安徽农业科学2025，Vol.53Issue(3)：47-52,76,7.DOI:10.3969/j.issn.0517-6611.2025.03.010
基于Sentinel-2的江淮分水岭撂荒地提取
Extraction of Abandoned Land in Jianghuai Watershed Based on Sentinel-2
摘要
Abstract
[Objective]To study the problem of low accuracy in extracting abandoned land in the Jianghuai Watershed due to terrain fragmen-tation,hilly undulations and severe cloud coverage.[Method]Based on the GEE cloud platform,taking Fengyang County as the research area,Sentinel-2 MSI image data was used to classify the land in Fengyang County from 2017 to 2023 using the random forest classification algorithm.Seven years of land classification data were obtained,and abandoned land identification rules were formulated.Based on this,abandoned land data in the study area was extracted.[Result]The overall land classification accuracy of remote sensing images was 82.45%-91.37%,which met the accuracy requirements for research.The maximum areas of abandoned land in the study area was 974.52 hm2,with the highest aban-donment rate of 0.91%in 2018.The abandoned land area showed a decreasing trend from 2018 to 2019,and increased year by year from 2019 to 2021;the maximum reclamation area of abandoned land from 2018 to 2023 was about 678.15 hm2,while the minimum reclamation area was about 78.23 hm2,the highest reclamation rate was 69.59%,the minimum reclamation rate was 37.02%.[Conclusion]This study provides strong data support for evaluating the risk of farmland abandonment and ecological environment management,and also provides reference for ex-tracting other abandoned lands.关键词
撂荒地/江淮分水岭/Sentinel-2/随机森林分类算法/凤阳县Key words
Abandoned land/Jianghuai watershed/Sentinel-2/Random forest classification algorithm/Fengyang County分类
农业科技引用本文复制引用
李元庆,宋宏利,刘欢,李伟涛,刘兴宇..基于Sentinel-2的江淮分水岭撂荒地提取[J].安徽农业科学,2025,53(3):47-52,76,7.基金项目
河北省自然科学基金项目(D2019402067) （D2019402067）
安徽省教育厅自然科学重大项目(KJ2021ZD0131). （KJ2021ZD0131）
