基于多源遥感协同的土地利用分类与碳排放时空格局——以青海湖流域为例

doi:10.13474/j.cnki.11-2246.2026.0329

摘要/Abstract

摘要： 在“双碳”战略背景下,精细识别区域土地利用变化及其碳排放时空格局至关重要。本文以青海湖流域为例,构建了“多源数据协同—土地利用分类—碳排放时空格局分析”的技术框架。通过集成Sentinel-1/2数据,以及样本自动化生成、多特征优化技术,构建了面向对象的随机森林(RF)分类模型,实现了青海湖流域土地利用的精确分类。在此基础上,结合系数法与能源消耗间接估算法,核算、分析了2017—2022年土地利用碳排放的时空演变。研究表明:①本文框架可显著减轻人工标注工作量并提高采样一致性,有效地将37个特征变量缩减至27个,并通过RF分类模型,使分类总体精度和Kappa系数均达96 %以上;②青海湖流域碳收支结构以建设用地为主要碳源(占总量93%以上),水体、林地和草地为主要碳汇;③2017—2022年,受建设用地扩张,以及林地、湿地碳汇能力下降等因素影响,流域净碳排放量由6.262万t增长至13.445万t,增幅达114.71%;④碳排放空间格局呈现“走廊聚源—滨水聚汇”的特征。本文建立的自动化分类与碳排放核算框架,可为区域生态管理与碳中和路径提供决策依据。

关键词: 土地利用分类, 碳排放, 时空格局, 多源遥感, 特征优选, 青海湖流域

Abstract: In the context of the “Dual Carbon” strategy,precise identification of regional land use changes and their spatiotemporal patterns of carbon emissions is crucial.Taking the Qinghai Lake basin as a case study,this paper establishes a technical framework integrating “multi-source data coordination,land use classification,and spatio-temporal analysis of carbon emissions”.By combining Sentinel-1/2 satellite data,automated sampling,and multi-feature optimization techniques,an object-oriented random forest classification model is developed to achieve accurate land use classification in the Qinghai Lake basin.Building on this foundation,the study combines coefficient analysis with indirect energy consumption estimation to calculate and analyze the spatiotemporal evolution of land use-related carbon emissions from 2017 to 2022.The findings demonstrate:①The proposed framework significantly reduces manual annotation workload and improves sampling consistency,effectively compressing 37 feature variables to 27.Through the implementation of the object-oriented random forest model,both overall classification accuracy and Kappa coefficient exceeds 96%.②The carbon balance structure in Qinghai Lake basin shows construction land as the primary carbon source (accounting for over 93%of total emissions),with water bodies,forests,and grasslands serving as major carbon sinks.③From 2017 to 2022,net carbon emissions in the basin surges from 6.262×10⁴ t to 13.445×10⁴ t,representing a 114.71%increase,driven by urban expansion and declining carbon sequestration capacity of forests and wetlands.④The spatial distribution of carbon emissions exhibits a “corridor-based concentration of sources—waterside concentration of sinks” pattern.The automated classification and carbon accounting framework develop in this study provides decision-making support for regional ecological management and carbon neutrality pathways.

Key words: land use classification, carbon emissions, spatio-temporal pattern, multi-source remote sensing, feature selection, Qinghai Lake basin

中图分类号:

P237

宋盈滨, 徐青, 叶飞, 罗伟. 基于多源遥感协同的土地利用分类与碳排放时空格局——以青海湖流域为例[J]. 测绘通报, 2026, 0(3): 174-180.

SONG Yingbin, XU Qing, YE Fei, LUO Wei. Land use classification and spatio-temporal pattern of carbon emissions based on multi-source remote sensing collaboration: a case study of Qinghai Lake basin[J]. Bulletin of Surveying and Mapping, 2026, 0(3): 174-180.

