A comparison of vegetation indices over a global set of TM images for EOS-MODIS

A. R. Huete, H. Q. Liu, K. Batchily, Willem van Leeuwen

Research output: Contribution to journalArticle

870 Citations (Scopus)

Abstract

A set of Landsat Thematic Mapper images representing a wide range of vegetation conditions from the NASA Landsat Pathfinder, global land cover test site (GLCTS) initiative were processed to simulate the Moderate Resolution Imaging Spectroradiometer (MODIS), global vegetation index imagery at 250 m pixel size resolution. The sites included boreal forest, temperature coniferous forest, temperate deciduous forest, tropical rainforest, grassland, savanna, and desert biomes. Differences and similarities in sensitivity to vegetation conditions were compared among various spectral vegetation indices (VIs). All VIs showed a qualitative relationship to variations in vegetation. However, there were significant differences among the VIs over desert, grassland, and forested biomes. The normalized difference vegetation index (NDVI) was sensitive to and responded primarily to the highly absorbing red reflectance band, while other indices such as the soil and atmosphere resistant vegetation index (SARVI) were more responsive to variations in the near-infrared (NIR) band. As a result, we found the NDVI to mimic red reflectances and saturate over the forested sites while the SARVI, by contrast, did not saturate and followed variations in NIR reflectances. In the arid and semiarid biomes, the NDVI was much more sensitive to canopy background variations than the SARVI. Maximum differences among vegetation index behavior occurred over the evergreen needleleaf forest sites relative to the deciduous broadleaf forests and drier, grassland, and shrub sites. These differences appear to be useful in complementing the NDVI for improved monitoring of vegetation, with the NDVI sensitive to fraction of absorbed photosynthetic active radiation and the SARVI more sensitive to structural canopy parameters such as leaf area index and leaf morphology.

Original languageEnglish (US)
Pages (from-to)440-451
Number of pages12
JournalRemote Sensing of Environment
Volume59
Issue number3
DOIs
StatePublished - Mar 1997

Fingerprint

Earth Observing System
moderate resolution imaging spectroradiometer
EOS
vegetation index
MODIS
Imaging techniques
NDVI
biome
reflectance
vegetation
atmosphere
grasslands
grassland
Landsat
deciduous forest
ecosystems
deserts
soil
near infrared
desert

ASJC Scopus subject areas

  • Computers in Earth Sciences
  • Earth-Surface Processes
  • Environmental Science(all)
  • Management, Monitoring, Policy and Law

Cite this

A comparison of vegetation indices over a global set of TM images for EOS-MODIS. / Huete, A. R.; Liu, H. Q.; Batchily, K.; van Leeuwen, Willem.

In: Remote Sensing of Environment, Vol. 59, No. 3, 03.1997, p. 440-451.

Research output: Contribution to journalArticle

Huete, A. R. ; Liu, H. Q. ; Batchily, K. ; van Leeuwen, Willem. / A comparison of vegetation indices over a global set of TM images for EOS-MODIS. In: Remote Sensing of Environment. 1997 ; Vol. 59, No. 3. pp. 440-451.
@article{27afbf052ddc40ed9aec218d1a70304e,
title = "A comparison of vegetation indices over a global set of TM images for EOS-MODIS",
abstract = "A set of Landsat Thematic Mapper images representing a wide range of vegetation conditions from the NASA Landsat Pathfinder, global land cover test site (GLCTS) initiative were processed to simulate the Moderate Resolution Imaging Spectroradiometer (MODIS), global vegetation index imagery at 250 m pixel size resolution. The sites included boreal forest, temperature coniferous forest, temperate deciduous forest, tropical rainforest, grassland, savanna, and desert biomes. Differences and similarities in sensitivity to vegetation conditions were compared among various spectral vegetation indices (VIs). All VIs showed a qualitative relationship to variations in vegetation. However, there were significant differences among the VIs over desert, grassland, and forested biomes. The normalized difference vegetation index (NDVI) was sensitive to and responded primarily to the highly absorbing red reflectance band, while other indices such as the soil and atmosphere resistant vegetation index (SARVI) were more responsive to variations in the near-infrared (NIR) band. As a result, we found the NDVI to mimic red reflectances and saturate over the forested sites while the SARVI, by contrast, did not saturate and followed variations in NIR reflectances. In the arid and semiarid biomes, the NDVI was much more sensitive to canopy background variations than the SARVI. Maximum differences among vegetation index behavior occurred over the evergreen needleleaf forest sites relative to the deciduous broadleaf forests and drier, grassland, and shrub sites. These differences appear to be useful in complementing the NDVI for improved monitoring of vegetation, with the NDVI sensitive to fraction of absorbed photosynthetic active radiation and the SARVI more sensitive to structural canopy parameters such as leaf area index and leaf morphology.",
author = "Huete, {A. R.} and Liu, {H. Q.} and K. Batchily and {van Leeuwen}, Willem",
year = "1997",
month = "3",
doi = "10.1016/S0034-4257(96)00112-5",
language = "English (US)",
volume = "59",
pages = "440--451",
journal = "Remote Sensing of Environment",
issn = "0034-4257",
publisher = "Elsevier Inc.",
number = "3",

