干旱区研究 ›› 2022, Vol. 39 ›› Issue (2): 594-604.doi: 10.13866/j.azr.2022.02.26

• 生态与环境 • 上一篇    下一篇

基于1990—2019年多时相影像的干旱区绿洲景观格局分析

宋奇1(),史舟2,冯春晖3,马自强4,纪文君5,彭杰3,高琪6,蒋学玮1()   

  1. 1.塔里木大学园艺与林学学院,新疆 阿拉尔 843300
    2.浙江大学环境与资源学院,浙江 杭州 310058
    3.塔里木大学农学院,新疆 阿拉尔 843300
    4.北京大学地球与空间科学学院 遥感与地理信息系统研究所,北京 100871
    5.中国农业大学土地科学与技术学院,北京 100193
    6.新疆昌吉州地质环境监测站,新疆 昌吉 831100
  • 收稿日期:2021-06-17 修回日期:2022-01-10 出版日期:2022-03-15 发布日期:2022-03-30
  • 通讯作者: 蒋学玮
  • 作者简介:宋奇(1996-),男,在读博士研究生,主要研究方向为遥感应用与地理空间数据分析. E-mail: tarimsongqi@163.com
  • 基金资助:
    兵团南疆重点产业创新发展支撑计划资助项目(2021DB015);兵团南疆重点产业创新发展支撑计划资助项目(2021DB019);兵团中青年创新领军人才项目(2020CB032);国家重点研发计划项目(2018YFE0107000)

Analysis of landscape pattern from 1990 to 2019 based on multi-temporal imagery in arid oasis

SONG Qi1(),SHI Zhou2,FENG Chunhui3,MA Ziqiang4,JI Wenjun5,PENG Jie3,GAO Qi6,JIANG Xuewei1()   

  1. 1. College of Horticulture and Forestry, Tarim University, Alar  843300, Xinjiang, China
    2. College of Environmental and Resource Science, Zhejiang University, Hangzhou  310058, Zhejiang, China
    3. College of Agriculture, Tarim University, Alar 843300, Xinjiang, China
    4. Institute of Remote Sensing and Geographical Information Systems, School of Earth and Space Sciences, Peking University 100871, Beijing, China
    5. College of Land Science and Technology, China Agricultural University 100193, Beijing, China
    6. Changji Geological Environment Monitoring Station, Changji 831100, Xinjiang, China
  • Received:2021-06-17 Revised:2022-01-10 Online:2022-03-15 Published:2022-03-30
  • Contact: Xuewei JIANG

摘要:

利用1990—2019年间阿拉尔垦区共226景的Landsat影像数据,对比不同分类方法,从中选出精度最高的分类方法,分别从斑块类型水平、景观水平、突变情况、空间分布方面分析阿拉尔垦区连续30 a间的景观格局变化特征。结果表明:多时相面向对象的分类结果最佳,总体精度为96.57%,Kappa系数为0.95。在斑块类型水平上,30 a间耕地趋于破碎化,未利用地的优势度降低,园地和林草地的景观形状趋于复杂化,水体和建设用地的景观趋于均衡化。在景观水平上,景观形状趋于复杂化和破碎化,景观连通性降低,景观丰富度和异质性增加。阿拉尔垦区在2005年发生景观格局突变。在空间分布上,呈现出以中部塔里木河区域向四周扩散的趋势。

关键词: Landsat, 多时相面向对象, 连续时间序列, 景观格局, 突变点检验, 阿拉尔垦区

Abstract:

In this study, 226 scenes of Landsat image data of Aral Reclamation Area from 1990 to 2019 were used to compare different classification methods, from which that with the highest accuracy was selected. The landscape pattern changes in the Aral Reclamation Area for more than 30 years were analyzed in terms of patch type level, landscape level, abrupt changes, and spatial distribution, respectively. The best results were obtained for the multi-temporal object-oriented classification, with an overall accuracy of 96.57% and a Kappa coefficient of 0.95. At the patch type level, arable land tends to fragment over 30 years, the dominance of unused land decreases, the landscape shape of parkland and woodland tends to become more complex, and the landscape of water bodies and built-up land tends to become more balanced. At the landscape level, landscape shapes become more complex and fragmented, landscape connectivity decreases, and landscape richness and heterogeneity increase. The Aral Reclamation Area underwent a sudden change in landscape pattern in 2005. In terms of spatial distribution, there is a tendency for the central Tarim River region to spread out in all directions.

Key words: Landsat, multi-temporal object-oriented, continuous time series, landscape pattern, mutation point inspection, Aral Reclamation Area