• No results found

Coupling of Coastal Hydrodynamics and Bio-Optical Color Components of a Tropical Estuarine Ecosystem

N/A
N/A
Protected

Academic year: 2022

Share "Coupling of Coastal Hydrodynamics and Bio-Optical Color Components of a Tropical Estuarine Ecosystem"

Copied!
214
0
0

Loading.... (view fulltext now)

Full text

(1)

AND

BIO-OPTICAL COLOR COMPONENTS OF A TROPICAL ESTUARINE ECOSYSTEM

THESIS SUBMITTED FOR THE AWARD OF THE DEGREE OF

DOCTOR OF PHILOSOPHY

IN

MARINE SCIENCE

(PHYSICAL OCEANOGRAPHY)

BY

ARJUN ADHIKARI

UNDER THE GUIDANCE OF

PROFESSOR HARILAL B. MENON

SCHOOL OF EARTH, OCEAN & ATMOSPHERIC SCIENCES GOA UNIVERSITY

GOA

FEBRUARY 2020

(2)

As required by the university ordinance OA-19(1–11) under section 24(1) of the Goa Uni- versity Act, 1984, I state that this thesis entitled Coupling of coastal hydrodynamics and bio-optical color components of a tropical estuarine ecosystem is my original contribution and none of the conducted work has been submitted elsewhere.

The literature concerned with the investigated problems has been cited comprehensibly, and any due acknowledgements are stated.

(Arjun Adhikari)

Date: February, 2020 Place: Goa University

(3)

It is certified that the thesis entitled Coupling of coastal hydrodynamics and bio-optical color components of a tropical estuarine ecosystem, submitted by Mr. Arjun Adhikari for the award of Doctor of Philosophy in Marine Science (Physical Oceanography) is his original contribution under my supervision. The thesis or any part thereof has not been previously submitted for any degree or diploma in any other University or Institution.

(Harilal B. Menon) Dean & Professor (HAG),

School of Earth, Ocean & Atmospheric Sciences, Goa University - 403206,

Taleigao Plateau, Goa INDIA.

(4)

you are my rock

(5)

–Isaac Newton

(6)

In my opinion, earning a Ph.D. is life’s utmost academic commitment that tests and helps developing one’s theoretical and experimental scientific approaches, and transforms one in to an independent researcher. It would not have been possible without the support and guidance that I received over-time from many people. First and foremost, I am earnestly grateful to my research advisor Prof. Harilal B. Menon, for all the support, encour- agement, patience, guidance, and enthusiasm, that inspired me to realize the power of critical thinking and what a hard-working scientist can accomplish. In our brainstorming discussions, he was a sounding board for all my research notions and developments, and a helpful critique in analyzing the hypothesis I put through in this thesis. I could not ask for a better advisor. He is also the dynamic core of all the logistics and preparations in retrieval of all data-sets used in this thesis.

I am sincerely thankful to my research committee, Dr. P. Vethamony (retired), Prof. V.

M. Matta, and Prof. M. K. Janarthanam. They were very generous in helping with their valuable comments and perspectives that improved my work. Especially, Dr. Vethamony’s generous help in providing ADCPs and critiquing data analysis improved my techniques.

Technical support from Mr. A. M. Almeida and Mr. P. S. Pednekar (retired) at National Institute of Oceanography, Goa, was crucial in the deployment and retrieval of moorings.

In addition, Prof. P. N. Vinayachandran and Dr. J. V. George at Indian Institute of Science, Bangalore, spared the vertical microstructure profiler for a cruise that allowed to collect crucial turbulence data set used in this thesis. With respect to microstructure mea- surements, technical support and discussions about data collection and processing with Dr. R. Lueck, Mr. E. Cervelli, and Mr. J. Hancyk at Rockland Scientific International Inc., Canada, were very insighful. During the first year of my stay at Goa University, Dr.

A. Lotliker at Indian National Centre for Ocean Information Services, Hyderabad, with

(7)

questions in a very effective way. Because of their precious support, they mean a lot to me.

Majority of the data sets used in this thesis was collected onboard ORV Sagar Manjusha with an efficient crew that made ingenious efforts in helping me collect data onboard. The conscientious assistance of Miss Nupur Phadte, Mr. Partha Patil, Miss Diksha Gaonkar and Miss Shruti Salgaonkar in data collection onboard cruise and over very many different field-surveys could not appreciated enough. They all made an invaluable contribution towards my Ph.D. I am also thankful to all the colleagues and administrative staff at the School of Earth, Ocean and Atmospheric Sciences, Goa University, for their unfailing assistance.

Finally, last but above all, I owe everything I do to my mother, sister, grandparents, and Heer for their continued and unconditional moral support in my life. You all have been my rock. My friends, with some it’s been over fifteen years, I thank you all for making me a part of your industrious lives.

I am not short of anything, but gratefulness!

You all have brought me immense happiness and I feel content.

(Arjun Adhikari)

(8)

Different studies carried out in this thesis were designed in a framework to understand and quantify the effect of climate change on optically complex waters through in-situ and satellite measurements of physical and optical oceanography. The thesis begins with the outcomes of investigations of the significance of tidal asymmetry in region over a spring- neap tidal cycle. Observations of water current structure during flood and ebb stages of tide exhibited strong asymmetry in vertical shear and stratification that evolved to dispro- portionately distribute momentum flux within the water column. Floods were relatively stronger, while the duration asymmetry of ebb flow was relatively longer indicating strati- fication evolved slowly. In accordance, the process of strain-induced-periodic-stratification (SIPS) was established as a major secondary driver of coastal circulation in region. With respect to SIPS, variations in the biological parameters such as chlorophyll-a were studied, in relation to phytoplankton bloom conditions that originate in coastal waters of region.

The biophysical interaction was strongly influenced by the bottom boundary layer (BBL), where low water current velocity during neap tide results in horizontal entrainment. Ad- ditionally, BBL was observed evolve with SIPS, such that strong horizontal and vertical flux of tracers was evident during spring tide due to the strength of straining that resulted in advection. But this scenario weakened during neap phase in low energy conditions and a constant supply of nutrients limited by the gradient Richardson number. This results in increased chlorophyll-a concentrations during the neap phase leading to a potential bloom condition. In order to accurately detect such changes in optically complex waters, a novel algorithm (Goa University Case 2, GUC2) was formulated. GUC2 was well tested with in situ regional data sets and satellite derived (MERIS) global data sets such as NO- MAD (NASA) and SATCORE (MoES). The wide applicability of GUC2 was established in Chesapeake Bay, where its accuracy in retrieving chlorophyll-a was positively validated in constantly changing biogeochemical conditions due to varying fluvial discharge rates.

(9)

coastal and estuarine waters revealed a strong relationship between particulate backscat- tering and remote sensing reflectance at 704 nm. Lastly, to evaluate the impact of climate change on global estuarine and coastal waters, optical partitioning of the World Ocean re- vealed a decadal increase in chlorophyll-a concentrations. This was compared with Global Warming Index and parameters of climate change, which revealed the weakening of global estuarine and coastal waters as a potential source of CO2 to atmosphere.

(10)

Page

List of Symbols v

List of Figures xiii

List of Tables xxii

1 Introduction 1

1.1 Tidal Straining . . . 2

1.2 Physical Controls on Phytoplankton . . . 7

1.3 Phytoplankton and Optical Complexity . . . 9

1.4 Optical Characterization of the Study Area . . . 14

1.5 Objectives and structure of the thesis . . . 15

2 Tidal Structure and Asymmetry 17 2.1 Introduction . . . 17

2.2 Study site, Dataset and Methodology . . . 20

2.2.1 Study site . . . 20

2.2.2 Instrumentation . . . 20

2.2.3 Data Analysis . . . 21

(11)

2.2.4 Computation of turbulence fluxes . . . 27

2.3 Tidal Structure . . . 29

2.4 Tidal Asymmetry . . . 38

2.5 Conclusion and Implications . . . 42

3 Tidal Straining and Phytoplankton Growth: Observations and Param- eterization of Turbulence 44 3.1 Introduction . . . 44

3.2 Field measurements . . . 47

3.2.1 Study site . . . 47

3.2.2 Data collection and analysis . . . 47

3.3 Observations and Result . . . 50

3.3.1 Conditions during spring tide . . . 51

3.3.2 Nutrient flux during spring tide . . . 51

3.3.3 Conditions during neap tide . . . 56

3.3.4 Nutrient flux during neap tides . . . 57

3.4 Discussion . . . 60

3.4.1 Vertical shear and BBL . . . 61

3.4.2 Vertical shear and stratification . . . 62

3.4.3 Vertical shear and frictional velocity . . . 65

3.4.4 Eddy diffusivity and nitrate flux . . . 65

3.5 Concluding remarks . . . 66

4 Remote Sensing of Chlorophyll-a over shallow continental shelves: Al- gorithm formulation and Global Applicability 68 4.1 Introduction . . . 68

4.2 Dataset and Methods . . . 71

4.2.1 In-situ Data . . . 71

4.2.2 Water sampling analysis . . . 73

(12)

4.2.3 Radiometric data set . . . 74

4.2.4 Satellite data, processing and validation . . . 75

4.2.5 Statistical evaluation . . . 76

4.3 Optical Complexity of eastern Arabian Sea . . . 78

4.4 Theoretical formulation . . . 81

4.4.1 Identification of λ2 . . . 83

4.4.2 Delineation of λ3 . . . 84

4.4.3 Variation in coefficients . . . 86

4.4.4 Turbid water indices . . . 89

4.5 Results . . . 92

4.5.1 Validation of algorithms with in-situ data set . . . 92

4.5.2 Validation of GUC2 with MERIS derived CHLvalues . . . 95

4.5.3 An application of validation of GUC2 in Chesapeake Bay . . . 96

4.6 Discussion . . . 99

4.6.1 Performance of GUC2 algorithm . . . 99

4.6.2 Non-applicability of GUC2 in open ocean waters/case I waters . . . 102

4.7 Conclusion . . . 103

5 Optical Partitioning of the World Ocean: Trends of Chlorophyll-a, Pri- mary Production & Anthropogenic Attributed Perturbations 105 5.1 Introduction . . . 105

5.2 Methodology . . . 107

5.2.1 Identification of optical domains. . . 108

5.2.2 Optical algorithms. . . 108

5.2.3 Computations of Oceanic Primary Production. . . 109

5.2.4 Global Warming Index. . . 109

5.2.5 Chesapeake Bay Dataset. . . 109

5.2.6 Statistical Analysis. . . 109

5.3 How to distinguish case 2 waters? . . . 110

(13)

5.4 Phytoplankton in global case 2 waters . . . 112

5.5 Carbon fixation by phytoplankton . . . 117

5.6 Climate Driven Variability . . . 119

5.7 CO2 residence in case 2 waters . . . 122

6 Non-Linear Interaction of the light field with total suspended matter: Regional Algorithm Formulation and Validation 127 6.1 Introduction . . . 127

6.2 Dataset and Methods . . . 131

6.2.1 In-situ dataset . . . 131

6.2.2 Water sampling analysis . . . 131

6.2.3 Radiometric Dataset . . . 131

6.2.4 Statistical Evaluation . . . 132

6.3 Optical complexity of the Zuary estuary . . . 132

6.3.1 Variability of IOPs . . . 133

6.3.2 Identification of water types . . . 134

6.3.3 Optical characterization and Rrs(λ) . . . 137

6.4 Theoretical Formulation of an algorithm . . . 137

6.4.1 Identification of λ2 . . . 139

6.4.2 Incorporating of backscattering coefficient . . . 140

6.5 Results . . . 143

6.5.1 Validation of algorithm with in-situ data set . . . 145

6.6 Discussion . . . 146

6.6.1 How backscattering accounts for non-linearity? . . . 146

6.6.2 Relation between particulate backscattering and Rrs . . . 147

6.7 Conclusion . . . 150

7 Conclusions 151

Bibliography 153

(14)

Variables

β Volume Backscattering Function

∂t partial derivative with respect to time u depth-averaged alongchannel velocity field bfbp particulate backscattering ratio

aCHL specific absorption coefficient of chlorophyll-a

aT SM specific absorption coefficient of total suspended matter an amplitude coefficient

aW absorption by water molecules

aCDOM absorption by colored organic dissolved matter aCHL absorption by chlorophyll-a

aT SM absorption by total suspended matter B vertical turbulent buoyancy flux

(15)

b buoyancy

bbCHL backscattering coefficient of chlorophyll-a

bbT SM backscattering coefficient of total suspended matter bb backscattering coefficient

bw backscattering coefficient of water molecules cCHL chlorophyll-a concentration

cM CHL modeled chlorophyll-a concentration cT SM concentration of total suspended matter Ed downwelling irradiance

Es surface reaching irradiance Eu upwelling irradiance

f Coriolis parameter

g acceleration due to gravity H depth of water column

hbbl bottom boundary layer height k total number of tidal constituents

kd diffuse attenuation coefficient of downwelling irradiance Kt diffuse attenuation coefficient of upwelling irradiance Lu upwelling radiance

N buoyancy frequency

(16)

n individual tidal constituent

N i nitrate

O oxygen

P production rate of turbulence kinetic energy

R reflectance

r thickness of water parcel Rrs remote sensing reflectance Rig gradient Richardson number

S salinity

u alongchannel velocity field

u(ζ) quadratic alongchannel tidal velocity profile u bed-frictional velocity

uz vertical profile of alongchannel velocity field ulp low-pass alongchannel velocity

utidal high-pass alongchannel velocity V volume of water parcel

v crosschannel velocity field w vertical velocity field

x, y, z components of cartesian coordinate system zc spectral optical attenuation depth

(17)

zt depth of sub-surface chlorophyll-a maxima aλ=70090 90th percentile of total absorption at 700 nm aλ=440avg averaged total absorption at 440 nm

aλ=700avg averaged total absorption at 700 nm aλ=440tot total absorption at 440 nm

aλ=700tot total absorption at 700 nm

bλ=700b90 90th percentile of backscattering at 700 nm bλ=700bavg averaged backscattering at 700 nm

bλ=700btot total backscattering at 700 nm btot total scattering

cM T SM algorithm estimated total suspended matter concentration Greek Symbols

α Fresnel reflection albedo from sun and sky dissipation rates of turbulence kinetic energy η surface elevation

η0 predicted tidal signal

ηw Fresnel refractive index of seawater Γ mixing efficiency

γoU velocity skewness γoζt duration asymmetry

(18)

κ von K´arm´an constant κρ eddy diffusivity

λ wavelength

ν kinematic viscosity

νn vertical turbulent nitrate flux

ωn frequency

φ potential energy anomaly

φa spectral power absorbed by water parcel φi spectral power

φn phase

φs spectral power scattered by water parcel

φt spectral power transmitted unchanged through water parcel φu, φv phase lag of horizontal components

ψ scatterance angle ψ(k) spectrum of shear

ρ seawater density

ρ0 reference seawater density

ρf Fresnel reflection index of seawater ρz seawater density at depth z

τb bed-shear stress

(19)

Acronyms

ADCP acoustic Doppler current profiler AOP apparent optical properties BBL bottom boundary layer

CDOM colored organic dissolved matter CHL chlorophyll-a

CO2 carbon dioxide

CTD conductivity-temperature-depth CZCS Coastal Zone Colour Scanner DIN Dissolved Inorganic Nitrogen ENSO El-Ni˜no-Southern Oscillation

GLODAP Global Ocean Data Analysis Project GUC2 Goa University Case II

HHW higher-high-water

IOP inherent optical properties LLW lower-low-water

MERIS MEdium Resolution Imaging Spectrometer MLD mixed layer depth

MODIS Moderate Resolution Imaging Spectroradiometer NDCI Normalized Difference Chlorophyll Index

(20)

NIR near infrared

NOMAD NASA bio-Optical Marine Algorithm Data set NPP Net Primary Production

OACs optically active constituents OBPG Ocean Biology Processing Group

OD optical depth

PAR photosynthetically active radiation PDO Pacific Decadal Oscillation

Rc radius of curvature RMSE root-mean-squared-error S2 squared-shear

SCI Synthetic Chlorophyll Index SDE standard error

SIPS strain induced periodic stratification SPM suspended particulate matter

SST Sea Surface Temperature SWIR Short-Wave-Infrared-Region TSM total suspended matter

VGPM Vertically Generalized Production Model VIIRS Visible Infrared Imaging Radiometer Suite

(21)

VMP vertical microstructure profiler

(22)

1.1 Schematic of tidal straining. (a) velocity profiles correspond to alongchan- nel velocity, where, solid line is effective flood and ebb profiles, dashed line show a symmetric tidal current velocity profile, and dotted line is a resid- ual mean velocity profile, (b) sketch of alongchannel velocity with overlaid horizontal density gradient in regions of freshwater influence, (c) illustra- tion of alongchannel cross-section of an estuary, and spiral arrows represent bottom generated mixing over a sloping continental shelf, and (d) repre- sents downslope ebb currents. Figure is inspired from the seminal works of Simpson et al. (1990) and Jay and Musiak (1994). . . 3 1.2 Schematic of the phytoplankton production that is the cross product of

biomass and growth rate regulated by mortality, light, temperature, and nutrients. . . 8 1.3 Defining inherent optical properties (adapted from Mobley (1994)). . . 11 1.4 Spectral absorption of optically active constituents in optically complex

waters. . . 12

(23)

2.1 Figure shows the study site (filled triangle) along with Zuary and Man- dovi estuary complex. White vertical line shows the location of mooring deployment. . . 19 2.2 Figure shows the spatially smoothed and temporally averaged salinity as

well as bathymetry along the axis of Zuary estuary for the month of January from year 2008 to 2017. . . 21 2.3 Figure shows the rotation of tidal ellipse. Close empty circle represent

semi-major axis where maximum velocity is experienced, semi-minor axis has minimum velocity. . . 24 2.4 Harmonic analysis of the observed water current time-series. Predicted

time-series is presented with an offset of 0.75 m. . . 30 2.5 A 40 hour low-pass Lanczos filter was used to separate the low frequency

and high frequency signals. Tidal signal (high pass signal) is strongly corre- lated to the horizontal velocity components, and circulation is least affected by wind. . . 32 2.6 M2 tidal ellipse at different depths of the water column, (a) depth 2–6 m,

(b)7–11 m, and (c) 12–16 m. . . 34 2.7 S2 tidal ellipse at different depths of the water column, (a) depth 2–6 m,

(b)7–11 m, and (c) 12–16 m. . . 34 2.8 K1 tidal ellipse at different depths of the water column, (a) depth 2–6 m,

(b)7–11 m, and (c) 12–16 m. . . 35 2.9 O1 tidal ellipse at different depths of the water column, (a) depth 2–6 m,

(b)7–11 m, and (c) 12–16 m. . . 35 2.10 Vertical structure of tidal ellipses representing major tidal constituents. . . 36 2.11 Mean currents observed during fully-developed flood and ebb of spring and

neap tides. . . 37 2.12 Observations of the computed BBL. . . 39

(24)

2.13 Mean currents observed during fully-developed flood and ebb of spring and neap tides. . . 40 3.1 Map showing the study region and study site (black filled circle) on the

western coast of India. . . 47 3.2 Observations during January 13–22, 2017. (a) Stick plot representing

wind velocity at study site, (b) water elevation (black) and depth-averaged alongchannel velocity component (gray) from bottom moored ADCP. The shaded area (light gray) exhibits the period when continuous microstruc- ture measurements from VMP-250 were collected during spring and neap tides, (c) depth-averaged salinity (black) and temperature (gray) from VMP-250 deployments. . . 50 3.3 Depth-averaged alongchannel velocity is shown at the top (<uz>). Obser-

vations and computations during spring and neap tides at the study site, (a) potential density anomaly (φz) and overlaid depth-averaged buoyancy frequency (red filled circles; log10(N2)), (b) bar plot of chlorophyll-a con- centrations (total of surface and zt [µg L−1]); red and blue filled circles represent measured depth-averaged nitrate and oxygen concentrations, re- spectively, (c) average of SPM concentrations observed at≈1 m above sea- bed and at surface, (d) depth-averaged alongchannel straining (black) and advection (red) terms, (e) depth-averaged rates of dissipation of turbulence kinetic energy () using VMP-250. . . 53 3.4 The figure shows vertical profiles representing flood (blue), early-ebb (green),

and ebb (red) stages during spring tide, (a) log10buoyancy frequency (s−2), (b) vertical turbulent salt flux (kg m−2 s−1), (c) vertical turbulent nitrate flux (mg m−2 s−1), (d) (W kg−1), (e) straining term (kg m−1 s−1), (f) advection term (kg m−1 s−1). . . 54

(25)

3.5 The figure shows vertical profiles representing flood (blue), early-ebb (green), and ebb (red) stages during neap tide, (a) log10 buoyancy frequency (s−2), (b) vertical turbulent salt flux (kg m−2 s−1), (c) vertical turbulent nitrate flux (mg m−2 s−1), (d) (W kg−1), (e) straining term (kg m−1 s−1), (f) advection term (kg m−1 s−1). . . 57 3.6 Representation of alongchannel shear absolute values (s−1) (color plot) with

BBL position overlaid as white filled circles. . . 60 3.7 Profiles of alongchannel velocity at 15 minute intervals showing water cur-

rent structure evolution during flood and ebb stages of a tidal cycle. . . 62 3.8 Representation of logarithm of computed Richardson number normalized

by the critical 0.25 (color plot), with BBL position overlaid as white filled circles. . . 63 3.9 Depth averaged alongchannel velocity is shown for referencing different

stages of a tidal cycle. (a) represents depth-averaged signed frictional ve- locity (m2 s−2), (b) depth averaged alongchannel shear (s−1). . . 63 3.10 The figure shows, (a) vertical structure of vertical turbulent buoyancy flux

buoyancy gradient (color plot), with overlaid BBL position as white filled circles, (b) depth averaged logarithm of ratio of production to dissipation rates of turbulence kinetic energy. . . 64 3.11 The figure shows logarithmic vertical profiles of (a) eddy diffusivity, and (b)

vertical turbulent nitrate flux, during flood (black) and ebb (gray) stages of a tidal cycle. . . 64

(26)

4.1 Geolocation of the study area. (a) Sampling stations (n = 51) along the coastal waters of eastern Arabian Sea (filled circles; n = 43) and locations of sampling station withTrichodesmium bloom sightings are shown as un- filled triangles (n = 8), (b) location of the in-situ data-points SATCORE