}

TY - JOUR

T1 - A comparison of vegetation indices over a global set of TM images for EOS-MODIS

AU - Huete, A. R.

AU - Liu, H. Q.

AU - Batchily, K.

AU - van Leeuwen, Willem

PY - 1997/3

Y1 - 1997/3

N2 - A set of Landsat Thematic Mapper images representing a wide range of vegetation conditions from the NASA Landsat Pathfinder, global land cover test site (GLCTS) initiative were processed to simulate the Moderate Resolution Imaging Spectroradiometer (MODIS), global vegetation index imagery at 250 m pixel size resolution. The sites included boreal forest, temperature coniferous forest, temperate deciduous forest, tropical rainforest, grassland, savanna, and desert biomes. Differences and similarities in sensitivity to vegetation conditions were compared among various spectral vegetation indices (VIs). All VIs showed a qualitative relationship to variations in vegetation. However, there were significant differences among the VIs over desert, grassland, and forested biomes. The normalized difference vegetation index (NDVI) was sensitive to and responded primarily to the highly absorbing red reflectance band, while other indices such as the soil and atmosphere resistant vegetation index (SARVI) were more responsive to variations in the near-infrared (NIR) band. As a result, we found the NDVI to mimic red reflectances and saturate over the forested sites while the SARVI, by contrast, did not saturate and followed variations in NIR reflectances. In the arid and semiarid biomes, the NDVI was much more sensitive to canopy background variations than the SARVI. Maximum differences among vegetation index behavior occurred over the evergreen needleleaf forest sites relative to the deciduous broadleaf forests and drier, grassland, and shrub sites. These differences appear to be useful in complementing the NDVI for improved monitoring of vegetation, with the NDVI sensitive to fraction of absorbed photosynthetic active radiation and the SARVI more sensitive to structural canopy parameters such as leaf area index and leaf morphology.

AB - A set of Landsat Thematic Mapper images representing a wide range of vegetation conditions from the NASA Landsat Pathfinder, global land cover test site (GLCTS) initiative were processed to simulate the Moderate Resolution Imaging Spectroradiometer (MODIS), global vegetation index imagery at 250 m pixel size resolution. The sites included boreal forest, temperature coniferous forest, temperate deciduous forest, tropical rainforest, grassland, savanna, and desert biomes. Differences and similarities in sensitivity to vegetation conditions were compared among various spectral vegetation indices (VIs). All VIs showed a qualitative relationship to variations in vegetation. However, there were significant differences among the VIs over desert, grassland, and forested biomes. The normalized difference vegetation index (NDVI) was sensitive to and responded primarily to the highly absorbing red reflectance band, while other indices such as the soil and atmosphere resistant vegetation index (SARVI) were more responsive to variations in the near-infrared (NIR) band. As a result, we found the NDVI to mimic red reflectances and saturate over the forested sites while the SARVI, by contrast, did not saturate and followed variations in NIR reflectances. In the arid and semiarid biomes, the NDVI was much more sensitive to canopy background variations than the SARVI. Maximum differences among vegetation index behavior occurred over the evergreen needleleaf forest sites relative to the deciduous broadleaf forests and drier, grassland, and shrub sites. These differences appear to be useful in complementing the NDVI for improved monitoring of vegetation, with the NDVI sensitive to fraction of absorbed photosynthetic active radiation and the SARVI more sensitive to structural canopy parameters such as leaf area index and leaf morphology.

UR - http://www.scopus.com/inward/record.url?scp=0030613490&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0030613490&partnerID=8YFLogxK

U2 - 10.1016/S0034-4257(96)00112-5

DO - 10.1016/S0034-4257(96)00112-5

M3 - Article

AN - SCOPUS:0030613490

VL - 59

SP - 440

EP - 451

JO - Remote Sensing of Environment

JF - Remote Sensing of Environment

SN - 0034-4257

IS - 3

ER -