sid.inpe.br/mtc-m21b/2016/06.08.16.27-TDI BIO-OPTICAL CHARACTERIZATION OF AMAZON FLOODPLAIN LAKES AND EVALUATION OF THE RETRIEVAL OF OPTICALLY ACTIVE CONSTITUENT USING REMOTE SENSING Lino Augusto Sander de Carvalho Doctorate Thesis Course Graduate in Remote Sensing, guided by Drs. Cláudio Clemente Faria Barbosa and Evlyn Márcia Leão de Moraes Novo, approved in April 18, 2016. URL of the original document: <http://urlib.net/ 8JMKD3MGP3W34P/3LRECPB> INPE São José dos Campos 2016 PUBLISHED BY: Instituto Nacional de Pesquisas Espaciais - INPE Gabinete do Diretor (GB) Serviço de Informação e Documentação (SID) Caixa Postal 515 - CEP 12.245-970 São José dos Campos - SP - Brasil Tel.:(012) 3208-6923/6921 Fax: (012) 3208-6919 E-mail: [email protected] COMMISSION OF BOARD OF PUBLISHING AND PRESERVATION OF INPE INTELLECTUAL PRODUCTION (DE/DIR-544): Chairperson: Maria do Carmo de Andrade Nono - Conselho de Pós-Graduação (CPG) Members: Dr. Plínio Carlos Alvalá - Centro de Ciência do Sistema Terrestre (CST) Dr. André de Castro Milone - Coordenação de Ciências Espaciais e Atmosféricas (CEA) Dra. Carina de Barros Melo - Coordenação de Laboratórios Associados (CTE) Dr. Evandro Marconi Rocco - Coordenação de Engenharia e Tecnologia Espacial (ETE) Dr. Hermann Johann Heinrich Kux - Coordenação de Observação da Terra (OBT) Dr. Marley Cavalcante de Lima Moscati - Centro de Previsão de Tempo e Estudos Climáticos (CPT) Silvia Castro Marcelino - Serviço de Informação e Documentação (SID) DIGITAL LIBRARY: Dr. Gerald Jean Francis Banon Clayton Martins Pereira - Serviço de Informação e Documentação (SID) DOCUMENT REVIEW: Simone Angélica Del Ducca Barbedo - Serviço de Informação e Documentação (SID) Yolanda Ribeiro da Silva Souza - Serviço de Informação e Documentação (SID) ELECTRONIC EDITING: Marcelo de Castro Pazos - Serviço de Informação e Documentação (SID) André Luis Dias Fernandes - Serviço de Informação e Documentação (SID) sid.inpe.br/mtc-m21b/2016/06.08.16.27-TDI BIO-OPTICAL CHARACTERIZATION OF AMAZON FLOODPLAIN LAKES AND EVALUATION OF THE RETRIEVAL OF OPTICALLY ACTIVE CONSTITUENT USING REMOTE SENSING Lino Augusto Sander de Carvalho Doctorate Thesis Course Graduate in Remote Sensing, guided by Drs. Cláudio Clemente Faria Barbosa and Evlyn Márcia Leão de Moraes Novo, approved in April 18, 2016. URL of the original document: <http://urlib.net/ 8JMKD3MGP3W34P/3LRECPB> INPE São José dos Campos 2016 Cataloging in Publication Data Carvalho, Lino Augusto Sander de. C253b Bio-optical characterization of Amazon floodplain lakes and evaluation of the retrieval of optically active constituent using remote sensing / Lino Augusto Sander de Carvalho. – São José dos Campos : INPE, 2016. xxvi + 175 p. ; (sid.inpe.br/mtc-m21b/2016/06.08.16.27-TDI) Thesis (Doctorate in Remote Sensing) – Instituto Nacional de Pesquisas Espaciais, São José dos Campos, 2016. Guiding : Drs. Cláudio Clemente Faria Barbosa and Evlyn Márcia Leão de Moraes Novo. 1. Amazon floodplain lakes. 2. Hydrological optics. 3. Lakes Biooptical Properties. 4. Semi-Analytical modeling. I.Title. CDU 528.8:551.435.12(811) Esta obra foi licenciada sob uma Licença Creative Commons Atribuição-NãoComercial 3.0 Não Adaptada. This work is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported License. ii Título: “ BIO-OPTICAL CHARACTERIZATION OF AMAZON FLOODPLAIN LAKES AND EVALUATION OF THE RETRIEVAL OF OPTICALLY ACTIVE CONSTITUENT USING REMOTE SENSING”. ACKNOWLEDGEMENTS Aos Orientadores Cláudio Clemente Faria Barbosa e Evlyn Márcia Leão de Moraes Novo pela valiosa orientação, suporte, crédito depositado em mim e pela oportu- nidade de trabalho em uma área repleta de desafios. Ao Programa de Pós-Graduação no INPE, professores e funcionários, pela oportu- nidade de estudos e utilização de suas instalações. Ao Professor Emmanuel Boss e à Universidade de Maine pelo período de doutorado sanduiche. À Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) que proveu a bolsa de estudos. Ao Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) através do Programa Ciência sem Fronteiras que me possibilitou o doutorado sanduiche. À FAPESP e ao Programa de Pós-Graduação em Sensoriamento Remoto pelo su- porte financeiro aos trabalhos de campo para coletas de dados. À NASA, pelo suporte financeiro e à Curtis Mobley e Emmanuel Boss pela indicação para a participação no curso Calibration and Validation for Ocean Color Remote Sensing nos anos de 2013 e 2015. À todos os meus amigos pelos bons momentos vividos. À meus pais, por todo o carinho e suporte. À minha esposa, pelo carinho e paciência. v ABSTRACT Amazon floodplain lakes play a substantial role in global and regional biogeochemical Amazonian processes. Due to their size, sampling strategies usually applied in limno- logical studies are not suitable and therefore, Remote Sensing (RS) techniques figure as an alternative due to the high temporal and synoptic characteristics. However, the use of RS demands precise lake bio-optical characterization in order to provide reliable estimates of optical active components (OAC).This work focused on the study of Curuai Lake which is a suitable example of a Brazilian Amazon floodplain lake. Curuai lake was sampled in four field campaigns (September/2012, February and August/2013 and April/2014) where Apparent Optical Properties-AOP (Rrs and K-functions), in situ Inherent Optical Properties-IOP (Attenuation, Absorp- tion, Backscattering profiles and Particle Size Distribution (PSD)) as well as labo- ratory analysis (AOC concentration and absorption) were measured. A data qual- ity assessment was performed to test the suitability of commercial instrumentation (ACS and Hydroscat) for turbid environments as well as commonly used AOP/IOP measurement methodologies. The optical characterization compared datasets from each fieldcampaign for surface and profile measurements. Also three semi-analytical inverse models (Nechad Algorithm (NECHAD et al., 2010), Quasi-Analytical Algo- rithm (QAA) (LEE et al., 2002) and Generalized ocean color inversion model (GIOP) (WERDELL et al., 2013)) were tested using measured AOP and IOPs. Data quality assessment show that sun/skyglint effects have the highest impact on above water Ab remote sensing measurements Rrs (Rrs ). Highest errors were found for In-water de- Iw rived AOPs (Rrs and K-functions), but despite the the different tested approaches their differences are commonly in the 10 to 15 % interval. For in situ IOPs, the Hy- drolight IOP/AOP closure experiments resulted in mismatches from 50 % to 100% depending on the field campaign. Among the corrections tested for ACS/Hydroscat errors, the Doxaran (DOXARAN et al., 2013) and Rottgers (RöTTGERS et al., 2013) methods were the most suitable. For Laboratory absorption measurements, differ- ences range from 10% to 80% but were below 15 % for biogeochemical variables concentration measurements. Curuai Lake Bio-optical characterization show that K-functions go from 1 to 12 m−1 in 676 nm depending on the season but can go as high as ≈ 20 m−1 in the blue range. For IOPs, total attenuation, absorption and scattering coefficients were higher for February/2013 (≈ 25, 4 and 21 m−1 re- spectively for 550 nm) and similar for August/2013 and April/2014 (≈ 12, 2.5 and 10 m−1 respectively for 550 nm). Backscattering coefficients also were higher for February/2013 (≈ 0.6 m−1 in 700 nm) and similar for August/2013 and April/2014 −1 b (≈ 0.3 m in 700 nm). However, the backscattering to scattering ratio ( bp/bp) was similar for February/2013, August/2013 and April/2014 (Median of 0.04), indicat- ing low particle type variability. However, differences were found for the exponent ∗ of scattering coefficient (γb) (Median = 1.3 to 1.5) and for specific scattering bp (0.42 to 0.67 m2g−1) showing a possible dominance of smaller inorganic particles for February/2013. Specific absorption coefficients for particulate (ap), non-algal par- ticles (anap) and phytoplankton (aφ) were within literature values but particularly vii −1 high for anap ( up to ≈ 6 m in the blue range). Particle Size Distribution shows a bimodal behavior with most of particles in the 10-15 µm range. Results for In- verse Models were not satisfactory for most of retrieved IOPs and biogeochemical variables (Chla and TSM). Best results for GIOP presented errors of at least 20 % while for QAA errors can reach more than 100 %. The Nechad algorithm shows errors that reaches 30 % for TSM estimates which may be considered reasonable. The main source of errors for semi-analytical inverse models rely on uncertainties of input parameters but further investigation is needed to understand the main source of error in the current dataset as well as help guiding future data acquisition. Keywords: Amazon floodplain lakes. Hydrological Optics. Lakes Biooptical Proper- ties. Semi-Analytical modeling. viii CARACTERIZAÇÃO BIO-OPTICA DE LAGOS DA PLANÍCIE INUNDÁVEL AMAZÔNICA E AVALIAÇÃO DA RECUPERAÇÃO DE CONSTITUINTES OPTICAMENTE ATIVOS UTILIZANDO SENSORIAMENTO REMOTO RESUMO Lagos de várzea amazônico desempenha um papel importante nos processos biogeo- químicos globais e regionais da Amazonia. Devido ao seu tamanho, estratégias de amostragem normalmente aplicados em estudos limnológicos não são adequadas e, portanto, técnicas de sensoriamento remoto (SR) se apresentam como uma alterna- tiva devido às sua alta resolução temporal e visão sinóptica. No entanto, o uso de SR exige uma caracterização bio-óptica precisa, a fim de fornecer estimativas confiáveis de componentes ativos ópticos (COA). Este trabalho teve como