测绘通报 ›› 2020, Vol. 0 ›› Issue (11): 23-27.doi: 10.13474/j.cnki.11-2246.2020.0348

• 生态环境动态监测 • 上一篇    下一篇

不同趋势法的宁夏长时序植被变化分析

康雄, 曹俊涛, 陈成, 杨杰, 王建雄   

  1. 云南农业大学水利学院云南省高校农业遥感与精准农业工程研究中心, 云南 昆明 350201
  • 收稿日期:2020-01-06 出版日期:2020-11-25 发布日期:2020-11-30
  • 通讯作者: 王建雄。E-mail:jianxiongw@126.com E-mail:jianxiongw@126.com
  • 作者简介:康雄(1994-),男,硕士生,主要从事资源与环境遥感方面的研究。E-mail:xiongkrs@outlook.com
  • 基金资助:
    云南省教育厅科学研究基金研究生项目(2019Y0080)

Analysis of long-term vegetation change in Ningxia with different trend methods

KANG Xiong, CAO Juntao, CHEN Cheng, YANG Jie, WANG Jianxiong   

  1. Research Center of Agricultural Remote Sensing and Precision Agriculture Engineering in Yunnan Universities, College of Water Conservancy, Yunnan Agricultural University, Kunming 350201, China
  • Received:2020-01-06 Online:2020-11-25 Published:2020-11-30

摘要: 宁夏地处黄土高原,植被变化趋势直接影响其生态保护。本文应用2005—2015年的MODIS NDVI月合成产品,并采用最大合成法得到年NDVI数据,分别利用一元线性回归法和Sen+Mann-Kendall法对宁夏11年间植被变化趋势和空间差异进行研究分析。结果表明,在月际变化上,月NDVI均值呈高斯分布,7—9月是植被长势最好的阶段;在年际变化上,2005—2007年NDVI值明显增长,2008—2012年NDVI值呈稳定性增长,2013—2015年NDVI值有下降趋势;在变化趋势上,一元线性回归法与Sen+Mann-Kendall法得到的植被变化趋势基本一致,均表现为北部植被整体改善,但局部城区植被退化较为严重;中部地区轻微改善,局部存在明显改善;南部植被有明显改善且植被改善面积较大;通过差异性分析二者差异性仅为22.95%,且Sen+Mann-Kendall法能更好地监测轻微变化区域,变化趋势更加准确。

关键词: 宁夏, NDVI, 最大合成法, 一元线性回归, Sen+Mann-Kendall

Abstract: Ningxia is located in the Loess Plateau, the trend of vegetation change directly affects the ecological protection in this place. This paper uses MODIS NDVI monthly synthetic products from 2005 to 2015 to get annual NDVI data with the maximum synthetic method, and adopts the unitary linear regression method and the Sen+Mann Kendall method to analyze the vegetation change trend and spatial difference in Ningxia in recent ten years. The results show that the monthly mean value of NDVI presents Gauss distribution, and the best stage of vegetation growth is from July to September. The yearly NDVI value increases significantly in 2005—2007, the yearly NDVI value increases steadily in 2008—2012, and the yearly NDVI value decreases in 2013—2015. The trends in vegetation change obtained by the one-dimensional linear regression method and the Sen+Mann-Kendall method are almost the same, with both the overall improvement of vegetation in the north, but local urban vegetation degradation is more serious; slight improvement in the central region, local obvious improvement. The vegetation in the south is significantly improved and the area of vegetation improvement was larger. The difference between the two methods by differential analysis is only 22.95%, and the Sen+Mann-Kendall method is better for monitoring of areas of slight variation and the trends are more accurate.

Key words: Ningxia, NDVI, maximum value composite, univariate linear regression, Sen+Mann-Kendall

中图分类号: