收藏设为首页 广告服务联系我们在线留言

干旱区地理 ›› 2020, Vol. 43 ›› Issue (4): 1023-1032.doi: 10.12118/j.issn.1000-6060.2020.04.17

• 地球信息科学 • 上一篇    下一篇

基于多时相Landsat影像的中亚地区植被覆盖遥感监测

王华1,2, 杨乾鹏1, 田云杰1, 郭山川3, 唐鹏飞3   

  1. 1 西藏林芝市气象局,西藏 林芝 860000;
    2 南京农业大学资源与环境科学学院,江苏 南京 210095;
    3 南京大学地理与海洋科学学院,江苏 南京 210023
  • 收稿日期:2019-06-19 修回日期:2020-03-20 出版日期:2020-07-25 发布日期:2020-11-18
  • 作者简介:王华(1989–),女,四川犍为人,在读硕士,工程师,研究方向为植被遥感监测. E-mail:825222166@qq.com
  • 基金资助:
    西藏自治区科技计划项目(XZ201801-GB-12); 国家自然科学基金项目(41472207,41631176)资助

Vegetation coverage monitoring in the Central Asian countries using multi-temporal Landsat images

WANG Hua1,2, YANG Qian-peng1, TIAN Yun-jie1, GUO Shan-chuan3, TANG Peng-fei3   

  1. 1 Nyingchi meteorological service. Nyingchi 860000,Xizang,China;
    2 College of Resources and Environmental Sciences,Nanjing Agricultural University,Nanjing 210095,Jiangsu,China;
    3 School of Geography and Ocean Science,Nanjing University,Nanjing 210023,Jiangsu,China
  • Received:2019-06-19 Revised:2020-03-20 Online:2020-07-25 Published:2020-11-18

摘要: 针对中亚地区的强生态脆弱性、高敏感性特征,有必要开展广域、长期的植被覆盖监测以匹配“绿色丝绸之路”的可持续发展目标。鉴于此,联合Landsat 5和Landsat 8卫星数据集,利用Google Earth Engine(GEE)地理空间数据云计算平台,估算了中亚地区1993—2018年间共12期的植被覆盖度。结果表明:(1)中亚地区植被覆盖总体水平较低,但也具有较为显著的空间异质性。(2)中亚地区1993—2018年间多数区域植被覆盖趋势较为稳定,哈萨克斯坦丘陵、费尔干纳盆地等区域植被覆盖度呈增加趋势,乌拉尔河流域和锡尔河流域等区域植被覆盖趋势为负。(3)植被覆盖度时序特征上,中亚地区1993—2018年间总体植被覆盖度累积增加3%,其中吉尔吉斯斯坦和塔吉克斯坦植被覆盖分别增加3.96%和5.86%。(4)裸土区呈退缩趋势,面积总计减少25.9×104 km2,低植被覆盖区、中植被覆盖区和高植被覆盖区范围在呈现出的振荡式增加。研究结合遥感大数据和地理云计算对中亚地区进行区域尺度的植被覆盖动态监测,能对中亚地区生态评估和演替分析提供技术支持和定量数据。

关键词: 植被覆盖监测, 中亚地区, GEE, 多时相Landsat影像, 归一化植被指数

Abstract: Central Asian regions with a typical temperate desert as well as grassland environment experience several ecological environmental problems such as water shortages and land degradation. Particularly in recent decades,the ecological disturbances caused by climate change and human activity have led to more consideration being given to environmental protection. As the most essential producer of an ecosystem,vegetation coverage status can directly affect an ecosystem’s function and structure. Therefore,to align with the sustainable development goal of the “Green Silk Road” ,there is a need to conduct vegetation coverage monitoring with wide-range and long-term abilities. In view of this,12 periods of vegetation coverage in the Central Asian regions from 1993 to 2018 were estimated using the Google Earth Engine geospatial data cloud computing platform from Landsat 5 together with Landsat 8 remote sensing data sets. To start,the research area boundary was used as the selected condition to filter the Landsat images. Through data preprocessing,including a cloud mask,filter,and split,we acquired 12 high-quality remote sensing images over the Central Asian regions. Then,we calculated the normalized difference vegetation index using the NIR and RED bands of the selected images. The dimidiate pixel model was used to estimate the vegetation coverage,followed by a linear regression analysis. The results showed as follows:(1) The overall level of vegetation coverage in Central Asia was low but had significant spatial heterogeneity. (2) The vegetation coverage change trend of most regions in Central Asia between 1993 and 2018 was relatively stable. The vegetation coverage over the Kazakhstan Hills and Fergana Valley areas showed an increasing trend,while the vegetation coverage in the Ural River Basin and Syr Darya River Basin areas showed a negative trend. (3) In terms of the time series characteristics of the vegetation coverage,the total vegetation coverage increased by 3% in the Central Asian regions between 1993 and 2018,during which the vegetation coverage in Kyrgyzstan and Tajikistan increased by 3.96% and 5.86%,respectively. (4) The bare soil area showed a retreating trend,and its total area decreased by 259 000 km2. (5) Lastly,the low vegetation,middle vegetation,and high vegetation cover areas showed an oscillatory increase. Overall,this study combined remote sensing data with geographical cloud computing to monitor vegetation coverage on a regional scale in Central Asia to provide technical support and quantitative data for ecological assessment and succession analysis in Central Asia.

Key words: vegetation coverage monitoring, the Central Asian Regions, Google Earth Engine, multi-temporal Landsat images, NDVI