@PhDThesis{Valerio:2018:AsSpVa,
author = "Valerio, Aline de Matos",
title = "Assessment of the spatiotemporal variability of optical and
biogeochemical parameters in the Lower Amazon region and of the
carbon content in the Amazon River continuum using in situ and
Remote Sensing data",
school = "Instituto Nacional de Pesquisas Espaciais (INPE)",
year = "2018",
address = "S{\~a}o Jos{\'e} dos Campos",
month = "2018-06-04",
keywords = "continuum do rio amazonas, par{\^a}metros
biogeoqu{\'{\i}}micos, propriedades bio-{\'o}pticas, CDOM,
balan{\c{c}}o de carbono, sensoriamento remoto da cor da
{\'a}gua, amazon river continuum, biogeochemical parameters,
bio-optical properties, CDOM, carbon budget, water colour remote
sensing.",
abstract = "O continuum do Rio Amazonas {\'e} uma regi{\~a}o determinante no
balan{\c{c}}o de carbono global, mas sua extens{\~a}o
geogr{\'a}fica dificulta sua observa{\c{c}}{\~a}o in situ.
Considerando as caracter{\'{\i}}sticas de alta
resolu{\c{c}}{\~a}o temporal e cobertura sin{\'o}ptica, o
sensoriamento remoto da cor da {\'a}gua (SRCA) representa uma
ferramenta importante para monitorar a distribui{\c{c}}{\~a}o e
variabilidade das fra{\c{c}}{\~o}es de carbono e outros
par{\^a}metros biogeoqu{\'{\i}}micos nas {\'a}guas
amaz{\^o}nicas. Entretanto, a efic{\'a}cia dos produtos gerados
por SRCA para estudar a din{\^a}mica biogeoqu{\'{\i}}mica de um
sistema aqu{\'a}tico, depende da acur{\'a}cia e precis{\~a}o
para representar as propriedades bio-{\'o}pticas da {\'a}rea
investigada. Este trabalho foi desenvolvido no continuum do Baixo
Rio Amazonas, onde uma descri{\c{c}}{\~a}o abrangente da
variabilidade espa{\c{c}}o-temporal de dados radiom{\'e}tricos,
bio-{\'o}pticos e biogeoqu{\'{\i}}micos foi realizada com o
objetivo de mapear as fra{\c{c}}{\~o}es de carbono por SRCA. Na
regi{\~a}o do Baixo Amazonas (RBA), foram feitas amostragens in
situ de reflect{\^a}ncia de sensoriamento remoto, par{\^a}metros
bio{\'o}pticos como os coeficientes de absor{\c{c}}{\~a}o pela
mat{\'e}ria org{\^a}nica colorida dissolvida (aCDOM), material
particulado total (ap), fitopl{\^a}ncton (aphy) e
part{\'{\i}}culas n{\~a}o-algais (anap), e par{\^a}metros
biogeoqu{\'{\i}}micos como o material particulado em
suspens{\~a}o (SPM), concentra{\c{c}}{\~a}o de clorofila-a,
concentra{\c{c}}{\~a}o de carbono org{\^a}nico dissolvido
(DOC), e press{\~a}o parcial do di{\'o}xido de carbono (pCO2)
para todas as esta{\c{c}}{\~o}es hidrol{\'o}gicas - enchente,
cheia, vazante e baixa, durante o per{\'{\i}}odo de 2014-2017.
