1

0545 June 22 2007 Recovery of Carbon Flux Explorer

Ocean Carbon and Biogeochemistry (OCB) Summer Workshop, Woods Hole July 24-28 2016IOS/Tully • SIO Instrument Development Group • LBNL Engineering • UCB Undergraduates.

OCE 0936143

Novel Observations of Carbon Export fromAutonomous Lagrangian Carbon Flux Explorers

Menu:

Carbon Flux ExplorerHigh Wintertime Sedimentation in California Coastal Waters

OCE 1538686

With… Michael Fong, Todd WoodTim Loew, Hannah Bourne, Mike Mclune, Mati Kahru

Christina Hamilton, Gabrille Weiss, Imari Walker, Xiao Fuand many more

Net Thermodynamic CO2 uptake, aka “Solubility Pump” quite well understood & empirically measured

The Ocean Bio-Carbon pump is fast, must be understood. Is today’s pump (EXPORT, UPWELLING) in balance?

How do we find out?

2

Photosynthetically Active RadiationSEA SURFACE

Particulate (C, Ca, Si, N, P …)

recycled nutrientsdissolved organic matter (DOM)

PHYTOPLANKTONBACTERIA

MICROZOOPLANKTON

BASE OF EUPHOTIC

ZONE

mixing /diffusionSinking

FECAL MATERIAL

MARINE SNOW

large particles

FECAL MATERIAL

MARINE SNOW

sloppy feeding

MACROZOOPLANKTON(INVERTEBRATES)

FISH

activemigration

dissolutionoxidation

respiration

small particles

excretion

BACTERIA

MICROZOOPLANKTON

DOM

Upwelling“New” nutrientsPO4, NO3, SiO3trace elements

Dust (Fe)BIOLOGICAL

CARBON PUMPN2O2

Lateral (Fe…)

Particle Concentration

and Flux Important

Indicators of Processes

Modified Scripps SOLO float

Fast Profiling (diurnal profiles to 1000m)

Long Lived ~1 year(battery limited)

Argos AntennaBi-directional Iridium Satellite

Telemetry & GPS

Temperature, SalinityParticulate Organic Carbon

(digital 10 Hz, 14 bits)Particulate Carbon Flux Index

Scattering (analog)

Particulate Inorganic Carbon(digital, 10Hz, 14 bits)

Mission life expected400,000 m of profiles

New Approach to Particle [ ] Dynamics: The Carbon Explorer 2.0

OCE 0964888

8

9

1

12

2

3

3

4 4

5

5

7 76

6

Recovery Loop

BallastingTubes

PIC sensor

Deployed at PAPA Feb 2013

3

Challenges of Measuring Particle Fluxes

1987 Martin et al. formula or varients: used in almost all

carbon cycle models

POC flux (mol C m-2 y-1)

week averaged sample collection at 5 locations N Pacific Ocean during summer.

Short deploymentsTied to ship availability

Surface Tethered Sediment Traps

3 Lightingmodes

5 Mpix Imaging(RGB)

Set f.no, shutter, & focal distance

Image resolution 13 µm

RGB

4

OPTICAL QUANTIFICATION OF EXPORT

Attenuance (derived from transmission) Proxy for Particulate Organic Carbon

Birefringence (Cross Polarized Photon Yield) Proxy for Particulate Inorganic Carbon.

(CaCO3) - shell materials that make aggregates sink

Darkfield: (side illuminated reflectance) Color of particles – Pigments – Fine

structures – e.g. Other modes possible.

POCATN FLUX --- units: mATN-cm2 cm-2 d-1

PICPOL FLUX --- units: ppm-cm2 cm-2 d-1

Bishop et al. (2016) Biogeosciences, 13, 3109–3129, 2016

RAW TRANSMITTED

ATN = -log10(Transmission)

TRANSMITTED reference

Transmission÷

Bishop et al. (2016) Biogeosciences, 13, 3109–3129, 2016

5

AttenuancePOCproxy

5 mm

Resolution15 µm

POL

Pteropod shellCoccolith hazein aggregates

Foraminifera

5 mm

Resolution15 µm

Cross PolarizedPIC proxy

(enhanced)

6

DarkField

5 mm

Resolution15 µm

33.58

33.62

33.66

33.70

33.74

33.78

-119.65 -119.45

Latit

ude

°N

Longitude °W

NH1204NH1301NH1304

(a)

5 km

May 2012Jan. 2013Mar. 2013

Moored trap deploymentsSanta Barbara Basin (SBB)

(e.g. Thunnel 1998)

San Pedro Channel (SPB)(e.g. Collins et al. 2011)

Santa Cruz BasinStudy Site

Does High Surface Chlorophyll mean high carbon export & sedimentation?

7

CFE MissionSanta Cruz Basin

250

500

0

Depth(m)

Time (hours)0800 1600 0000 0800 1600…

Clean, image @ Hour (0, 0.4,

0.8, … 1.6), Clean … repeatSurface every ~8 Hrs,

GPS, data in/outDive to next depth

In animations (later): 1s ~ 2hrs

JAN 2013MAY 2012

Dark Field Illumination CFE001 various depths 150 to 900 m

Does High Surface Chlorophyll mean high carbon sedimentation?

ANSWER : NO

1 cm

8

-10103050

151.5 152.0 152.5 153.0 153.5 154.0 154.5 155.0

0

50

86.0 86.5 87.0 87.5 88.0 88.5 89.0

0

100

200

300

400

18.0 18.5 19.0 19.5 20.0

Jan. 2013

Mar. 2013

May 2012

164 271 503 173261

144320

506

144320

465 715 915 435 670 880 395 635

513167 262

CFE Depth (m)

515

AVERAGE SAMPLE LOADING TIME SERIES POCATN (mATN).

.0

.Days Since Jan 1 0:00 [UTC]

Sequential imaging of

accumulatedParticlesover 1.6

hoursfollowed by

cleaning

Bishop et al. (2016) Biogeosciences.

Calculation of Attenuance Flux:1) Volume Attenuance = Attenuance*Stage area

units: mATN-cm2

2) Volume Attenuance Flux =

ΔVol. Attenuance/ trap area /Δtunits: mATN-cm2 cm-2 d-1

0

100

200

300

400

19.3 19.5 19.7 19.9

Stage area = 5.07 cm2

Trap area = 186.2 cm2

mATN

Bishop et al. (2016) Biogeosciences, 13, 3109–3129, 2016

9

-250

2550

86.0 87.0 88.0 89.0

-25

0

25

50

151.5 152.0 152.5 153.0 153.5 154.0 154.5 155.0

0255075

100125150175

18.0 18.5 19.0 19.5 20.0

.

.0

.Days Since Jan 1 0:00 [UTC]

Jan. 2013

Mar. 2013

May 2012

144320

506

144320

465 915 435 670 880 395 635

CFE Depth (m)

164 271 503 173261

513167 262 515

715

Multiply by ~3 to get mmol C m-2 d-1

Dots – 20 min fluxBars – 1.6 hr averaged flux

POCATN FLUX (mATN-cm2 cm-2 d-1)

Bishop et al. (2016) Biogeosciences, 13, 3109–3129, 2016

-1000

100200300400

18.0 18.5 19.0 19.5 20.0

-25

0

25

151.5 152.0 152.5 153.0 153.5 154.0 154.5 155.0

-1000

100200

86.0 86.5 87.0 87.5 88.0 88.5 89.0

(630) Carapace

.

Days Since Jan 1 0:00 [UTC]

.

+

+ carapace and swimmer effect excluded from avg

.

Dots – 20 min fluxBars – 1.6 hr averaged flux

PICPOL FLUX (ppm-cm2 cm-2 d-1)

144320

506144

320

164 271 503 173261

513167 262 515

465 915 435 670 880 395 635715

Bishop et al. (2016) Biogeosciences, 13, 3109–3129, 2016

10

0.1

1

10

100 300 5002012 days

Satellite POC (μM) and Chlorophyll (mg m-3)

0

5

10

15

20

25

30

0.0

1.0

2.0

3.0

Jan Mar May

POC

(μM

)

Chlo

roph

yll (

mg

m-3

)

CHLPOC

in-s

itu

MA

Y 2

012

JAN

201

3

MA

R 2

013

In situ POC = 27*cp

In-Situ Beam attenuation Coefficient

0

5

10

15

20

25

30

0.0

1.0

2.0

3.0

Jan Mar May

POC

(μM

)

Chlo

roph

yll (

mg

m-3

)

CHLPOC

in-s

itu

0

200

400

600

800

1000

0.1 1.0 10.0 100.0

Dep

th

POCATN FLUX (mATN-cm2 cm-2 d-1)

NH1301NH1304NH1204

0

200

400

600

800

1000

0.1 1.0 10.0 100.0

DEP

TH (m

)

PICPOL FLUX (ppm-cm2 cm-2 d-1)

JAN 2013MAR 2013

MAY 2012

CARBON EXPORT -REMOTE SENSED

BIOMASSComplex

Variability lower at high flux!

