Sensitivity of common vegetation indices to the canopy structure of field crops

Xiaochen Zou, Matti Mõttus

Research output: Contribution to journalArticleScientificpeer-review

4 Citations (Scopus)

Abstract

Leaf inclination angle distribution is an important canopy structure characteristic which directly impacts the fraction of the intercepted solar radiation. Together with the leaf area index (LAI) it determines the structure and fractional cover of a homogeneous crop canopy. Unfortunately, this key canopy parameter has usually been ignored when applying common multispectral vegetation indices to the mapping of LAI, although its impact is known from model simulations. Therefore, we measured leaf angles and determined their distribution (quantified using the leaf mean tilt angle, MTA) for six crop species with different structures growing on 162 plots with a broad range of LAI (1.1-5.0) and leaf chlorophyll content (26-94 μg cm-2). Next, we calculated six vegetation indices widely used for LAI monitoring-the normalized difference vegetation index (NDVI), the enhanced vegetation index (EVI), the two band enhanced vegetation index (EVI2), the modified triangular vegetation index (MTVI2), the optimized soil adjusted vegetation index (OSAVI) and the wide dynamic range vegetation index (WDRVI)-from airborne imaging spectroscopy data. We calculated the Spearman's correlation coefficient Rs, a non-parametric statistic chosen because of the non-normal distribution of canopy parameters. All studied indices depended on the LAI (0.50 ≤ Rs ≤ 0.71), but the dependence on the MTA was of similar magnitude (-0.83 ≤ Rs ≤ -0.53) with EVI, EVI2, OSAVI and MTVI2 depending more strongly on MTA than on LAI. All studied indices were good proxies (0.78 ≤ Rs ≤ 0.88) for vegetation fractional cover (Fcover) which, for homogeneous crop canopies, is a nonlinear function of LAI and MTA. EVI2 and MTVI2 were the most strongly correlated with Fcover, although the difference to the other studied indices was small. This first study involving a large range of crop structures confirms the results from canopy reflectance simulations and emphasizes the necessity of leaf angle information for the successful mapping of LAI with Earth observation data.

Original languageEnglish
Article number994
JournalRemote Sensing
Volume9
Issue number10
DOIs
Publication statusPublished - 1 Oct 2017
MoE publication typeA1 Journal article-refereed

Fingerprint

vegetation index
leaf area index
canopy
crop
tilt
canopy reflectance
NDVI
simulation
solar radiation
chlorophyll
soil
spectroscopy
vegetation
monitoring

Keywords

  • Canopy structure
  • EVI
  • EVI2
  • Imaging spectroscopy
  • Leaf angle distribution
  • Mean tilt angle
  • MTVI2
  • NDVI
  • OSAVI
  • PROSAIL
  • WDRVI

Cite this

@article{6e7471a4e28040d397a061d458ab7c42,
title = "Sensitivity of common vegetation indices to the canopy structure of field crops",
abstract = "Leaf inclination angle distribution is an important canopy structure characteristic which directly impacts the fraction of the intercepted solar radiation. Together with the leaf area index (LAI) it determines the structure and fractional cover of a homogeneous crop canopy. Unfortunately, this key canopy parameter has usually been ignored when applying common multispectral vegetation indices to the mapping of LAI, although its impact is known from model simulations. Therefore, we measured leaf angles and determined their distribution (quantified using the leaf mean tilt angle, MTA) for six crop species with different structures growing on 162 plots with a broad range of LAI (1.1-5.0) and leaf chlorophyll content (26-94 μg cm-2). Next, we calculated six vegetation indices widely used for LAI monitoring-the normalized difference vegetation index (NDVI), the enhanced vegetation index (EVI), the two band enhanced vegetation index (EVI2), the modified triangular vegetation index (MTVI2), the optimized soil adjusted vegetation index (OSAVI) and the wide dynamic range vegetation index (WDRVI)-from airborne imaging spectroscopy data. We calculated the Spearman's correlation coefficient Rs, a non-parametric statistic chosen because of the non-normal distribution of canopy parameters. All studied indices depended on the LAI (0.50 ≤ Rs ≤ 0.71), but the dependence on the MTA was of similar magnitude (-0.83 ≤ Rs ≤ -0.53) with EVI, EVI2, OSAVI and MTVI2 depending more strongly on MTA than on LAI. All studied indices were good proxies (0.78 ≤ Rs ≤ 0.88) for vegetation fractional cover (Fcover) which, for homogeneous crop canopies, is a nonlinear function of LAI and MTA. EVI2 and MTVI2 were the most strongly correlated with Fcover, although the difference to the other studied indices was small. This first study involving a large range of crop structures confirms the results from canopy reflectance simulations and emphasizes the necessity of leaf angle information for the successful mapping of LAI with Earth observation data.",
keywords = "Canopy structure, EVI, EVI2, Imaging spectroscopy, Leaf angle distribution, Mean tilt angle, MTVI2, NDVI, OSAVI, PROSAIL, WDRVI",
author = "Xiaochen Zou and Matti M{\~o}ttus",
note = "Project code: 113325",
year = "2017",
month = "10",
day = "1",
doi = "10.3390/rs9100994",
language = "English",
volume = "9",
journal = "Remote Sensing",
issn = "2072-4292",
publisher = "MDPI",
number = "10",

}

Sensitivity of common vegetation indices to the canopy structure of field crops. / Zou, Xiaochen; Mõttus, Matti.

