Post on 13-May-2022
Rrs(λ) IOPs
What to do with the retrieved IOPs?
Chlorophyll concentration: [Chl]
)(pha
)()()()( 21 dgphw aMaMaa
))(),(()( brs baFR
Examples: Carder et al (1999), GSM (2002), GIOP (2013)
:)(pha Varies with temperature in Carder et al (1999); Fixed globally in GSM (Maritorena et al 2002); Varies with [Chl] in GIOP (Werdell et al 2013)
“On a global scale, marine phytoplankton consume fifty thousand million tones of carbon every year in a process referred to as primary production.” -- IOCCG Report #2
“One of the principal applications of satellite ocean color data is to derive net primary production (NPP).” --- McClain (Annu. Rev. Mar. Sci., 2009)
food chain; carbon cycle
1. Estimate global primary production (PP)
Why chlorophyll concentration?
Elements of photosynthesis:
CO2 + H2O + phytoplankton + light + nutrient chemical energy
Components for PP quantification:
(Platt and Sathyendranath, 2007)
1. Phytoplankton 2. Light penetration
3. Photosynthesis parameters
[chl]
Traditional strategy: [chl] centered system
Works for waters where IOPs co-vary with [Chl].
Ocean color spectrum
Light field
color
PP
auxi.
Works for most waters.
Kd()
IOPs
a, bb, aph
IOP-based approach:
1.1 Estimation of diffuse attenuation coefficient (Kd) (for light field)
)(
)()490(
2
1
Rrs
RrsKd
Or,
ChlKd )(
2
3
b
1
a
21
)490(
aa)490(
XbK
or
XK
d
d
)551(
)488(
)551(
)488(
Rrs
RrsX
or
Lwn
LwnX
with
(Darecki and Stramisk, 2004)
Ratio-derived Kd
odbdoubuu EbrEbraEdz
d )(
odbdoubud EbraEbrEdz
d)(
Aas (1987) two-stream solution:
')','()','()cos(4
wdLLczd
Ld
oubuodbdodd EbrEbrbEcEdz
d )(
u
u
bud
d
bdd E
brE
braE
dz
d
b
u
u
d
d
d
d bRrra
K
oubuodbdd EbrEbraEdz
d )(
)( bdd baDK
Empirical approximation:
bd baK v)005.01( s
)52.01(18.4v 8.10 ae
Rrs a&bb dK
(Lee et al. 2005)
No division of “Case 1” or “Case 2” waters.
Kd(490) - measured [m-1
]
0.03 0.05 0.3 0.5 3 50.1 1
Kd(4
90
) [
m-1
]
0.03
0.05
0.3
0.5
3
5
0.1
1
1:1
Arabian Sea
GOM
Baltic
)490(dK
(IOP-model)
(Lee et al. 2005) 0.03 0.05 0.3 0.5 3 50.1 1
0.25
0.75
1.25
1.75
2.25
0.00
0.50
1.00
1.50
2.00Arabian Sea
GOM
Baltic
)4
90
(d
er
dK
)4
90
(m
ead
K/
)490(dK - measured [m-1]
(IOP-model)
0.75
1.25
Oceanic & Coastal waters
Other wavelengths:
X Data
350 400 450 500 550
Y D
ata
0.0
0.1
0.2
0.3
profile-Kd
IOPs-Kd
X Data
350 400 450 500 550
Y D
ata
0.00
0.02
0.04
0.06
0.08
profile-Kd
IOPs-Kd
Wavelength [nm]
Kd [
m-1
]
Wavelength [nm]
Kd [
m-1
]
(a) (b)
(Lee et al 2013)
zKPARePARzPAR
)0()(
1.2 Attenuation of PAR (KPAR)
Good for earlier days, Not so good for the 21st century
(Morel, 1988, JGR)
Change of light quality with depth:
deEzPAR zzK
700
400
),(
0 )0,()(
(Kirk 1994)
Light at deeper depth is associated with lower attenuation coefficient
)0(
)(ln
1
PAR
zPAR
zK PAR
KPAR is light-quality weighted!
Broadband attenuation is light dependent, so no-longer additive!
zzK parePARzPAR)(
)0()(
5.0
21
)1()(
z
KKzKPAR
Euphotic depth (zeu):
%1)0(
)(
PAR
zPAR
)(
6.4
euPAR
euzK
z
Key: KPAR varies with depth, especially in the upper water column!
[Chl]
Zeu (water clarity) rsR
PAR(0)
PAR(zeu)
1.3 Euphotic depth (Zeu)
%1)0(
)(
PAR
zPAR eu
(Morel 1988)
[Chl] approach
[Chl] (empirical) approach:
Rrs(λ) [Chl] zeu
IOP approach:
Rrs(λ) a(λ)&bb(λ) KPAR(z) zeu
5.0
21
)1(
))490(&)490(()490(&)490()(
z
baKbaKzK b
bPAR
Measured Zeu
[m]
0 20 40 60 80
Rrs
deri
ved Z
eu [
m]
0
20
40
60
80 1:1
1:1 R
rs-d
eri
ved
Zeu [m
]
Measured zeu [m]
IOP approach
Water clarity (zeu)
(Lee et al 2005)
water clarity (m)
Global distribution of Zeu
Zeu can be ‘measured’ with modern optical-electronic system
Ocean color Kd, aph PP
dtdzaztEzPP ph ),(),,()( 0
1.4 IOP based PP estimation:
(Kiefer and Mitchell, 1983, L&O)
ϕ
represents ‘photosynthesis’
φ: quantum yield of photosynthesis
measured production (mol/l/day)
0.3 31 10
calc
ula
ted
pro
duct
ion
(
mol
/l/d
ay)
0.3
3
1
10 aph-centered
Chl-centered
1:1
(Lee et al., Appl. Opt., 1996) (Lee et al., JGR, 2011)
Remotely-estimated PP compared with measured PP
Essence of present satellite Chl product:
b
brs
ba
bGR
)550()550(
)440()440(
)440(
)550(
)550(
)440(
bpbw
bpbw
rs
rs
bb
bb
a
a
R
R
Chlaaa
Chlaaa
a
a
R
R
phdgw
phdgw
rs
rs
)440()440()440(
)550()550()550(
)440(
)550(
)550(
)440(*
*
Change of Rrs band ratio not necessarily represents change of [Chl]!
)(
)(
2
1
rs
rs
R
RfunChl
Why aph based approach is likely better?
(Szeto et al 2011, JGR)
The change of Rrs ratio really reflects change of total absorption!
(Lee et al, 2010, JGR)
Rat
io-d
eriv
ed C
hl
(Carder et al, 1999, JGR)
To accurately retrieve spatially and temporally varying Chl from Rrs, we need:
1. Remove the influence of detritus/CDOM and particles 2. Take into account the spatial/temporal variation of a*ph
A promising effort:
Brief summary about current satellite Chl product: • The map of Rrs ratio represents a map of total absorption coefficient. • The GSM-derived Chl represents more of phytoplankton absorption coefficient.
pb:
(Platt et al 2008, RSE)
Pb is also dependent on a*ph,
which is not a constant either for a given Chl nor for varying Chl!
(Behrenfeld and Falkowski, 1997)
*
phm
b ap
Carbon based Production Model (CbPM)
)(/)( 0 geu IfIhZChlNPP
actually aph from GSM )(/ gIf
C
Chl
(Behrenfeld et al 2005; Westberry et al 2008)
Another example of using remotely sensed IOPs for PP
bbp
aph from GSM
‘IOP-based growth rate’
‘IOP-based NPP’
color
Works for most waters.
Kd()
KPAR
IOP
a, bb, aph
[chl]
ZSD, Zeu
[SPM]
{others}
2. Example of other applications of IOP products
PP
2.1 Secchi depth (ZSD)
Water clarity
(Olmanson et al 2008)
empirical approach:
(Doron et al 2007) )490(&)490( bba
IOP approach:
(Arnone et al 2004)
2.2 Water mass classification
(Carnizzaro et al 2006)
2.3 HAB identification-1
aph(657)
(Lee and Carder 2004)
2.3 HAB identification-2
Rrs(λ) aph(λ) = a(λ) – adg(λ) – aw(λ) QAA
(Craig et al 2006)
K. b
revi
s
(Shang et al, 2012)
2.4 Bloom dynamics
Distribution of aph(440) at Luzon Street
(Shang et al, 2012)
Monthly distribution of aph(440) at Luzon Street
2.5 Global physiology of ocean phytoplankton
φ: quantum yield of fluorescence (Behrenfeld et al 2009)
(Behrenfeld et al 2009)
IOP
(Vodacek et al 1997)
2.6 Salinity estimation
(Castellio et al 1999) SSS = x aCDOM + y
Lohrenz and Cai (2006) pCO2 = f(T, S, Chl)
2.7 pCO2 estimation
2.8 Concentrations of mineral particles (SPM; TSM)
Many publications in the literature
bbp [SPM]
(Neukermans et al 2009)
Key Points:
1. Many applications traditionally built around [Chl] can be built around IOPs.
2. Remote sensing and applications centered around IOPs avoided, when necessary, concentration-normalized optical properties.
3. With IOPs as inputs, many products, e.g. Kd, Zeu, PP, could be estimated easily and more accurately.
4. When IOPs are known, many other applications could
be carried out. Be creative!!!