Development and application of ocean color algorithms for estimating particulate organic carbon in the Southern Ocean from satellite observations

Empirical algorithms have been developed for estimating surface concentration of particulate organic carbon (POC) from remotely-sensed ocean color in the Southern Ocean using field data POC, spectral remote-sensing reflectance, R/rs[lambda], and the inherent optical properties (IOPs) of seawater. Se...

Full description

Bibliographic Details
Main Author: Allison, David Benjamin
Format: Other/Unknown Material
Language:unknown
Published: eScholarship, University of California 2010
Subjects:
Online Access:https://escholarship.org/uc/item/45q7q67b
id ftcdlib:oai:escholarship.org/ark:/13030/qt45q7q67b
record_format openpolar
spelling ftcdlib:oai:escholarship.org/ark:/13030/qt45q7q67b 2023-05-15T18:24:55+02:00 Development and application of ocean color algorithms for estimating particulate organic carbon in the Southern Ocean from satellite observations Allison, David Benjamin 2010-01-01 application/pdf https://escholarship.org/uc/item/45q7q67b unknown eScholarship, University of California qt45q7q67b https://escholarship.org/uc/item/45q7q67b public UCSD Dissertations Academic Oceanography. (Discipline) etd 2010 ftcdlib 2020-06-06T07:56:28Z Empirical algorithms have been developed for estimating surface concentration of particulate organic carbon (POC) from remotely-sensed ocean color in the Southern Ocean using field data POC, spectral remote-sensing reflectance, R/rs[lambda], and the inherent optical properties (IOPs) of seawater. Several algorithm formulations have been considered. The best algorithm performance was obtained for the power function fit POC (mg m⁻³) = 189.29 [R/ rs(443)/R/rs(555)]⁻⁰·⁸⁷ with mean bias of 3%, normalized mean square error 7%, and determination coefficient 0.93. Analysis of match-up comparisons between satellite-derived and in situ POC support application of this algorithm in the Southern Ocean. The bio-optical relationships on which the POC algorithms are based exhibit significant variability mainly due to differing particulate assemblages. To quantify the sources of this variability, Mie scattering modeling and empirical data were used to calculate IOPs, POC, and chlorophyll-a content for 21 representative classes of particles. These classes represent colloids, organic detritus, minerals, and various plankton species. By using this reductionist approach, 38 different bulk models of seawater were constructed and analyzed. The utility of this approach in advancing an understanding of variability in the POC algorithms is shown; for example, the relationship between POC and particulate backscattering is investigated. The POC retrieval algorithm based on the reflectance band ratio was applied to SeaWiFS satellite data to demonstrate seasonal and interannual variability in POC in the Southern Ocean (south of 35°S) from 1997 through 2007. Typically the surface POC concentrations range from 30 to 120 mg mg m⁻³ while the monthly means range from 70-80 mg m⁻³. The seasonal maximum stock of POC (0.6 Pg) integrated within the top 100 m of the ocean occurs in December. The seasonal range of area-normalized POC is 5.5 - 6.6 g m⁻². The region south of 55°S provides a dominant contribution to the accumulation of POC during the productive period of the season. During the austral spring, the area-normalized POC accumulates in these high-latitude waters at rates from about 0.2 to 0.7 g m⁻² month⁻¹. The comparison of these rates with large-scale satellite-based estimates of net primary production indicates that only a small fraction (<10%) of production accumulates as POC Other/Unknown Material Southern Ocean University of California: eScholarship Southern Ocean Austral Lambda ENVELOPE(-62.983,-62.983,-64.300,-64.300)
institution Open Polar
collection University of California: eScholarship
op_collection_id ftcdlib
language unknown
topic UCSD Dissertations
Academic Oceanography. (Discipline)
spellingShingle UCSD Dissertations
Academic Oceanography. (Discipline)
Allison, David Benjamin
Development and application of ocean color algorithms for estimating particulate organic carbon in the Southern Ocean from satellite observations
topic_facet UCSD Dissertations
Academic Oceanography. (Discipline)
description Empirical algorithms have been developed for estimating surface concentration of particulate organic carbon (POC) from remotely-sensed ocean color in the Southern Ocean using field data POC, spectral remote-sensing reflectance, R/rs[lambda], and the inherent optical properties (IOPs) of seawater. Several algorithm formulations have been considered. The best algorithm performance was obtained for the power function fit POC (mg m⁻³) = 189.29 [R/ rs(443)/R/rs(555)]⁻⁰·⁸⁷ with mean bias of 3%, normalized mean square error 7%, and determination coefficient 0.93. Analysis of match-up comparisons between satellite-derived and in situ POC support application of this algorithm in the Southern Ocean. The bio-optical relationships on which the POC algorithms are based exhibit significant variability mainly due to differing particulate assemblages. To quantify the sources of this variability, Mie scattering modeling and empirical data were used to calculate IOPs, POC, and chlorophyll-a content for 21 representative classes of particles. These classes represent colloids, organic detritus, minerals, and various plankton species. By using this reductionist approach, 38 different bulk models of seawater were constructed and analyzed. The utility of this approach in advancing an understanding of variability in the POC algorithms is shown; for example, the relationship between POC and particulate backscattering is investigated. The POC retrieval algorithm based on the reflectance band ratio was applied to SeaWiFS satellite data to demonstrate seasonal and interannual variability in POC in the Southern Ocean (south of 35°S) from 1997 through 2007. Typically the surface POC concentrations range from 30 to 120 mg mg m⁻³ while the monthly means range from 70-80 mg m⁻³. The seasonal maximum stock of POC (0.6 Pg) integrated within the top 100 m of the ocean occurs in December. The seasonal range of area-normalized POC is 5.5 - 6.6 g m⁻². The region south of 55°S provides a dominant contribution to the accumulation of POC during the productive period of the season. During the austral spring, the area-normalized POC accumulates in these high-latitude waters at rates from about 0.2 to 0.7 g m⁻² month⁻¹. The comparison of these rates with large-scale satellite-based estimates of net primary production indicates that only a small fraction (<10%) of production accumulates as POC
format Other/Unknown Material
author Allison, David Benjamin
author_facet Allison, David Benjamin
author_sort Allison, David Benjamin
title Development and application of ocean color algorithms for estimating particulate organic carbon in the Southern Ocean from satellite observations
title_short Development and application of ocean color algorithms for estimating particulate organic carbon in the Southern Ocean from satellite observations
title_full Development and application of ocean color algorithms for estimating particulate organic carbon in the Southern Ocean from satellite observations
title_fullStr Development and application of ocean color algorithms for estimating particulate organic carbon in the Southern Ocean from satellite observations
title_full_unstemmed Development and application of ocean color algorithms for estimating particulate organic carbon in the Southern Ocean from satellite observations
title_sort development and application of ocean color algorithms for estimating particulate organic carbon in the southern ocean from satellite observations
publisher eScholarship, University of California
publishDate 2010
url https://escholarship.org/uc/item/45q7q67b
long_lat ENVELOPE(-62.983,-62.983,-64.300,-64.300)
geographic Southern Ocean
Austral
Lambda
geographic_facet Southern Ocean
Austral
Lambda
genre Southern Ocean
genre_facet Southern Ocean
op_relation qt45q7q67b
https://escholarship.org/uc/item/45q7q67b
op_rights public
_version_ 1766205930583621632