参考文献

[1] MENG Zirui,HE Mengxuan,LI Xuemei,et al.Spatio-temporal analysis and driving forces of urban ecosystem resilience based on land use:a case study in the Great Bay area[J].Ecological Indicators,2024,159:111769.
[2] ZHANG Jinlong,QI Yuan,YANG Rui,et al.Impacts of climate change and land use/cover change on the net primary productivity of vegetation in the Qinghai Lake basin[J].International Journal of Environmental Research and Public Health,2023,20(3):2179.
[3] VIZZARI M,LESTI G,ACHARKI S.Crop classification in Google Earth Engine:leveraging Sentinel-1,Sentinel-2,European CAP data,and object-based machine-learning approaches[J].Geo-spatial Information Science,2025,28(3):815-830.
[4] ZHOU Yi,ZHANG Rui,WANG Shixin,et al.Feature selection method based on high-resolution remote sensing images and the effect of sensitive features on classification accuracy[J].Sensors,2018,18(7):2013.
[5] SUN Jialin,YAN Shuai,ALEXANDRIDIS T,et al.Enhancing crop mapping through automated sample generation based on segment anything model with medium-resolution satellite imagery[J].Remote Sensing,2024,16(9):1505.
[6] CUI Yanglin,ZHAO Chunjiang,QU Xuzhou,et al.Mapping land-cover dynamics in arid regions using spectral mixture analysis and representative training samples[J].IEEE Transactions on Geoscience and Remote Sensing,2024,62:5643413.
[7] POTT L P,AMADO T J C,SCHWALBERT R A,et al.Satellite-based data fusion crop type classification and mapping in Rio Grande do Sul,Brazil[J].ISPRS Journal of Photogrammetry and Remote Sensing,2021,176:196-210.
[8] TANG Zhonglin,WANG Yuting,FU Min,et al.The role of land use landscape patterns in the carbon emission reduction:empirical evidence from China[J].Ecological Indicators,2023,156:111176.
[9] 孙宇栋,毛德华,邱冬冬.长沙市土地利用碳排放的时空演变特征及多情景模拟[J].农业资源与环境学报,2025,42(2):529-542.
[10] ROHLF G.Building new China,Colonizing Kokonor[M].[S.l.]:Lexington Books,2016.
[11] CUI Buli,LI Xiaoyan,WEI Xinghua.Isotope and hydrochemistry reveal evolutionary processes of lake water in Qinghai Lake[J].Journal of Great Lakes Research,2016,42(3):580-587.
[12] LEBAKULA V,SIMS K,REITH A,et al.LandScan global 30 arcsecond annual global gridded population datasets from 2000 to 2022[J].Scientific Data,2025,12:495.
[13] 张增祥,汪潇,温庆可,等.土地资源遥感应用研究进展[J].遥感学报,2016,20(5):1243-1258.
[14] HUANG Huabing,WANG Jie,LIU Caixia,et al.The migration of training samples towards dynamic global land cover mapping[J].ISPRS Journal of Photogrammetry and Remote Sensing,2020,161:27-36.
[15] TASSI A,VIZZARI M.Object-oriented LULC classification in google earth engine combining SNIC,GLCM,and machine learning algorithms[J].Remote Sensing,2020,12(22):3776.
[16] ZHOU Yi,TIAN Shufang,CHEN Jianping,et al.Research on classification of open-pit mineral exploiting information based on OOB RFE feature optimization[J].Sensors,2022,22(5):1948.
[17] XU Mengxia,WANG Mingchang,WANG Fengyan,et al.Extraction of gully erosion using multi-level random forest model based on object-based image analysis[J].International Journal of Applied Earth Observation and Geoinformation,2025,137:104434.
[18] BELGIU M,DRǍGUŢ L.Random forest in remote sensing:a review of applications and future directions[J].ISPRS Journal of Photogrammetry and Remote Sensing,2016,114:24-31.
[19] YANG Lingbo,WANG Limin,ABUBAKAR G A,et al.High-resolution rice mapping based on SNIC segmentation and multi-source remote sensing images[J].Remote Sensing,2021,13(6):1148.
[20] YAO Jinxi,WU Ji,XIAO Chengzhi,et al.The classification method study of crops remote sensing with deep learning,machine learning,and google earth engine[J].Remote Sensing,2022,14(12):2758.
[21] 曹周亮,张欣蓉,员学锋,等.基于土地利用变化的陕西省县域碳排放时空变化及影响因素研究[J].水土保持通报,2022,42(5):376-385.
[22] 李佳,焦利民.土地利用碳收支精细化核算与时空特征分析[J].测绘学报,2024,53(8):1480-1492.
[23] 刘畅,苏筠,黎玲玲.中国县域能源消费碳排放估算及其空间分布[J].环境污染与防治,2020,42(1):113-119.
[24] HUANG Rongming,ZHANG Xiaocun,LIU Kaihua.Assessment of operational carbon emissions for residential buildings comparing different machine learning approaches:a study of 34 cities in China[J].Building and Environment,2024,250:111176.