干旱气象 ›› 2024, Vol. 42 ›› Issue (2): 274-282.DOI: 10.11755/j.issn.1006-7639(2024)-02-0274

• 新气象数据检验与订证 • 上一篇    下一篇

基于MODIS数据的三种植被指数在海南岛植被监测中的适用性研究

杨静1,2(), 张亚杰1,2(), 陈金威2,3, 朱晶晶1,2, 张明洁1,2, 林绍伍1,2   

  1. 1.海南省气候中心,海南 海口 570203
    2.海南省南海气象防灾减灾重点实验室,海南 海口 570203
    3.海南省气象服务中心,海南 海口 570203
  • 收稿日期:2022-09-23 修回日期:2023-04-14 出版日期:2024-04-30 发布日期:2024-05-12
  • 通讯作者: 张亚杰(1987—),男,河南省郑州市人,硕士,高级工程师,主要从事应对气候变化、健康气象及应用气象等研究。E-mail:zhyajie87@163.com
  • 作者简介:杨静(1993—),女,重庆市大足区人,硕士,助理工程师,主要从事健康气象研究。E-mail:graceyangOUC@163.com
  • 基金资助:
    海南省科技专项(ZDYF2022SHFZ036);海南省自然科学基金项目(823MS119);及海南省气象局技术提升项目(HNQXJS202305)

Research on the applicability of three vegetation indices based on MODIS data in vegetation monitoring of Hainan Island

YANG Jing1,2(), ZHANG Yajie1,2(), CHEN Jinwei2,3, ZHU Jingjing1,2, ZHANG Mingjie1,2, Lin Shaowu1,2   

  1. 1. Hainan Climate Center, Haikou 570203, China
    2. Key Laboratory of South China Sea Meteorological Disaster Prevention and Mitigation of Hainan Province, Haikou 570203, China
    3. Hainan Meteorological Service Center, Haikou 570203, China
  • Received:2022-09-23 Revised:2023-04-14 Online:2024-04-30 Published:2024-05-12

摘要:

在大力推进生态文明建设背景下,开展植被指数在海南岛植被监测中的适用性研究具有重要参考价值。基于海南岛2001—2020年中分辨率成像光谱仪(Moderate-Resolution Imaging Spectroradiometer,MODIS)提取的归一化植被指数(Normalized Differnce Vegetation Index, NDVI)、增强型植被指数(Enhanced Vegetation Index, EVI)、差值植被指数(Difference Vegetation Index, DVI)数据,海南岛18个国家气象观测站逐月平均气温和降水量数据,以及2015、2017年海南省地表覆盖数据,利用一元线性拟合、均方根误差分析、相关分析等方法开展了NDVI、EVI、DVI 3种植被指数在海南岛植被监测中的适用性研究。结果表明:(1)3种植被指数中,EVI对基于海南岛地表覆盖数据提取的植被覆盖面积的拟合效果最好,均方根误差占实际均值的9.57%。(2)3种植被指数中,EVI最能反映海南岛植被的季节变化特征,2月开始缓慢增加,8月达到峰值后缓慢下降,次年2月降至最低;EVI值域分布最广,最能反映不同茂密程度植被的覆盖状况,能更好克服NDVI在高植被区易饱和的问题;EVI与气候因子的相关性最大,最能反映植被对气候的响应特征。(3)综合评估3种植被指数对海南岛植被覆盖面积的拟合效果,对植被季节变化、茂密程度的表征能力,确定最适合表征海南岛植被特征的植被指数为EVI。

关键词: 海南岛, 植被指数, 适用性研究

Abstract:

Under the background of vigorously promoting the construction of ecological civilization, it is of great reference value to study the applicability of vegetation indices in vegetation monitoring in Hainan Island. Based on the NDVI (Normalized Differnce Vegetation Index), EVI (Enhanced Vegetation Index), DVI (Difference Vegetation Index) data extracted from MODIS (Moderate-Resolution Imaging Spectroradiometer), monthly mean temperature and precipitation data of 18 national meteorological observation stations from 2001 to 2020 and land cover data in 2015 and 2017 of Hainan Island, the applicability of three vegetation indices NDVI, EVI and DVI in vegetation monitoring of Hainan Island was studied by using the univariate linear fitting, root mean square error analysis and correlation analysis method. The results are as follows: (1) Among the three vegetation indices, EVI has the best fitting effect on the vegetation coverage area extracted from the land cover data of Hainan Island, the root mean square error accounts for 9.57% of the actual mean. (2) EVI can best reflect the seasonal variation characteristics of vegetation in Hainan Island, it began to increase slowly in February, slow decline after reaching its peak in August, and fell to the lowest in February next year; EVI has the widest range, which can best reflect the vegetation coverage with different thicknesses, and can better overcome the NDVI saturation problem in high vegetation area; EVI has the greatest correlation with climate factors and can best reflect the characteristics of vegetation's response to climate. (3) The fitting effect of three vegetation indices on the vegetation coverage area of Hainan Island was comprehensively evaluated, and the ability to characterize the seasonal variation and density of vegetation was evaluated. The EVI was determined to be the most suitable vegetation index for characterizing the vegetation characteristics of Hainan Island.

Key words: Hainan Island, vegetation indexes, applicability study

中图分类号: