Effect of bio-optical parameter variability and uncertainties in reflectance measurements on the remote estimation of chlorophyll-a concentration in turbid productive waters: Modeling results

Giorgio Dall'Olmo, Anatoly A. Gitelson

Research output: Contribution to journalArticle

100 Citations (Scopus)

Abstract

Most algorithms for retrieving chlorophyll-a concentration (Chla) from reflectance spectra assume that bio-optical parameters such as the phytoplankton specific absorption coefficient (aφ*) or the chlorophyll-a fluorescence quantum yield (η) are constant. Yet there exist experimental data showing large ranges of variability for these quantities. The main objective of this study was to analyze the sensitivity of two Chla algorithms to variations in bio-optical parameters and to uncertainties in reflectance measurements. These algorithms are specifically designed for turbid productive waters and are based on red and near-infrared reflectances. By means of simulated data, it is shown that the spectral regions where the algorithms are maximally sensitive to Chla overlap those of maximal sensitivity to variations in the above bio-optical parameters. Thus, to increase the accuracy of Chla retrieval, we suggest using spectral regions where the algorithms are less sensitive to Chla, but also less sensitive to these interferences. a φ appeared to be one of the most important sources of error for retrieving Chla. However, when the phytoplankton backscattering coefficient (bb,φ) dominates the total backscattering, as is likely during algal blooms, variations in the specific bb,φ may introduce large systematic uncertainties in Chla estimation. Also, uncertainties in reflectance measurements, which are due to incomplete atmospheric correction or reflected skylight removal, seem to affect considerably the accuracy of Chla estimation. Instead, variations in other bio-optical parameters, such as η or the specific backscattering coefficient of total suspended particles, appear to have minor importance. Suggestions regarding the optimal band locations to be used in the above algorithms are finally provided.

Original languageEnglish (US)
Pages (from-to)3577-3592
Number of pages16
JournalApplied optics
Volume45
Issue number15
DOIs
StatePublished - May 20 2006

Fingerprint

Reflectometers
chlorophylls
Chlorophyll
reflectance
water
Water
Backscattering
backscattering
phytoplankton
Phytoplankton
Uncertainty
atmospheric correction
sensitivity
Quantum yield
algae
coefficients
retrieval
suggestion
absorptivity
Fluorescence

ASJC Scopus subject areas

  • Atomic and Molecular Physics, and Optics
  • Engineering (miscellaneous)
  • Electrical and Electronic Engineering

Cite this

Effect of bio-optical parameter variability and uncertainties in reflectance measurements on the remote estimation of chlorophyll-a concentration in turbid productive waters : Modeling results. / Dall'Olmo, Giorgio; Gitelson, Anatoly A.

In: Applied optics, Vol. 45, No. 15, 20.05.2006, p. 3577-3592.

Research output: Contribution to journalArticle

@article{fdd19bbe0d1b49ccbc05550fab8f272a,
title = "Effect of bio-optical parameter variability and uncertainties in reflectance measurements on the remote estimation of chlorophyll-a concentration in turbid productive waters: Modeling results",
abstract = "Most algorithms for retrieving chlorophyll-a concentration (Chla) from reflectance spectra assume that bio-optical parameters such as the phytoplankton specific absorption coefficient (aφ*) or the chlorophyll-a fluorescence quantum yield (η) are constant. Yet there exist experimental data showing large ranges of variability for these quantities. The main objective of this study was to analyze the sensitivity of two Chla algorithms to variations in bio-optical parameters and to uncertainties in reflectance measurements. These algorithms are specifically designed for turbid productive waters and are based on red and near-infrared reflectances. By means of simulated data, it is shown that the spectral regions where the algorithms are maximally sensitive to Chla overlap those of maximal sensitivity to variations in the above bio-optical parameters. Thus, to increase the accuracy of Chla retrieval, we suggest using spectral regions where the algorithms are less sensitive to Chla, but also less sensitive to these interferences. a φ appeared to be one of the most important sources of error for retrieving Chla. However, when the phytoplankton backscattering coefficient (bb,φ) dominates the total backscattering, as is likely during algal blooms, variations in the specific bb,φ may introduce large systematic uncertainties in Chla estimation. Also, uncertainties in reflectance measurements, which are due to incomplete atmospheric correction or reflected skylight removal, seem to affect considerably the accuracy of Chla estimation. Instead, variations in other bio-optical parameters, such as η or the specific backscattering coefficient of total suspended particles, appear to have minor importance. Suggestions regarding the optimal band locations to be used in the above algorithms are finally provided.",
author = "Giorgio Dall'Olmo and Gitelson, {Anatoly A.}",
year = "2006",
month = "5",
day = "20",
doi = "10.1364/AO.45.003577",
language = "English (US)",
volume = "45",
pages = "3577--3592",
journal = "Applied Optics",
issn = "1559-128X",
publisher = "The Optical Society",
number = "15",

}

TY - JOUR

T1 - Effect of bio-optical parameter variability and uncertainties in reflectance measurements on the remote estimation of chlorophyll-a concentration in turbid productive waters

T2 - Modeling results

AU - Dall'Olmo, Giorgio

AU - Gitelson, Anatoly A.

PY - 2006/5/20

Y1 - 2006/5/20

N2 - Most algorithms for retrieving chlorophyll-a concentration (Chla) from reflectance spectra assume that bio-optical parameters such as the phytoplankton specific absorption coefficient (aφ*) or the chlorophyll-a fluorescence quantum yield (η) are constant. Yet there exist experimental data showing large ranges of variability for these quantities. The main objective of this study was to analyze the sensitivity of two Chla algorithms to variations in bio-optical parameters and to uncertainties in reflectance measurements. These algorithms are specifically designed for turbid productive waters and are based on red and near-infrared reflectances. By means of simulated data, it is shown that the spectral regions where the algorithms are maximally sensitive to Chla overlap those of maximal sensitivity to variations in the above bio-optical parameters. Thus, to increase the accuracy of Chla retrieval, we suggest using spectral regions where the algorithms are less sensitive to Chla, but also less sensitive to these interferences. a φ appeared to be one of the most important sources of error for retrieving Chla. However, when the phytoplankton backscattering coefficient (bb,φ) dominates the total backscattering, as is likely during algal blooms, variations in the specific bb,φ may introduce large systematic uncertainties in Chla estimation. Also, uncertainties in reflectance measurements, which are due to incomplete atmospheric correction or reflected skylight removal, seem to affect considerably the accuracy of Chla estimation. Instead, variations in other bio-optical parameters, such as η or the specific backscattering coefficient of total suspended particles, appear to have minor importance. Suggestions regarding the optimal band locations to be used in the above algorithms are finally provided.

AB - Most algorithms for retrieving chlorophyll-a concentration (Chla) from reflectance spectra assume that bio-optical parameters such as the phytoplankton specific absorption coefficient (aφ*) or the chlorophyll-a fluorescence quantum yield (η) are constant. Yet there exist experimental data showing large ranges of variability for these quantities. The main objective of this study was to analyze the sensitivity of two Chla algorithms to variations in bio-optical parameters and to uncertainties in reflectance measurements. These algorithms are specifically designed for turbid productive waters and are based on red and near-infrared reflectances. By means of simulated data, it is shown that the spectral regions where the algorithms are maximally sensitive to Chla overlap those of maximal sensitivity to variations in the above bio-optical parameters. Thus, to increase the accuracy of Chla retrieval, we suggest using spectral regions where the algorithms are less sensitive to Chla, but also less sensitive to these interferences. a φ appeared to be one of the most important sources of error for retrieving Chla. However, when the phytoplankton backscattering coefficient (bb,φ) dominates the total backscattering, as is likely during algal blooms, variations in the specific bb,φ may introduce large systematic uncertainties in Chla estimation. Also, uncertainties in reflectance measurements, which are due to incomplete atmospheric correction or reflected skylight removal, seem to affect considerably the accuracy of Chla estimation. Instead, variations in other bio-optical parameters, such as η or the specific backscattering coefficient of total suspended particles, appear to have minor importance. Suggestions regarding the optimal band locations to be used in the above algorithms are finally provided.

UR - http://www.scopus.com/inward/record.url?scp=33745768370&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33745768370&partnerID=8YFLogxK

U2 - 10.1364/AO.45.003577

DO - 10.1364/AO.45.003577

M3 - Article

C2 - 16708105

AN - SCOPUS:33745768370

VL - 45

SP - 3577

EP - 3592

JO - Applied Optics

JF - Applied Optics

SN - 1559-128X

IS - 15

ER -