”match-up” with MERIS sensor used for the validation of algorithm in the eastern Arabian Sea (n = 14). The estuarine data used at different steps of algorithm formulation was collected from Zuary and Mandovi estuaries shown with arrows (n = 349), (c) location of the in-situ NOMAD (n = 82) data set ”match-up” with MERIS sensor along California coast (n = 37) and northeastern Gulf of Mexico (n = 45). . . 72 4.2 The plot showing the variability of magnitude and shape of Rrs(λ) in the

visible part of electromagnetic spectrum of three water types encountered in the study area, (a)Trichodesmiumbloom dominated (n = 8), (b) inshore turbid (n = 40), and (c) offshore waters (n = 103). . . 77 4.3 Absorption coefficient of CHL (green) showing primary and secondary

peaks at 443 nm and 663 nm. The gray (shaded region), yellow, and red lines represents the mean spectral absorption of CHL(one-standard- deviation), CDOM, TSM, while blue line represents water molecules. The figure shows a very low relative absorption byTSM andCDOM at 663 nm with respect toCHL, with water molecules being dominant of all. . . 80 4.4 The figure shows (a) iteration of λ2 with respect to λ1 and λ3 to obtain

minimum RMSE when λ2 = 623 nm, (b) iteration ofλ3 with respect toλ1 and λ2 shows minimum RMSE when λ3 = 636 nm, and (c) iteration of λ1, between 650 and 675 nm, to obtain minimum RMSE whenλ1 = 663 nm. . 86 4.5 The linear relationship of (aW+aCDOM+aT SM) at 663 nm and 623 nm, a)

a correlation using 41 data points used for algorithm formulation withr2 = 0.9212, and, b) a correlation using 349 data points which include estuarine data too (see Figure 4.1b) with r2 = 0.99. . . 87

(27)

4.6 Figure showing the GUC2 algorithm formulation by regressing 41 in-situ data points (cCHL) with GUC2 CHLindex results in r2 = 0.8584. . . 88 4.7 The figure shows the results of the regression of turbid water indices with

in-situCHLconcentration. Same data (n = 41) are used to formulate GUC2. 91 4.8 Figure showing the GUC2 algorithm validation, (a) a linear fit was obtained

between 110 spectrophotometrically derivedCHL(cCHL) ranging from 0.11 to 25.56 µg L−1 and estimated concentration from GUC2 (cM CHL) that resulted in r2 = 0.9964, RMSE = 0.4449, bias = 0.0001, and p<0.0001, (b) cCHL limited up to 3 µgL−1 and a total of 83 data points still shows a strong correlation indicating the algorithm to hold when tested with varying concentrations, i.e., r2 = 0.8885, RMSE = 0.3037, bias = -0.0069, and p<0.0001. . . 94 4.9 The validation of MERIS derivedCHLusing GUC2 and in-situ data set i.e.

NOMAD (n = 82; open squares) and SATCORE (n=14; open circles). A total of 96 data points were used for validation. A correlation is obtained with r2 = 0.9046, RMSE = 1.2499 and bias = -0.0013. . . 95 4.10 Figure showing the comparison of CHL in Chesapeake Bay retrieved by

applying GUC2, representing data points from September 2011 to January 2012, (a) sampling locations HUC8 (filled light gray triangles), HUC12 (darker gray filled circles), and CBSeg2003 (open squares), (b) shows the daily discharge rates (m3 s−1), (c) a linear correlation between in-situCHL and GUC2 (MERIS) derived CHLwith r2 = 0.9872, RMSE = 1.4116, and bias = -0.0001 for 455 data points. . . 97 5.1 Sampling sites from where the data set of dissolved inorganic nitrogen were

retrieved from Chesapeake Bay. . . 110 5.2 Figure shows MERIS entire mission averaged re-processed CHL (µg L−1)

computed using a combination of OC4E and GUC2. . . 111

(28)

5.3 Figure shows, (a) a ratio of defaultCHL(OC4E evaluated) to re-processed data set (GUC2 and OC4E evaluated) showing the extent of case 2 waters in World Ocean and overestimations occurred using default data set, (b) decadal trend of global annually-averaged euphotic zone depth-integrated CHL biomass (Tg, 1 Tg = 1012g), (c) decadal trend of annually averaged surface area of identified case 2 waters (estuarine and coastal) globally. . . 112 5.4 MERIS annually averaged CHL concentration (µg L−1) computed using a

combination of OC4E and GUC2 (re-processed data set), (a) global, (b) case 1 waters (open ocean), (c) case 2 waters (estuarine and coastal), and, (d) global decadal trend of CHL in form of correlation coefficient, r, at statistical significance, p<0.01. . . 114 5.5 Figure shows the computation of global oceanic primary production using

the VGPM model. . . 116 5.6 Figure shows CHL computed decadal trends of NPP, (a) global, (b) case

1 waters (open ocean), (c) case 2 waters, and (d) global decadal trend of NPP in form of correlation coefficient,r, at statistical significance, p<0.01. 119 5.7 A comparison ofCHLanomaly and Global Warming Index is presented as a

change in global temperatures caused due to, (a) Anthropogenic activities, (b) Naturally occurred, and (c) observed change in temperature through measurements. . . 120 5.8 A comparison of CHLanomaly and indicators of climate change (annually

averaged), (a) land cover in million hectares, (b) land use percentages, and (c) amount of fertilizers used in hundred-million-tonnes globally, where excess run-off in rivers finally reaches in estuaries and cycles in to coastal waters. . . 122

(29)

5.9 A comparison of CHL anomaly and DIN anomaly computed for selected sampling sites (shown in figure 5.1) in Chesapeake Bay. The figure shows that an excess increase in DIN input to upper stretches to Chesapeake Bay resulted in CHLgrowth. . . 123 5.10 Decadal trend ofpCOsw2 in identified estuarine and coastal waters globally

showing a decline of these waters as a source of CO2 to atmosphere. . . 124 5.11 Figure shows the identified case 2 waters of the World Ocean with labeled

major rivers and coastal areas. . . 126 5.12 Figure shows division of World Ocean in 12 major basins. . . 126 6.1 Figure shows the study area with sampling locations (Z5–Z21) in an opti-

cally complex Zuary Estuary, Goa. . . 131 6.2 Figure shows the frequency distribution (n = 166) of, (a) TSM, (b) CHL,

and, (c) CDOM observed at all sampling sites. . . 133 6.3 Figure shows the frequency distribution (n = 166) of inherent optical prop-

erties, (a) total backscattering coefficient at 700 nm (bbλ=700tot ), (b) total absorption coefficient at 700 nm (aλ=700tot ), and, (c) total absorption coeffi- cient at 700 nm (aλ=440tot ). . . 134 6.4 Figure shows the separation of data set in two optical water types depend-

ing on the linear correlation between total backscattering coefficient and total absorption coefficient at 700 nm. The presence of two optical water types, with optical characteristics included. . . 135 6.5 Variability of spectral Remote sensing reflectance (Rrs(λ)) is presented for,

(a) Group 1, and (b) Group 2. . . 136 6.6 Figure showing the partial derivative analysis of remote sensing reflectance

with respect to specific TSM absorption coefficient as, (a) First derivative (∂Rrs/∂aT SM), and (b) Second derivative (∂2Rrs/∂(aT SM)2). . . 138 6.7 A sensitivity analysis showing the simulated remote sensing reflectance for

the TSM concentration 1–1000 mg L−1. . . 140

(30)

6.8 Figure showing the linear correlation fit between the absorption of CHL, CDOM, and water molecules at λ1 = 679 nm, and λ2 = 695 nm. . . 141 6.9 Figure showing the best fit between the absorption of CHL, CDOM, and

water molecules at λ1 = 679 nm, and λ2 = 695 nm. . . 143 6.10 Figure showing the linear correlation fit between algorithm estimatedTSM

(CMT SM) and in-situ TSM (cT SM) concentrations. A strong correlation for n = 116 was observed with r2 = 0.88, RMSE = 3.0605, and mean bias

= -0.3309. . . 144 6.11 Figure showing the linear correlation fit between algorithm estimatedTSM

(CMT SM) and in-situ TSM (cT SM) concentrations in range 20–40 mg L−1. A strong correlation for n = 103 was observed with r2 = 0.81, RMSE = 2.116, and mean bias = 0.2249. . . 145 6.12 Figure showing the distribution of TSM concentrations with respect to

remote sensing reflectance (Rrs) and bbp:bp. . . 147 6.13 Figure showing the distribution ofTSM concentrations in range 20–40 mg

L−1 with respect to remote sensing reflectance (Rrs) and bbp:bp. . . 148 6.14 Figure showing the variability of spectral Rrs with respect to TSM con-

centrations. . . 149 6.15 Figure showing the linear correlation variability of spectral Rrswith respect

to particulate backscattering coefficients. . . 149

(31)

2.1 Tidal ellipse values derived from the harmonic analysis of the tidal currents showing the major tidal constituents. . . 31 2.2 Percentage variance of high-pass and low-pass signals from 40 hour Lanczos

filter is shown. It is evident that dominant fluctuations occur at higher frequencies, while an input from lower frequencies is intermittent. Shear represents the difference between the top-most bin and the lower-most bin. 33 3.1 The change in depth-averaged concentrations of oxygen, chlorophyll-a and

nitrate are shown in form of percentages from flood to ebb during spring tides. O, CHL, and Ni represents oxygen, chlorophyll-a, and nitrate con- centrations, respectively. Although all values are shown, but calculated percentages reflect flood to ebb change. . . 52 4.1 The performance of GUC2, turbid water approaches, and OC4E is pre-

sented when validated with in-situCHLconcentrations (n = 110). In table, λ1 andλ2 are 663 nm and 623 nm, respectively, whileλ4 throughλ9 repre- sent 560 nm, 620 nm, 665 nm, 681 nm, 709 nm, and 753 nm, respectively.

SDE and p represent standard error and p-value. . . 93

(32)

4.2 A comparison of linear correlation is presented for NOMAD and SATCORE repository individual data-sets. . . 96 4.3 A comparison of GUC2 derived CHL and in-situ measured CHL from

September 2011 to January 2012 in Chesapeake Bay. The sampling sites HUC8, HUC12 and CBSeg2003 represent Hydrologic Unit, small water- shed and Monitoring Segment, respectively. In table, n, r2, cmin, andcmax represents encountered match-up points, correlation coefficient, minimum CHLobserved, and maximum CHLobserved, respectively. . . 98 5.1 Annual-averaged decadal changes in chlorophyll-a concentrations globally,

in case 1, and in case 2 waters, where r is correlation coefficient tested at statistical significance, p<0.01. In parentheses, corresponding trends of sea-surface temperature in form ofr are presented. . . 115 5.2 Annual-averaged decadal changes in oceanic net primary production glob-

ally, in case 1, and in case 2 waters, wherer is correlation coefficient tested at statistical significance, p<0.01. In parentheses, corresponding trends of sea-surface temperature in form ofr are presented. The area of each basin is an approximate in million km2. . . 118

(33)

INTRODUCTION

In a physical perspective, estuaries are realms where vertical stress divergence vector experiences a shift in direction and magnitude depending on the intensity of baroclinic and barotropic flows. In weakly-mixed and well-mixed cases of stratification in an estuary, biogeochemical processes are heavily influenced by tides (external forcing agents) governed by estuarine morphology (Dronkers, 1986), tidal range (Allen et al., 1980), surface wind stress (Alvarez-Salgadoet al., 1996), alongchannel and lateral circulation (Becherer et al., 2016), and several other processes over a shallow continental shelf. Estuarine circulation over shallow horizontally stratified continental shelves is a well-understood process, where the near-bottom residual currents are directed landward and surface directed to seaward.

This characteristic flow is a density dependent and was originally thought to be a part of the gravitational circulation caused due to horizontal density gradient (Pritchard, 1954, 1955; Hansen and Rattray, 1965). However, subsequent studies revealed that a number of additional physical processes impact the estuarine exchanges (see, Burchard and Baumert (1998) for details). In sight of the investigated processes in this thesis, estuarine exchange flows associated with internal tidal asymmetry (Jay and Musiak, 1994) generate flood-ebb asymmetry in the velocity structure of the alongchannel currents that results in residual

(34)

flow oriented in direction of the classical gravitational circulation (Lerczak and Geyer, 2004). This is often an important physical mechanism in well-mixed tidally energetic regions, over gravitational circulation (Burchard and Hetland, 2010; Bechereret al., 2011;

Burchard et al., 2018).

1.1 Tidal Straining

Mathematically, covariance between eddy viscosity and vertical shear stress of the along- channel velocity field represents tidal straining, i.e., straining induced by the vertical shear of periodic tidal currents overlaid by horizontal density gradients. Through field obser- vations and modeling approach, van Aken (1986) and Simpson et al. (1990) showed the formation of asymmetry in stratification evolution varying at flood-ebb phases of a tidal cycle. The bottom boundary friction influences the tidal currents to shear near-bed, that results in differential advection of the horizontal buoyancy gradient. In Figure 1.1, the flood and ebb current profiles are shown. It should be noted here that flood is defined as the part of tidal cycle where the flow is directed towards higher buoyancy, and ebb is vice-versa. During floods, the offshore saline denser water is transported over shallower continental shelf faster near surface causing unstable stratification, while ebb currents ad- vect fresher less dense waters from up-estuary yielding opposite flow profile and enhancing stratification vertically. Therefore, the term strain induced periodic stratification (SIPS) was coined by Simpson et al. (1990).

Jay and Musiak (1994) first acknowledged the importance of tidal asymmetries in shaping eddy viscosity and longitudinal vertical shear that directly impacts estuarine circulation in landward-based near-bed tidal velocities during both, ebb and flood, phases. The dif- ference in the vertical momentum flux as observed in Figure 1.1 during ebb and flood currents results in a difference in the transport of along-channel momentum flux. The enhancement of the turbulence momentum flux during flood yields a vertically homo- geneous profile, which in association of the vertically more sheared ebb flows causes a residual longitudinal exchange flow. Therefore, the resulting profile of circulation has the

(35)

Figure 1.1 Schematic of tidal straining. (a) velocity profiles correspond to

alongchannel velocity, where, solid line is effective flood and ebb profiles, dashed line show a symmetric tidal current velocity profile, and dotted line is a residual mean velocity profile, (b) sketch of alongchannel velocity with overlaid horizontal density gradient in regions of freshwater influence, (c) illustration of alongchannel cross-section of an estuary, and spiral arrows represent bottom generated mixing over a sloping continental shelf, and (d) represents downslope ebb currents. Figure is inspired from the seminal works of Simpson et al. (1990) and Jay and Musiak (1994).

same orientation as classical gravitational circulation, of which tidal straining is an in- ternal tidal asymmetry resulting in periodic vertical stratification and inhibiting vertical momentum flux (Figure 1.1).

SIPS or tidal straining contributes to as much as two-third of the total estuarine circu- lation (Burchard and Hetland, 2010) in partially mixed estuaries. The implications of SIPS inherently includes the horizontal dispersion of particulate and dissolved matter of organic and inorganic origin such as nutrients, salt, pollutants, sediments, and biological

(36)

on variety of parameters that contribute to estuarine and coastal ecosystem health. Jay and Musiak (1994) showed the flood-ebb varying asymmetric mixing that yields in an estuarine exchange flow resulting in a difference in average vertical distribution of the along-channel momentum flux. For example, enhanced turbulent momentum flux during flood is relatively more homogeneous vertically. Suggestively, based on the asymmetric flux of vertical momentum, tidal straining is characterized as an internal mixing asym- metry which is associated with vertical stratification and its inhibiting role of vertical momentum flux is applicable in regions of freshwater influence.

In well-mixed estuaries (as a case in this thesis), tidal currents are vertically sheared to produce turbulence due to bottom friction (Figure 1.1). The major semi-diurnal (M2;S2) and diurnal (K1;O1) tidal constituents interact with boundaries longitudinally and lat- erally to generate higher harmonics, called as overtides (e.g., M4 from M2). Therefore, the ratio of amplitude of major to higher harmonics including the phase shift at differ- ent stages of the tidal wave can cause asymmetry in the strength of flood and ebb flow (Dronkers, 1986). This differential asymmetry in the flood-ebb currents causes differential production of turbulence, that ultimately affects the residual exchange flow. In addition, asymmetric induced transformed vertical stratification limits the turbulence production and dissipation rates by controlling turbulent mixing length scales that inhibits strati- fication (Smyth and Moum, 2000). In simplest terms, periodic systematic stratification results in asymmetric vertical profile of the alongchannel tidal currents. Through observa- tions (e.g., Stacey et al., 2001) and numerical modeling (e.g., Burchard and Schuttelaars, 2012), several studies has reported that the effects of tidal straining might be of greater influence than it is realized and may even be more crucial than classical gravitational circulation in some cases in terms of generating estuarine exchange flow.

It is crucial at this point for better understanding that the basic equation governing almost any geophysical flows and with respect to SIPS over shallow continental shelves should be addressed. Momentum equations that solve for flux in form of Reynolds averaged Navier

(37)

Stokes equations are solved with an application of Boussinesq and hydrostatic assump- tions. Therefore, the horizontal components in a three-dimensional curved coordinate system (Huijts et al., 2009) are written as,

∂u

∂t =−u ∂u

∂x − v ∂u

∂y −w ∂u

∂z +f v− 1

Rcuv−g∂η

∂x + Z η

z

∂b

∂xdz0− ∂hu0w0i

∂z , (1.1)

∂v

∂t =−u ∂v

∂x − v ∂v

∂y − w ∂v

∂z −f u− 1

Rcuu−g∂η

∂y + Z η

z

∂b

∂ydz0− ∂hv0w0i

∂z , (1.2) where, ∂t, ∂x, ∂y, and ∂z represent the partial derivative with respect to time and space,Rc is radius of curvature,f is Coriolis parameter,η is surface elevation at a reference time in space, andbdenotes buoyancy. The velocity fields,uandvare the horizontal components, and wis the vertical component such that when the curved coordinated system is fit over channeled flows, x is perpendicular to Rc and is called the alongchannel component.

Similarly, y parallel to Rc is the cross-channel component and z is vertically positive upwards in any orientation of time and space. The angled brackets, h·i, represents the Reynolds averaged, and primes denote fluctuating components that deviate from Reynolds averaged velocity fields. For the flows over shallow continental shelves, it is logical to imply assumption of incompressible fluid case, where the three velocity components are represented in form of continuity equation, written as,

∂u

∂x +∂v

∂y + ∂w

∂z = 0 (1.3)

The hydrodynamics dealt with in this thesis show the impact of straining of density on optically active constituents (OACs) such as chlorophyll-a (CHL), colored organic dissolved matter (CDOM), and total suspended matter (TSM). Therefore, the basic parameter relating to density changes is buoyancy (as in equations 1.1 and 1.2) defined as,

b=−gρ−ρ0

ρ (1.4)

(38)

where, g is acceleration due to gravity, ρ is density, and ρ0 is a reference density. The transport that occurs due to changes in buoyancy is represented as follows,

∂b

∂t =−u∂b

∂x −v∂b

∂y −w∂b

∂z −u∂hb0w0i

∂z (1.5)

where,b0 represents the fluctuating part in buoyancy from Reynolds average. Further, the stratification in terms of buoyancy is buoyancy frequency (N2) that denotes the frequency at which water parcel (arbitrary volume of water) moves vertically when perturbed from the state of rest, and is represented as,

N2 = ∂b

∂z =g(−1 ρ

∂ρ

∂z) [radians s−1]2 (1.6)

Equations 1.1–1.6 form the basis of interaction between barotropic and baroclinic flows over shallow continental shelves that results in periodic vertical mixing and stratification depending on the part of tidal cycle. Therefore, equations 1.1–1.6 in direct perspective with tidal straining is understood in terms of potential density anomaly following the seminal works of van Aken (1986) and Simpson et al. (1990). Potential density anomaly (φ) represents a balance between flood induced mixing and ebb induced stratification.

φ increases positively with water column stability, represents work done to mix water column, and signifies stratification in a water column. φ is represented as,

φ= 1 H

Z 0

−H

(ρ−ρz)gz dz, and ρ= 1 H

Z 0

−H

ρzdz (1.7)

where, H is the depth of water column. Further, the interaction of barotropic flows overlaid with horizontal density gradient and constantly evolving stratification can be quantified in terms of φ, as,

∂φ

∂t = g H

∂ρ

∂x Z 0

−H

(uz −u)z dz (1.8)

(39)

Equation 1.8 shows the change in density or explains constant evolution of stratification only in the alongchannel direction due to advection. ∂x∂ρ, uz, and u represent the hori- zontal density gradient, vertical profile of alongchannel velocity field, and depth-averaged alongchannel velocity, respectively. As a result, the changes in stratification enacted upon by tidally sheared flows due to bottom friction, and ∂ρ∂x is explained by representing tidal velocity profile using the quadratic formula proposed by Bowden and Fairbairn (1952), given as,

u(ζ) =u(1.15−0.425ζ2), ζ =z(H) (1.9) which implies that the rate of change of φ is given as,

∂φ

∂t

str

=gh u ∂ρ

∂x Z 0

−1

(0.15−0.425ζ2)ζ dζ = 0.031gh u ∂ρ

∂x (1.10)

1.2 Physical Controls on Phytoplankton

The frequency of occurrence and intensity of phytoplankton blooms, and harmful algal blooms in global coastal waters have significantly increased over the past few decades (Landa et al., 2016). The accumulation of phytoplankton species or rapid growth under optimal conditions such as optimal light, nutrient availability, and temperature enhances biological productivity and plays a vital role in the regulation of atmospheric carbon to deeper waters (Falkowski and Oliver, 2007). Furthermore, harmful algal blooms cause hypoxia or anoxic conditions during the senescence phase, and are responsible for spread- ing toxicity in the food-chain, thereby, affecting the ecosystem structure and function (Anderson et al., 2012).

Monitoring programs globally are dedicated completely to algal research have provided detailed data sets that helped in understanding of the ecological niche these species thrive- in, but it also requires an understanding of the mechanism that drives their productivity.

The matter in suspension in coastal waters with strong tidal flows is always under the direct influence of physical processes. In general, two factors governs the patchiness of

(40)

Figure 1.2 Schematic of the phytoplankton production that is the cross product of biomass and growth rate regulated by mortality, light, temperature, and nutrients.

phytoplankton (Lucaset al., 1999), i.e., local balance between phytoplankton growth and disintegration, and spatio-temporal transports of water-borne parameters and plankton.

Local growth rates of phytoplankton may drastically vary longitudinally depending on water depth, fluvial discharge, nutrient concentrations, grazing, turbidity, and time-scales of vertical transport. Therefore, the accumulation of biomass depends on positive phyto- plankton growth and pathways (mechanism) that transport phytoplankton populations as well as nutrients that boosts growth (Figure 1.2).

The dependence of phytoplankton growth on various physical processes is illustrated in Figure 1.2. In ecosystems predominant with tidal currents overlaid freshwater influence, stratification (tidal straining in this case) leading to shallower depths of pycnocline than euphotic depth provides optimal irradiance conditions for photosynthetic organisms near the surface. However, the same pycnocline restricts the availability of inorganic nutrients that makes deep waters rich. In regimes of weak horizontal density gradients, bottom generated turbulence re-suspended sea-bed suspended matter and redistribute inorganic nutrients from waters near-bed to near-surface euphotic layer and could transport phyto- plankton from surface to poorly lit-zones (Pingree et al., 1978). In addition, horizontal

(41)

advection of nutrients to surface layers also acts as input that allows maximum chlorophyll concentrations near-surface on the landward side of stratification (Landeira et al., 2014) and decreases offshore (Franks, 1992). Moreover, horizontal and vertical mixing rates redistribute suspended matter including phytoplankton and their grazers, and finally in- fluences community structure and functional dynamics (Lewitus et al., 1998; Domingues et al., 2010). These environmental variations in presence of strong horizontal density gra- dients directly affects upper mixed layer, euphotic depths, nutrient supply, enhance phy- toplankton growth, and allows development of constantly varying different phytoplankton assemblages (Margalef, 1978).

Much notions at the juncture of nutrient dynamics and external forcing agents were exem- plified by sediment dispersal dynamics (e.g., Glangeaud, 1938; Postma and Kalle, 1955), that later became the focus for several seminal observation (e.g., Jay and Musiak, 1994;

Scully and Friedrichs, 2007) and numerical studies (e.g., Burchard and Baumert, 1998).

Similarly, phytoplankton were studied from the biological perspective by including nu- trient dispersal and diffusion with an aim to improve the understanding of fluctuations (biological-mediated turbulence under constant external conditions) in intrinsic popula- tion in phytoplankton communities (e.g., Beninc´aet al., 2008). Therefore, it is crucial to understand the complex interplay between intrinsic nutrient dynamics and nature-driven variations.

Hence, characterization of horizontal displacement of water masses at different tidal pe- riodicity with respect to variability (intrinsic as well as extrinsic) that allow differential nutrient flux to favor growth is critical to further understanding of horizontal transport mechanisms, a key to ascertain estuarine and coastal water ecosystem health.

1.3 Phytoplankton and Optical Complexity

The resultant of the bio-physical coupling addressed in section 1.1–1.2, causes differential growth patterns that (along with other matter) differentially interacts with irradiance. In estuarine and coastal waters, the three optically active constituents (OACs) depending

(42)

on their interaction with visible spectrum of electromagnetic radiation are, (a) CHL, (b) CDOM, and (c) TSM; excluding CHL, that are inherent to water quality (Cloern et al., 2014), and acts as organic source and sink of carbon (Fichot and Benner, 2012), and turbidity (Burchard et al., 2018), respectively. These three OACs are monitored globally through optical remote sensing for decades and provides high resolution spatio-temporal synoptic view influencing ecosystem health with direct implications on climate change (Miller and Cruise, 1995). With respect to the objectives of thesis, the focus is stressed upon accurate optical remote sensing of CHL growth that is a resultant of straining induced advection that supplies nutrients over short-term tidal time scales.

In optical remote sensing of water bodies, the spatio-temporal variations in the colum- nar attenuation of light depends fundamentally on the prevailing dissolved and particu- late matter of organic and inorganic origin. Jerlov (1968) introduced the concept of the classification water types that was based on the spectral optical attenuation depth, i.e., zc = 1/kd, where kd is the attenuation coefficient of downwelling irradiance. This criteria allowed classification of water types in five general oceanic sets and nine coastal water sets (see table XX in Jerlov, 1968). A clear distinction was observed between sets as a function of transmittance (%) per meter of water column, where transmission percentage was the highest at lower wavelengths of the visible spectrum in the clearest of oceanic waters.

Further, Morel and Prieur (1977) modified the characterization of water types based on reflectance variability, and introduced, R = Eu/Ed, where Eu denotes upwelling irradi- ance and Ed represents downwelling irradiance. Following this, a notion of two different classes of water types were introduced, namely, case I and case II.

Traditionally, case I water types corresponds to waters where chlorophyll-a concentration is relatively higher than the mass-specific-scattering coefficients that results a minima in spectral R formed around 440 nm and a maxima shifts around 565–570 nm, in addition to water molecules. In accordance, a second spectral minima in spectral R is observed around 665 nm, and a second maxima is formed around 681 nm due to fluorescence by

(43)

Figure 1.3 Defining inherent optical properties (adapted from Mobley (1994)).

chlorophyll itself. In contrast, case II waters have the effect of other particulate and dissolved matter of organic and inorganic origin in addition to chlorophyll. Therefore, spectral R is generally higher at longer wavelengths when compared with case I, and has a different shape. The spectral R conforms between 400 and 560 nm to become smoother and flat around 560 nm. The spectral values of R increases with turbidity, and these waters appear blue-green to green, and yellowish to red and red-brown.

The bio-optical nature of optically complex case II water types resulting due to integral effect of the biogeochemical processes that are responsible for the alterations in concentra- tion of major optically active constituents (OACs), i.e., CHL,CDOM, and TSM. It must be noted here that TSM includes all particulate matter except for active-CHL. Therefore to put in perspective of optical remote sensing, when a photon interacts with matter, it is either absorbed or scattered. Disappearance of a photon is indicative of absorption which is conversion to another form of energy such as heat while change in direction and/or energy of photon is termed as scattering (Figure 1.3). Collectively, the process of ab- sorption and scattering are termed as the inherent optical properties (IOPs). Figure 1.3 shows incident monochromatic light of spectral power φi(λ) (W m−2) is passed through a volume of water parcel, ∆V, and thickness, ∆r. Some part ofφi is absorbed within the water parcel, i.e., φa(λ), and some part is scattered (φs(λ, ψ)) at an angle ψ. Therefore, the remaining energy is scattered in all directions while φt passes through media without

(44)

Figure 1.4 Spectral absorption of optically active constituents in optically complex waters.

any change in direction, and by conservation of energy it is shown that,

φi(λ) =φa(λ) +φs(λ) +φt(λ) (1.11)

Hence, in a nutshell, absorption coefficient describes how the medium absorbs light, and scattering subjects to how light is directed in different directions and enables to compre- hend the optical properties of media.

In the visible part of the electromagnetic spectrum (400–700 nm), the general absorption spectra of CHL, CDOM, and TSM in Figure 1.4 shows absorption is additive in nature and is an integral effect due to all OACs. Therefore, properties of spectral R likewise, is also a sum of all OACs, and hence, is direct in application optical remote sensing. The parameters such as R, are termed as apparent optical properties (AOPs), constitute from the combined effect of inherent optical properties of each optically active constituents (for details see Kirk, 1994). R is fundamental to optical remote sensing as it is the form of energy perceived by sensor, and is prone to various forms of degradation, thereby, highly sensitive forms of algorithms for detecting different OACs is critical for coastal and estuarine optical oceanography applications.

(45)

Bio-optical studies have in various parts of the world have focused on near-shore marine environments because of their importance of coastal areas for fisheries, recreation, and of direct assimilation from anthropogenic activities. Dynamic processes occurring in the coastal waters alter in-water IOPs and have a significant impact on the measurements and interpretation of upwelled radiance and thus, remote sensing data. Darecki and Stramski (2004) reported the poor performance of standard algorithms in case II waters in estimat- ing the various in-water constituents from ocean color remote sensing (Odermatt et al., 2012; Blondeau-Patissier et al., 2014; Sathyendranath et al., 2017). Several works in dif- ferent optically complex case II waters showed the necessity to develop regional empirical algorithms associated with limited spatial and temporal application for coastal waters.

In turbid estuarine and coastal waters development of algorithm becomes mandatory, as its application to widely monitor water quality parameters such as water transparency, suspended sediment concentrations and presence of harmful algal blooms in estuarine and coastal environment which can further help maintaining health of the system (Carder and Steward, 1985; Tassan, 1994; Kahru and Mitchell, 1998).

Remote sensing or satellites provides even broader spatial coverage of phytoplankton abundance. The technique is based on the fact that radiance reflected from the sea sur- face in the visible or Photosynthetic-ally Active Radiation (PAR) spectrum (400–700 nm) is related to the concentration of CHL. Because CHL is green, and water color changes from blue to green as chlorophyll concentration increases, the relative color differences can be used as a measure of chlorophyll concentration (Lalli and Parsons, 1997). The satellite based retrieval of ocean color parameters demands an accurate bio-optical algo- rithm especially in the coastal waters. Most operational satellite algorithms are empirical switching band-ratio and developed for open ocean waters. Such algorithms tend to fail in optically complex case II waters. Studies that focus on the relationship between IOPs and AOPs are vital and such is the base of this thesis.

(46)

1.4 Optical Characterization of the Study Area

Optical complexity in the study area (mapped later) alters critically with the onset on southwest Indian monsoon, where increase in riverine discharge adds organic matter and sediments of terrestrial origin. In addition, the presence of a strong horizontal stratifi- cation (Vijith, 2014) limits estuarine exchange that alters spatial distribution of OACs and at a temporal horizontal buoyancy of greater than 10−6 s2. For ease of presentation, the study area is temporally divided in to three seasons, i.e., monsoon (June–September), post-monsoon (October–January), and pre-monsoon (February–May). In monsoon sea- son, the bio-optical data collection (in-situ or satellite measurements) is restricted due to unstable weather and cloud coverage. Therefore, pre-monsoon and post-monsoon seasons are effectively safe for data collection in-field and approach scientific efforts. Further, a background on optical complexity pertaining to the process studies in this thesis and their implications are limited to post-monsoon season here.

In post-monsoon season, the riverine flux decreases that results in the study to eventually become tidally dominated with weaker horizontal stratification. The immediate impact of this is observed in spatio-temporal distribution of the OACs, where effective cycling to organic and inorganic matter in to coastal seas becomes possible. This was first presented by Menon et al., 2011 in form of the CDOM distribution through satellite and in-situ measurements. The spatio-temporal variations in the horizontal distribution of CDOM was delineated from the remnants of freshwater discharge (plume dissociation) and its impact was found significant up to middle reaches. In the lower reaches of the study area, a dissolution of plume resulted in decreased CDOM while upper and middle zones were indistinct (Menon et al., 2011).

Additionally, the spatio-temporal variability of CHL and TSM is dynamically inter- connected in the post-monsoon season. In upper reaches of the study area, the partially- mixed state, shallow depths, less turbidity, and ample light conditions favors CHL con- centration, which is linked with in-situ production of CDOM (Menon et al., 2005b). In

(47)

contrast, higher concentrations of TSM were reported in the lower reaches of the study area that is consistent with elevated tidal action in presence of relatively stronger stratifi- cation in upper reaches (Manojet al., 2009). Therefore, it is worthwhile to note here that interaction of OACs in dynamically active regions are immediate parameters of ecosystem health and directly or indirectly impact phytoplankton growth.

Primary productivity in regions of shallow continental shelves contribute significantly to global carbon fixation. The transformation of inorganic carbon to organic source is the basis of primary growth and is constantly affected by physical processes of different spatio-temporal scales. The research carried out in this thesis primarily focuses on tidal hydro-mechanics and its direct impact of primary growth in a tropical estuarine ecosys- tem. The resultant of prevailing bio-physical coupling resides in ecosystem setting for phytoplankton growth that includes the differential interaction of chlorophyll-a, colored organic dissolved matter, and total suspended matter with incoming light-field and re- sults in optical complexity. Therefore, efforts were made to provide solutions pertinent to the prevailing optical complexity over shallow continental shelves of western India. This resulted in development of two optical algorithms applicable to remote sensor. The first algorithm is shown to be applicable in global optically complex waters, while secondarily, an approach developed to understand non-linearity of suspended matter interaction with light-field resulting in a regional index. Lastly, using the notions developed in different case-studies of this thesis and its implication on climate change through optical remote sensing is presented. The results obtained from objectives of the thesis are applicable to different scenarios and case-studies of global estuarine and coastal waters.

1.5 Objectives and structure of the thesis

The major aim of this thesis is to bridge gaps in the scientific understanding of tidal dynamics, its impact on phytoplankton (an index of primary production) and presents solutions to optical complexity in nearshore coastal with direct global applications.

The thesis begins by investigating the tidal structure and tidal asymmetry of the region,

(48)

presented in chapter 2. The analysis presents an interplay between barotropic and baro- clinic forcing that governs general estuarine circulation of the region. In chapter 3, the process of tidal straining is established in the region which is the first observation from this area. Further, measurements of tidal straining along with microstructure measure- ments of turbulence, and in-situ concentrations of chlorophyll-a were analyzed based on the time evolution of the vertical salinity gradient to provide details on the mechanism of transport pathways (advection) that significantly affects and alter primary productivity constantly.

Secondly, the resultant of the proven bio-physical coupling in region that significantly accounts for optical complexity through various means is addressed in chapter 4. Using the results, a novel remote sensing algorithm was formulated to retrieve chlorophyll-a from optically complex waters. It was for the first time that a chlorophyll-a algorithm could be applied to global optically complex water without fine-tuning to the turbidity of region. The algorithm was subjected to validation in different optical domains with unaltered successful accuracy.

In chapter 5, the implications of developed chlorophyll-a algorithm on global primary pro- ductivity is presented. An analysis of global trends of chlorophyll-a and modeled euphotic depth integrated primary productivity is explained. The affect of accurate computations of primary productivity on climate change are majorly discussed with a strong reasoning on advancement on predictability.

Assessment of the optical complexity that prevails within total suspended matter (excludes living matter) particles is presented in chapter 6. A thorough analysis of backscattering and remote sensing reflectance values of the total suspended matter provided insight as to what causes the non-linear interaction with electromagnetic field. The methodology adopted in the objective was validated with in-situ concentrations of total suspended matter by formulating a regional algorithm. Concluding in chapter 7, major outcomes of this thesis are listed.

References

Related documents

These gains in crop production are unprecedented which is why 5 million small farmers in India in 2008 elected to plant 7.6 million hectares of Bt cotton which

3 Collective bargaining is defined in the ILO’s Collective Bargaining Convention, 1981 (No. 154), as “all negotiations which take place between an employer, a group of employers

The impacts of climate change are increasingly affecting the Horn of Africa, thereby amplifying pre-existing vulnerabilities such as food insecurity and political instability

40 Besides the immediate and direct impact for groups at risk and those workers who have already lost their incomes, well-designed social protection measures can also contribute

Harmonization of requirements of national legislation on international road transport, including requirements for vehicles and road infrastructure ..... Promoting the implementation

In the most recent The global risks report 2019 by the World Economic Forum, environmental risks, including climate change, accounted for three of the top five risks ranked

China loses 0.4 percent of its income in 2021 because of the inefficient diversion of trade away from other more efficient sources, even though there is also significant trade

In shallow coastal areas, such as the study region, sediments exhibit higher biological mixing intensities while a shallow water column and increased productivity (chapter 3) provide