK模型的平均误差（ME）低于GWR模型,与OLS模型相近.GWRK模型的预测精度为83.2%,较OLS模型（73.7%）和GWR模型（77.3%）分别提高6%和10%,拟合精度明显提高。

Forest carbon stock exerts a significant impact on global climate change. Previous model estimations failed to consider the spatial autocorrelation of model residuals and the non-stationarity of carbon stock data, which compromised the prediction accuracy of the models. In this study, based on ETM+ remote sensing image data from the Maoershan Experimental Forest Farm of Northeast Forestry University and 193 permanent sample plots, we established a regression model linking forest carbon stock with remote sensing and topographic factors using Geographically Weighted Kriging Regression (GWRK). Meanwhile, we compared the prediction accuracy of the Ordinary Least Squares model (OLS), Geographically Weighted Regression model (GWR), and the GWRK model. The results show that: For the estimation of forest carbon stock in the Maoershan area, the Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) of the GWRK model are lower than those of the OLS and GWR models; the Mean Error (ME) of the GWRK model is lower than that of the GWR model and similar to that of the OLS model. The prediction accuracy of the GWRK model reaches 83.2%, which is 6% and 10% higher than that of the OLS model (73.7%) and GWR model (77.3%) respectively, indicating a significant improvement in fitting accuracy.

搜集汇总

背景与挑战

背景概述

该数据集聚焦于帽儿山地区森林碳储量的空间分布研究，采用地理加权回归克里格（GWRK）模型，结合ETM+遥感影像和193块固定样地数据，以提高预测精度。结果表明，GWRK模型相比传统最小二乘和地理加权回归模型，预测精度提升至83.2%，有效解决了残差空间相关性和数据非平稳性问题。数据集属于国家重点研发计划项目，以文档格式提供，适用于森林碳储量估算和空间分析方法验证。