In: Remote Sensing, Vol. 9, No. 10, 994, 01.10.2017.

Research output: Contribution to journalArticleScientificpeer-review

TY - JOUR

T1 - Sensitivity of common vegetation indices to the canopy structure of field crops

AU - Zou, Xiaochen

AU - Mõttus, Matti

N1 - Project code: 113325

PY - 2017/10/1

Y1 - 2017/10/1

N2 - Leaf inclination angle distribution is an important canopy structure characteristic which directly impacts the fraction of the intercepted solar radiation. Together with the leaf area index (LAI) it determines the structure and fractional cover of a homogeneous crop canopy. Unfortunately, this key canopy parameter has usually been ignored when applying common multispectral vegetation indices to the mapping of LAI, although its impact is known from model simulations. Therefore, we measured leaf angles and determined their distribution (quantified using the leaf mean tilt angle, MTA) for six crop species with different structures growing on 162 plots with a broad range of LAI (1.1-5.0) and leaf chlorophyll content (26-94 μg cm-2). Next, we calculated six vegetation indices widely used for LAI monitoring-the normalized difference vegetation index (NDVI), the enhanced vegetation index (EVI), the two band enhanced vegetation index (EVI2), the modified triangular vegetation index (MTVI2), the optimized soil adjusted vegetation index (OSAVI) and the wide dynamic range vegetation index (WDRVI)-from airborne imaging spectroscopy data. We calculated the Spearman's correlation coefficient Rs, a non-parametric statistic chosen because of the non-normal distribution of canopy parameters. All studied indices depended on the LAI (0.50 ≤ Rs ≤ 0.71), but the dependence on the MTA was of similar magnitude (-0.83 ≤ Rs ≤ -0.53) with EVI, EVI2, OSAVI and MTVI2 depending more strongly on MTA than on LAI. All studied indices were good proxies (0.78 ≤ Rs ≤ 0.88) for vegetation fractional cover (Fcover) which, for homogeneous crop canopies, is a nonlinear function of LAI and MTA. EVI2 and MTVI2 were the most strongly correlated with Fcover, although the difference to the other studied indices was small. This first study involving a large range of crop structures confirms the results from canopy reflectance simulations and emphasizes the necessity of leaf angle information for the successful mapping of LAI with Earth observation data.

AB - Leaf inclination angle distribution is an important canopy structure characteristic which directly impacts the fraction of the intercepted solar radiation. Together with the leaf area index (LAI) it determines the structure and fractional cover of a homogeneous crop canopy. Unfortunately, this key canopy parameter has usually been ignored when applying common multispectral vegetation indices to the mapping of LAI, although its impact is known from model simulations. Therefore, we measured leaf angles and determined their distribution (quantified using the leaf mean tilt angle, MTA) for six crop species with different structures growing on 162 plots with a broad range of LAI (1.1-5.0) and leaf chlorophyll content (26-94 μg cm-2). Next, we calculated six vegetation indices widely used for LAI monitoring-the normalized difference vegetation index (NDVI), the enhanced vegetation index (EVI), the two band enhanced vegetation index (EVI2), the modified triangular vegetation index (MTVI2), the optimized soil adjusted vegetation index (OSAVI) and the wide dynamic range vegetation index (WDRVI)-from airborne imaging spectroscopy data. We calculated the Spearman's correlation coefficient Rs, a non-parametric statistic chosen because of the non-normal distribution of canopy parameters. All studied indices depended on the LAI (0.50 ≤ Rs ≤ 0.71), but the dependence on the MTA was of similar magnitude (-0.83 ≤ Rs ≤ -0.53) with EVI, EVI2, OSAVI and MTVI2 depending more strongly on MTA than on LAI. All studied indices were good proxies (0.78 ≤ Rs ≤ 0.88) for vegetation fractional cover (Fcover) which, for homogeneous crop canopies, is a nonlinear function of LAI and MTA. EVI2 and MTVI2 were the most strongly correlated with Fcover, although the difference to the other studied indices was small. This first study involving a large range of crop structures confirms the results from canopy reflectance simulations and emphasizes the necessity of leaf angle information for the successful mapping of LAI with Earth observation data.

KW - Canopy structure

KW - EVI

KW - EVI2

KW - Imaging spectroscopy

KW - Leaf angle distribution

KW - Mean tilt angle

KW - MTVI2

KW - NDVI

KW - OSAVI

KW - PROSAIL

KW - WDRVI

UR - http://www.scopus.com/inward/record.url?scp=85032879967&partnerID=8YFLogxK

U2 - 10.3390/rs9100994

DO - 10.3390/rs9100994

M3 - Article

VL - 9

JO - Remote Sensing

JF - Remote Sensing

SN - 2072-4292

IS - 10

M1 - 994

ER -