基于遥感数据产品的亚热带地区土壤有机碳储量评估——以长株潭绿心区为例  

Assessment of Soil Organic Carbon Storage in Subtropical Regions Based on Remote Sensing Data Products:A Case Study of Changsha-Zhuzhou-Xiangtan Green Heart Region

作　　者：陈铸 肖海 全思湘 殷梓强 徐思源 CHEN Zhu;XIAO Hai;QUAN Si-xiang(Hunan Second Surveying and Mapping Institute,Changsha,Hunan 410000;Hunan Engineering Research Center for Monitoring of Natural Ecosystem Carbon Sinks,Changsha,Hunan 410000;Key Laboratory of Natural Resources Monitoring and Supervision in Southern Hilly Areas of the Ministry of Natural Resources,Changsha,Hunan 410000)

机构地区：[1]湖南省第二测绘院,湖南长沙410000 [2]自然生态系统碳汇监测湖南省工程研究中心,湖南长沙410000 [3]自然资源部南方丘陵区自然资源监测监管重点实验室,湖南长沙410000

出　　处：《安徽农业科学》2025年第7期35-40,55,共7页Journal of Anhui Agricultural Sciences

基　　金：湖南省自然资源厅科技计划项目(20230158CH);湖南省自然资源领域推进碳达峰碳中和重大技术研究项目(湘自科[2022]5号)。

摘　　要：以长株潭绿心中央公园核心区为研究对象,基于实地采集的土壤样点有机碳数据、多光谱遥感数据、数字高程模型(DEM)、气象数据(降水和温度)、国土年度变更调查数据及其他辅助数据,选择多元线性回归模型、逐步回归模型和随机森林模型分别对0~20 cm表层及剖面有机碳密度、有机碳储量进行反演和评估,比较不同模型方法的估算结果精度差异,分析各影响因子的重要性。结果表明:随机森林模型对0~20 cm表层土壤有机碳密度的反演精度最高(R^(2)=0.88),影响较大的因子为地表反射率蓝光波段、高程、水流强度指数、地形湿度指数和地形起伏度;逐步回归模型对剖面土壤有机碳密度的反演精度最高(R^(2)=0.54),影响较大的因子为高程、水流强度指数、地形起伏度、绿度指数、归一化植被指数、土壤调节植被指数和平均气温;0~20 cm表层土壤有机碳密度空间分布特征与地形起伏度空间分布十分吻合,地形地貌对表层土壤有机碳储量分布的影响较大;研究区不同土地利用方式0~20 cm表层土壤有机碳储量从大到小依次为林地>水田>园地>草地>旱地。Taking the core area of Changsha-Zhuzhou-Xiangtan Green Heart Central Park as the research object,based on on-site collected data including soil organic carbon data,multispectral remote sensing data,digital elevation models(DEM),meteorological data(precipitation and temperature),land annual change survey data,and other auxiliary data,multiple linear regression models,stepwise regression models and random forest models were selected to invert and evaluate the organic carbon density and storage in the 0-20 cm surface and profile,respectively.The accuracy differences of the estimation results of different model methods were compared,and the importance of each influencing factor was analyzed.The result showed that the random forest model demonstrated the highest accuracy in estimating the organic carbon density of 0-20 cm surface soil(R^(2)=0.88),with significant factors including surface reflectance in the blue light band,elevation,water flow intensity index,terrain humidity index and terrain undulation.The stepwise regression model showed the highest accuracy in estimating the profile soil organic carbon density(R^(2)=0.54),with significant factors including elevation,water flow intensity index,terrain undulation,greenness index,normalized vegetation index,soil-adjusted vegetation index and average temperature.The spatial distribution characteristics of 0-20 cm surface soil organic carbon density coincide with the spatial distribution of terrain undulation,indicating significant influence of terrain topography on the distribution of surface soil organic carbon reserves.The order of surface soil organic carbon density from highest to lowest among different land use types in the study area was forest land>paddy field>orchard land>grassland>dry land.

关 键 词：遥感数据产品 土壤有机碳密度 土壤有机碳储量 亚热带地区 

分 类 号：X144[环境科学与工程—环境科学]