Publication:
Evaluation and development of bio-optical algorithms for chlorophyll retrieval in western Puerto Rico

dc.contributor.advisor Armstrong, Roy A.
dc.contributor.author Rosado Torres, Marcos A.
dc.contributor.college College of Arts and Sciences - Sciences en_US
dc.contributor.committee García Sais, Jorge R.
dc.contributor.committee Gilbes Santaella, Fernando
dc.contributor.committee Ramírez Beltrán, Nazario D.
dc.contributor.committee Aponte, Nilda E.
dc.contributor.department Department of Marine Sciences en_US
dc.contributor.representative Macchiavelli, Raúl
dc.date.accessioned 2018-12-04T17:05:57Z
dc.date.available 2018-12-04T17:05:57Z
dc.date.issued 2008
dc.description.abstract A series of cruises were carried out in Mayagüez Bay from January, 1997, to January, 2004, with the purpose of studying phytoplankton dynamics, optically active water components and ultimately developing bio-optical algorithms for the estimation of chlorophyll a (Chl a). Fourth derivative spectroscopic analyses were applied to phytoplankton absorption spectra in order to determine phytoplankton pigments that could be identified as chemotaxonomic markers and to quantify Chl a. The spectral maxima found in the fourth derivative spectra were associated with chlorophylls a, c and carotenoid pigments, probably fucoxanthin. Good correlation was found between in situ Chl a concentration and the fourth derivative peaks associated to photosynthetic pigments. These results suggest that diatoms are the dominant phytoplankton group in Mayagüez Bay. Surface Rrs spectra were modeled with Hydrolight using a four component case 2 model with in situ data for input. The moderate agreement of the model for some of the data was attributed to the error associated with using TSS as a proxy for mineral concentrations. The effect of the bio-optical components concentration on the Rrs curve was then evaluated using Hydrolight. It was determined that for the range of Chl a values normally observed in Mayaguez Bay (0.1 to 1 μg l-1), mineral values over 5 mg l-1 were sufficient to mask the Chl a signal. These results were used to tentatively set a TSS threshold of 5 mg l-1 over which Chl a concentrations cannot be derived from the Rrs spectra. Three empirical algorithms were developed for Mayagüez Bay. When tested with a data set different that the one used for their development, the fourth order polynomial and the red band ratio algorithms produced a 120.15 and 79.00 mean percent error in Chl a estimation respectively. The quasi-analytical algorithm (QAA) from Lee et al. (2002) was also evaluated with Mayagüez Bay data. The QAA performed poorly when the derived inherent optical properties were compared to measured data. The poor performance of the QAA was attributed to its inadequacy in the modeling of the spectral particle backscattering coefficients for Mayagüez Bay. en_US
dc.description.abstract Una serie de cruceros se llevaron a cabo en la Bahía de Mayagüez desde enero de 1997 hasta enero de 2004 con el propósito de estudiar la dinámica de fitoplancton, los componentes ópticamente activos y finalmente desarrollar algoritmos bio-ópticos para estimar clorofila a (Chl a). Análisis espectroscópico de cuarta derivada fueron aplicados a los espectros de absorción de fitoplancton para determinar pigmentos fitoplanctónicos que pudiesen ser identificados como marcadores quimiotaxonómicos. Los máximos espectrales encontrados en las curvas de cuarta derivada fueron asociados con clorofila a, c y pigmentos carotenoides, probablemente fucoxantina. Se encontró una buena correlación entre la concentración de Chl a in situ y los máximos de cuarta derivada asociados con pigmentos fotosintéticos. Estos resultados sugieren que las diatomeas son el grupo fitoplanctónico dominante en la Bahía de Mayagüez. Espectros de superficie de Rrs fueron modelados con Hydrolight usando un modelo caso 2 de cuatro componentes con datos in situ como insumo. El ajuste moderado del modelo para algunos de los datos fue atribuido al error asociado con usar TSS como estimador de concentración de minerales. El efecto de estos componentes en la curva de reflectancia fue evaluado usando Hydrolight. Se determinó que para el intervalo de valores de Chl a observados normalmente en la Bahía de Mayagüez (0.1 to 1 μg l-1), concentraciones de minerales mayores de 5 mg l-1 fueron suficientes para enmascarar la señal de Chl a. Estos resultados fueron usados para fijar un umbral de TSS de 5 mg l-1 sobre el cual la señal de Chl a no se puede extraer de la curva de Rrs. Tres algoritmos empíricos fueron desarrollados con los datos de la Bahía de Mayagüez. Dos de ellos produjeron un por ciento de error promedio de 120.15 y 79.00 respectivamente. El algoritmo cuasi-analítico (QAA) de Lee et al. (2002) también fue evaluado con datos de Mayagüez. El algoritmo QAA se desempeñó pobremente cuando las propiedades ópticas derivadas fueron comparadas con las medidas de campo. El pobre desempeño del QAA se atribuyó a su incapacidad para modelar los coeficientes espectrales de retrodispersión de partículas en la Bahía de Mayagüez. en_US
dc.description.graduationSemester Spring en_US
dc.description.graduationYear 2008 en_US
dc.description.sponsorship NASA Graduate Student Researchers Program grant NGT 13-52746; NASA grants NAG13-99029, NAG13-54, and NCC5-518; NOAA Cooperative Remote Sensing Science and Technology Center (CREST) program under grant number NA06OAR4810162 en_US
dc.identifier.uri https://hdl.handle.net/20.500.11801/1581
dc.language.iso en en_US
dc.rights.holder (c) 2008 Marcos A. Rosado Torres en_US
dc.rights.license All rights reserved en_US
dc.subject Phytoplankton analysis - Mayagüez Bay (P.R.) en_US
dc.subject Bio-optical algorithms – Development – Chlorophyll – Analysis - Mayagüez Bay (P.R.) en_US
dc.subject Phytoplankton pigments - Chemotaxonomy - Mayagüez Bay (P.R.) en_US
dc.subject.lcsh Seawater--Puerto Rico--Mayagüez--Optical properties en_US
dc.subject.lcsh Hyperspectral imaging en_US
dc.subject.lcsh Ocean color--Puerto Rico--Mayagüez en_US
dc.title Evaluation and development of bio-optical algorithms for chlorophyll retrieval in western Puerto Rico en_US
dc.type Dissertation en_US
dspace.entity.type Publication
thesis.degree.discipline Marine Sciences en_US
thesis.degree.level Ph.D. en_US
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
CIMA_RosadoTorresMA_2008.pdf
Size:
3.28 MB
Format:
Adobe Portable Document Format
Description:
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
1.64 KB
Format:
Item-specific license agreed upon to submission
Description: