Sensing primary production from ocean color: Puzzle pieces and their status

ZhongPing LeeUniversity of Massachusetts Boston

An effort started half century ago …

~15 Gt/year

From over 7000 measurements

Global PP:

Longhurst et al (1995): 45-50 Gt C year-1

Antoine et al (1996): 36.5-45.6 Gt C year-1

Behrenfeld and Falkowski (1997): 43.5 Gt C year-1

~3 times higher than estimates in the 50-60s’!

(Behrenfeld et al 2005)

Large spatial differences from different sensing models

Puzzle pieces to sense PP in the ocean

(Platt and Sathyendranath, 2007)

3. Phytoplankton index2. Light at depth

4. Energy conversion

(spectral attenuation)

(nutrient)

1. Input energy

1. Input energy

(Frouin et al 1989, 2003)

PAR (Photosynthetic Available Radiation)

Spectral irradiance

(Gregg and Carder 1990)

zKdd

deEzE )()0,(),(

Q: How to get Kd(λ) of varying water bodies?

2. Light at depth(spectral attenuation)

Rrs

Method 1:empirical

Rrs(λ1)/Rrs(λ2)

Algorithms to get Kd

Current operational standard

Rrs [Chl]

Method 2: empirical empirical

Rrs(λ1)/Rrs(λ2)

KddK

Kd )490(dK

Rrs a&bb

Method 3: semi-analytical semi-analyticalKddK

(QAA)

Kd(490) - measured [m-1]

0.03 0.05 0.3 0.5 3 50.1 1

K d(49

0) [

m-1]

0.03

0.05

0.3

0.5

3

5

0.1

1

1:1Arabian SeaGOMBaltic

(b: Method 2)

Kd(490) - measured [m-1]

0.03 0.05 0.3 0.5 3 50.1 1

Kd(

490)

[m

-1]

0.03

0.05

0.3

0.5

3

5

0.1

1

1:1Arabian SeaGOMBaltic

)49

0(dK

)490(dK

)49

0(dK

Kd(490) - measured [m-1]

0.03 0.05 0.3 0.5 3 50.1 1

Kd(

490)

[m

-1]

0.03

0.05

0.3

0.5

3

5

0.1

1

1:1Arabian SeaGOMBaltic

(a: Method 1)

)49

0(dK

)490(dK )490(dK

(c: Method 3)

(Lee et al. 2005)

Oceanic &Coastal waters

Wavelength [nm]

Spec

tral

Kd

[m-1

]

X Data

350 400 450 500 550

Y D

ata

0.00

0.02

0.04

0.06

0.08

measuredsensed

X Data

0.03 0.30.01 0.1 1

Y D

ata

0.03

0.3

0.01

0.1

11:1IOP

IOPs

-Kd(4

90)

[m-1

]

Profile Kd(490) [m-1]

The NOMAD set (1243 data points)

Kd through IOPs

Profile-Kd(490) [m-1]

Ratio

-der

ived

Kd(4

90)

[m-1

]

X Data

0.00 0.05 0.10 0.15 0.20

Y D

ata

0.00

0.05

0.10

0.15

0.20

0.25

< 301.0930-600.99> 600.89

IOP-

base

d K d(4

90)

[m-1

]

Profile-Kd(490) [m-1]X Data

0.00 0.05 0.10 0.15 0.20

Y D

ata

0.00

0.05

0.10

0.15

0.20

0.25

< 300.9930-601.00> 601.00

Different sun angles:

Empirical ratio Through IOPs

How Kd in the UVA/UVB varies globally?Challenges:

Spectral Kd can be well derived based on physics!

(Behrenfeld and Falkowski, 1997)

VGPM:

Chl became the index!!

3. Phytoplankton index

Essence of Rrs-ratio derived Chl product:

b

brs ba

bGR

)550()550()440()440(

)440()550(

)550()440(

bpbw

bpbw

rs

rs

bbbb

aa

RR

ChlaaaChlaaa

aa

RR

phdgw

phdgw

rs

rs

)440()440()440()550()550()550(

)440()550(

)550()440(

*

*

)()(

2

1

rs

rs

RRfunChl

Simple ratio actually involves more than one variable!

(Szeto et al 2011, JGR) Simple ratio dismissed spatial/temporal variation!

Nature of ratio-derived “Chl”

At the center of South Pacific Gyre

May 2009, Global, MODIS

Ratio-derived “Chl” is re-scaled total absorption coefficient!

Rrs

-rat

io d

eriv

ed C

hl [

mg/

m3 ]

Analytically derived a(443) [m-1] Rrs

-rat

io d

eriv

ed C

hl [

mg/

m3 ]

Analytically derived a(443) [m-1]

(Behrenfeld and Falkowski 1997)

4. Energy conversion

(Platt et al, RSE, 2008)

Variation of phytoplankton- (or chlorophyll-) specific absorption coefficient (a*ph) contributes largely to the variation of PB

opt.

DLzChlPARPPP euBopteu

DLzaPARPP eupheu

Centered on Chl

Centered on absorption

Both PBopt and Chl have a*ph associated

Increase uncertainty in PP

No engagement of a*ph

“Site-specific and previously published global models of primary production both perform poorly and account for less than 40% of the variance in ʃPP,” (Siegel et al 2001)

“significant improvements in estimating oceanic primary production will not be forthcoming without considerable advance in our ability to predict temporal and spatial variability in PB

opt”. (Behrenfeld and Falkowski 1997)

Chl is NOT the direction to go.

PP estimation based on phytoplankton absorption (aph):

Ocean color aph PP

dtdzaztEztzPP ph ),(),,(),()( 0

Phytoplankton index

Quantum yield for photosynthesisRemotely sensible

(Marra et al 2007, Deep Sea Res.)

R2 = 0.84 R2 = 0.78

0

0bpbp )()( bb

Rrs()

)()( 0 aa

)(),(),(F)( 0bw0020bp baub

)(),(),(F)( bwbp3 bbua

η (± Δη)

)(F)( 1 rsRu

))(),(),(),((F)(

))(),(),(),((F)(

2152

2142

aaa

aaa

dg

ph

U1

U2

U3 U4

(Lee et al. Appl. Opt., 2002)The Quasi-analytical algorithm (QAA)

)()()()(

b

b

babu

a ph(λ

) (m

-1)

(Lee et al 2004)

Measured vs sensed aph

measured production (mol/l/day)0.3 31 10

calc

ulat

ed p

rodu

ctio

n (

mol

/l/da

y )

0.3

3

1

10 aph-centeredChl-centered1:1

(Lee et al. 1996)(Lee et al. 2010)

Absorption-based PP compared with measured PP

100 200 300 400 500 600 700100

200

300

400

500

600

700

Chl-PPB

Chl-PPA

Aph-PP

R2 = 0.29

R2 = 0.74

PPeu from on-deck incubation [mg C m-2 d-1] PP

eufr

om m

odel

[m

g C

m-2

d-1 ]

1:1R2 = 0.44

Where is the global model for φ?

Challenges:

Which ‘ground truth’ we remote-sensors should aim at?

Summary:1. A frame work for sensing primary production is well established.

2. Optical/light related parameters can now be well retrieved from satellite measurements, at least for oceanic waters.

3. Demands support and hard work to understand and quantify photo-physiological effects.

4. Demands true “ground truth”!

Questions?