Na pluma do Rio Amazonas (PRA), dados in situ de pCO2, salinidade
da superf{\'{\i}}cie do mar (SSS) e temperatura da
superf{\'{\i}}cie do mar (SST) foram adquiridos durante as
esta{\c{c}}{\~o}es de cheia, vazante e baixa durante os anos de
2010-2012. As caracter{\'{\i}}sticas bio-{\'o}pticas descritas
neste trabalho permitiram uma clara distin{\c{c}}{\~a}o entre o
corpo de {\'a}gua principal do Rio Amazonas, dominado por NAP e
CDOM, e os tribut{\'a}rios de {\'a}guas claras e dominados por
CDOM. A an{\'a}lise da variabilidade espa{\c{c}}o-temporal das
propriedades bio-{\'o}pticas evidenciou: 1) o impacto da
dilui{\c{c}}{\~a}o dos par{\^a}metros biogeoqu{\'{\i}}micos
causados pela contribui{\c{c}}{\~a}o dos tribut{\'a}rios assim
como os processos de degrada{\c{c}}{\~a}o do DOM no curso do Rio
Amazonas; 2) a homogeneidade das caracter{\'{\i}}sticas
bio-{\'o}pticas durante as esta{\c{c}}{\~o}es hidrol{\'o}gicas
de cheia, vazante e baixa, em contraste com a esta{\c{c}}{\~a}o
da enchente (caracter{\'{\i}}stica predominante do SPM); 3) a
vulnerabilidade das {\'a}guas amaz{\^o}nicas {\`a}s
condi{\c{c}}{\~o}es hidrol{\'o}gicas excepcionais; 4) a pouca
influ{\^e}ncia de processos de menor escala (por exemplo, xiv
efeito de mar{\'e}) nas caracter{\'{\i}}sticas bio-{\'o}pticas
regionais. Foram desenvolvidos algoritmos de invers{\~a}o
regionais de CDOM, (aCDOM(412) e da inclina{\c{c}}{\~a}o da
curva no intervalo do UV, S275-295), DOC e pCO2. Primeiro foram
desenvolvidas formula{\c{c}}{\~o}es emp{\'{\i}}ricas de
aCDOM(412) (N = 100, R2 = 0.67, p<0.05) e S275-295 (N = 100, R2 =
0.83, p<0.05), baseados em modelo linear multivariado e n{\~a}o
linear, respectivamente. Devido a diferentes padr{\~o}es sazonais
de DOC (clara distin{\c{c}}{\~a}o da enchente em
rela{\c{c}}{\~a}o {\`a}s demais), sua estimativa foi
particularmente complexa. A pCO2 foi estimada satisfatoriamente a
partir de uma rela{\c{c}}{\~a}o multivariada usando CDOM e
temperatura (N = 69, R2 = 0.80, p<0.05). Os modelos desenvolvidos
para estimar aCDOM(412), S275-295, DOC e pCO2, foram aplicados em
imagens sazonais do sensor orbital Medium Resolution Imaging
Spectrometer (MERIS) para os anos de 2010- 2011 para demonstrar a
din{\^a}mica das fra{\c{c}}{\~o}es de carbono nas {\'a}guas
amaz{\^o}nicas. O Rio Amazonas foi fonte de carbono durante todas
as esta{\c{c}}{\~o}es com a maior (menor) emiss{\~a}o de
carbono durante a cheia (baixa). A variabilidade intra-sazonal
destaca a forte din{\^a}mica em {\'a}reas de
transi{\c{c}}{\~a}o entre {\'a}guas de rio e oceano. Na PRA, a
pCO2 foi satisfatoriamente estimada a partir de uma
rela{\c{c}}{\~a}o multivariada usando SSS e SST (N = 76, R2=
0.74, p<0.05) e o modelo foi aplicado em dados do sensor Soil
Moisture and Ocean Salinity (SMOS) para os anos de 2010-2014. A
an{\'a}lise dos mapas da pCO2-SMOS evidenciou o impacto do
padr{\~a}o hidrol{\'o}gico na variabilidade inter e intra-anual
na pCO2. Em geral, a PRA durante as esta{\c{c}}{\~o}es de
enchente e cheia atuou como uma fonte de CO2, enquanto que durante
as esta{\c{c}}{\~o}es de vazante e baixa se comportou como um
sumidouro de CO2. Os resultados aqui apresentados demonstram que a
PRA sequestra menos carbono do que se presume atualmente e que
inclusive, pode tamb{\'e}m agir como emissor de CO2 durante
alguns per{\'{\i}}odos do ano. Este estudo enfatiza a
necessidade de se obter informa{\c{c}}{\~o}es adicionais in situ
(principalmente na {\'a}rea de transi{\c{c}}{\~a}o entre o rio
e o oceano) para refinar e melhorar a valida{\c{c}}{\~a}o dos
modelos aqui desenvolvidos e assim obter uma melhor
compreens{\~a}o do papel do continuum do Rio Amazonas no
balan{\c{c}}o global de carbono. Os m{\'e}todos propostos por
este estudo para estimar as fra{\c{c}}{\~o}es de carbono no
continuum do Rio Amazonas tem potencial para aplica{\c{c}}{\~a}o
em outros grandes rios globais, especialmente em regi{\~o}es
tropicais. ABSTRACT: The Amazon River continuum plays a crucial
role to the global carbon budget but its geographic extension
challenges in situ observations. Due to its high temporal and
synoptic coverage, the water colour remote sensing (WCRS)
represents a relevant observation tool to monitor the distribution
and variability of carbon content and other biogeochemical
parameters on the Amazon waters. However, the optimal exploitation
of the information provided by WCRS for investigating
biogeochemical dynamics of a water system relies on accurate
retrieval of bio-optical properties of the area investigated. This
work focused on the Lower Amazon River continuum where a
comprehensive description of the spatiotemporal variability of
radiometric, bio-optical and biogeochemical parameters was
performed with the aim of mapping carbon content from remote
sensing observation. In the Lower Amazon River region (LAR), in
situ sampling of remote sensing reflectance, bio-optical
parameters (absorption properties of the coloured dissolved
organic matter (aCDOM), total particulate matter (ap),
phytoplankton (aphy) and non-algal particles (anap)) and
biogeochemical parameters (suspended particulate matter (SPM),
chlorophyll-a, dissolved organic carbon (DOC) concentration, and
partial pressure of dioxide carbon (pCO2)) were acquired for all
regional hydrological seasons (i.e. rising water (RW), high water
(HW), falling water (FW) and low water (LW)), over the 2014-2017
time period. In the Amazon River plume (ARP), in situ pCO2, sea
surface salinity (SSS) and water surface temperature (SST) were
acquired during HW, FW and LW seasons during 2010-2012. The
general description of the bio-optical characteristics of the
Lower Amazon River performed from this original data set has
allowed a clear optical distinction between waters from the Amazon
mainstream (NAP and CDOM dominated) and those corresponding to the
Amazon tributaries (clear waters and CDOM dominated). The analysis
of the spatiotemporal variability of the Lower Amazon bio-optical
properties emphasized: 1) the predominant impact of Amazon
tributaries dilution on biogeochemical parameters and degradation
processes of the DOM along the Amazon course; 2) the homogeneity
in the Amazon bio-optical characteristics during the HW, FW, LW
seasons the latter contrasting with the RW season (SPM major
characteristics); 3) the sensitivity of the Amazon water to
exceptional hydrological conditions; 4) the restricted influence
of small scale processes (e.g. tidal effects) on the regional
biooptical characteristics. Regional CDOM (aCDOM(412) and CDOM
spectral slope in the UV, S275-295), DOC and pCO2 inversion
algorithms were developed. aCDOM(412) and S275-295 empirical
formulations based on a multiband linear relationship (N = 100, R2
= 0.67, p<0.05) and a nonlinear relationship (N = 100, R2 = 0.83,
p<0.05), respectively, were first developed. The DOC retrieval
from aCDOM(412) in the LAR has been shown to be relatively complex
relying on the specific consideration of the seasonal pattern in
the algorithm definition (distinction between RW and the other
seasons). pCO2 was satisfactorily retrieved from a unique
algorithm using CDOM and temperature as input parameters (N = 69,
R2 = 0.80, p<0.05). The models developed here for estimating
aCDOM(412), S275-295, DOC and pCO2 were applied on Medium
Resolution Imaging xii Spectrometer (MERIS) seasonal composite
images for the years of 2010-2011 to illustrate the spatiotemporal
dynamics of the carbon contents in the Amazon waters. Amazon River
was found to represent a source of carbon during all seasons, with
the highest (lowest) carbon export during the RW (LW). The
intra-seasonal variability underlines the strong dynamics of the
transition areas between the river and oceanic waters. In the ARP,
pCO2 was satisfactorily retrieved using SSS and SST as proxies (N
= 76, R2= 0.74, p<0.05) and the model was applied on Soil Moisture
and Ocean Salinity (SMOS) images for the years of 2010-2014. The
analysis of SMOS-based pCO2 maps has illustrated the impact of the
hydrological pattern on inter and intra-annual pCO2 variability.
The ARP during the RW and HW seasons was generally representing a
net source of CO2. Conversely, during the FW and LW seasons, the
ARP was a net sink of CO2. The latter results restricted carbon
sink area when compared to previous observations and net source of
CO2 (during some periods of the year) are particularly original.
This study emphasized the crucial need of additional in situ
information (especially in the river to ocean transition area) for
refining and better validating the models developed in this study
and thus obtain a better insight into the understanding of the
role of the Amazon River Continuum on the global carbon budget.
The methods here proposed to assess the carbon content in the
Amazon River continuum might be potentially applied to other large
river systems, especially over tropical areas.",
committee = "Novo, Evlyn Marcia Le{\~a}o de Moraes (presidente) and Kampel,
Milton (orientador) and Vantrepotte, Vincent (orientador) and
Rudorff, Nat{\'a}lia de Moraes and Alc{\^a}ntara, Enner Herenio
de and Loisel, Hubert",
englishtitle = "Avalia{\c{c}}{\~a}o da variabilidade espa{\c{c}}o-temporal dos
par{\^a}metros {\'o}pticos e biogeoqu{\'{\i}}micos na
regi{\~a}o do Baixo Amazonas e das fra{\c{c}}{\~o}es de carbono
no continuum do Rio Amazonas utilizando dados in situ e de
sensoriamento remoto",
language = "en",
pages = "237",
ibi = "8JMKD3MGP3W34R/3R677US",
url = "http://urlib.net/ibi/8JMKD3MGP3W34R/3R677US",
targetfile = "publicacao.pdf",
urlaccessdate = "06 maio 2024"
}