新疆天山北坡城市群土地覆被数据产品精度评价  

Evaluation of land cover data product accuracy in urban agglomeration on the northern slope of Tianshan Mountains in Xinjiang

作　　者：闫兆进 孙雨晴 HE Rong 王冉[1,3] 阮晓光[5] 杨慧 慈慧[1,3] YAN Zhaojin;SUN Yuqing;HE Rong;WANG Ran;RUAN Xiaoguang;YANG Hui;CI Hui(School of Resources and Geosciences,China University of Mining and Technology,Xuzhou 221116,Jiangsu,China;Jiangsu Key Laboratory of Coal-Based Greenhouse Gas Control and Utilization,China University of Mining and Technology,Xuzhou 221008,Jiangsu,China;Key Laboratory of Coal Bed Gas Resources and Forming Process of Ministry of Education,China University of Mining and Technology,Xuzhou 221116,Jiangsu,China;Department of Civil,Environmental and Sustainable Engineering,Santa Clara University,Cingifornia 95053,USA;College of Geomatics and Municipal Engineering,Zhejiang University of Water Resources and Electric Power,Hangzhou 310018,Zhejiang,China)

机构地区：[1]中国矿业大学资源与地球科学学院,江苏徐州221116 [2]中国矿业大学江苏省煤基温室气体减排与资源化利用重点实验室,江苏徐州221008 [3]中国矿业大学煤层气资源与成藏过程教育部重点实验室,江苏徐州221116 [4]圣克拉拉大学土木、环境与可持续工程系,美国加州95053 [5]浙江水利水电学院测绘与市政工程学院,浙江杭州310018

出　　处：《干旱区地理》2024年第11期1852-1862,共11页Arid Land Geography

基　　金：新疆维吾尔自治区重点研发计划项目课题(2022B01012-1);科技部第三次新疆综合科学考察项目课题(2022xjkk1006);国家自然科学基金项目(42201451);中国博士后科学基金面上项目(2022M723379);中央高校基本科研业务费专项(2024ZD-PYCH1003);浙江省社科联项目(2024N085)资助。

摘　　要：土地覆被产品为全球各种地球系统科学应用提供了重要的地表覆被信息,如CLCD30(30 m)、GlobeLand30(30 m,简称Globe30)、GLC_FCS30(30 m)、FROM-GLC10(10 m)、Esri_Land_Cover_2020(10 m,简称Esri10),以及ESAWorldCover2020(10 m,简称ESA10),然而其在局地的精度和适用性如何尚不明确。基于Sentinel-2影像,通过样本精度评价和类别混淆评价对上述6套土地覆被数据产品在天山北坡城市群的精度及误差情况进行了研究和分析,并探讨了误差成因和不同数据产品的适用性。结果表明:(1)6套数据产品中,除Esri10外,其余5套数据产品的类型构成、面积占比相对一致。(2)GLC-FCS30、Globe30、CLCD30、FROM-GLC10、ESA10、Esri10的总体精度分别为0.8080、0.8147、0.7880、0.8531、0.8047、0.4725。(3)从产品适用性来看,GLC_FCS30适用于对耕地、裸地的分析,CLCD30适用于对林地、裸地的分析,FROM-GLC10适用于对草地、水体、冰/雪以及建筑的分析,ESA10适用于对耕地和草地的分析,Esri10适用于对林地、冰/雪以及建筑的分析,Globe30在各类别精度评价结果上更均衡。(4)类别混淆主要是耕地、林地以及草地之间及与其他类别的相互混淆,尤其是在土地覆被复杂的地区,如城市边缘区域。Land cover data products,including CLCD30(30 m resolution),GlobeLand30(30 m,referred to as Globe30),GLC_FCS30(30 m),FROM-GLC10(10 m),Esri_Land_Cover_2020(10 m,referred to as Esri10),and ESAWorldCover2020(10 m,referred to as ESA10),provide essential surface cover information for various earth system science applications globally.However,their accuracy and suitability for specific local areas remain uncertain.This study evaluates the accuracy and errors of these six land cover products in the northern slope of the Tianshan Mountains in Xinjiang,China,based on Sentinel-2 imagery,using sample accuracy assessment and category confusion evaluation.Furthermore,the causes of errors and the applicability of different products were analyzed.The results indicate that:(1)Among the six data product sets,except for Esri10,the composition and area percentages of land cover types are relatively consistent across the other five products.(2)The overall accuracies of these products are:GLC_FCS30(0.8080),Globe30(0.8147),CLCD30(0.7880),FROM-GLC10(0.8531),ESA10(0.8047),and Esri10(0.4725).(3)Regarding product applicability,GLC_FCS30 is effective for analyzing cropland and bareland;CLCD30 is optimal for forest and bareland;FROM-GLC10 excels in representing grassland,water bodies,snow/ice,and built-up areas;ESA10 is suitable for cropland and grassland;Esri10 performs well in forest,snow/ice,and built-up areas;and Globe30 maintains overall stability in accuracy across all categories.(4)Major category confusion occurs between cropland,forest,grassland,and other categories,especially in areas with complex land cover,such as urban fringes.

关 键 词：土地覆被数据产品 遥感 精度分析 类别混淆 天山北坡城市群 

分 类 号：P237[天文地球—摄影测量与遥感]