Bishop et al. (2016) Biogeosciences, 13, 3109–3129, 2016

11

(2) CALIBRATION OF CFE DATABUOY SYSTEM: Optical Sediment Recorder

Twin OSR Instrumentscollect samples for calibration

Our Calibration is not yet realized due to bias.Only fragments or subset of particles sampled

Surface Tethered OSRCollected 1/20th of the

particlesMissed > 1.5 mm

particlesCFE

BUOY

2.5 cm

320 m

507 m

144 m

Buoy saw little >2 mm material • except • when

currents across trap <2 cm/sec

•

12

California Coastal Waters:• High surface chlorophyll does not imply high

sedimentation. • Low surface chlorophyll (January): a lot hungry grazers –

efficient export.• High surface chlorophyll left-over veggies• Most carbonate is calcite (Forams, Coccoliths)

vs. Preliminary CFE at PAPA – Aragonite &Pteropods• Coastal Zone Carbon Cycle:

– POC Flux (surface tethered baffled traps) likely underestimated by a minimum factor of 3 … at times as high as 20!

– January 2013 Rates exceed highest trap POC flux measured in San Pedro Basin (and Santa Barbara Basin) by 8x.

– Challenge for continuous observations. Cooperative autonomy or modified Glider?

NEXT STEPS for CFE• August 2016. R/V Oceanus.

First deployments of sample collecting CFE’s. & PIT traps and McLane Pumps.

• Expect to have image analysis runningaboard CFEs. Codes as described above.

• Mission capability 8 months @ hourly

PENDING – full autonomy • Detection of living organisms. • Onboard size analysis

particle classification. E.g. coccoliths vs. Foraminifera vs. Pteropods. E.g. aggregate type. Size distribution…

See Hannah Bourne’s poster. OBVIOUS : merge Carbon Explorer and CFE = > OCO float• Add transmissometer, PIC, and scattering …

13

Anthropogenic Perturbation of the Global Carbon Cycle

Perturbation of the global carbon cycle caused by anthropogenic activities, averaged globally for the decade 2003–2012 (GtC/yr)

Source: Le Quéré et al 2013; CDIAC Data; NOAA/ESRL Data; Global Carbon Project 2013

14

JAN 2013

CFE001 various depths 150 to 500 m1 cm

ATN POLDRK

Estimate of POC FLUX

Size analysis => equivalent circular diameter (d)

using ImageJ

+ empirical thickness, h (fn of d)From Bishop et al. 1978.

Aggregate volume = 0.113 cm3

Dry weight Aggregate density 0.087 g/cm3

From Bishop et al. 1978.

and ORG Matter fraction = 60%=> ORG Mass = 0.0059 gOM:C 1.88 (From Hedges)

C moles = 0.00026Then use

Trap area = 0.0186 m2

Time days = 0.0766

POC Flux = 183 mmol C/m2/dVol Attn Flux = 66.2 => factor = 2.820130120_123947_bckr

15

Merged 1 Km Chlorophyll from Modis Aqua/Terra and VIIRS

20130212013020201301920130182013017

20130162013012 201301520130142013013

2013011201301020130082013007

http://spg.ucsd.edu/Satellite_Data/California_Current/

Chlorophyll-a (mg m-3)831.00.40.10.05

Mati Kahru mkahru@ucsd.eduReplotted by Michael Fong

Study Site

Uncorrected MergedResults.

??

Were there significant spatial gradients near our study site at 1 km resolution? We searched 1 km satellite data within 2 km radius of 33.73N, 119.50W and 33.69N 119.58W. The differences were small. Left: mean ± s.d. Right: relative difference ± s.d. between sites.

Jan

2013

Mar

201

3

May

201

2

? ?

http://spg.ucsd.edu/Satellite_Data/California_Current/

CFE

- Ship -

16

2013008 2013014

2013016

4 Km Modis Aqua/TerraAnd VIIRS processed dataFrom Mati Kahrumkahru@ucsd.edu

201302120130202013019

2013012

20130182013017

201301520130142013013

2013011201301020130082013007

Chlorophyll-a

120 W

119 W

34 N

33 N

120 W

119 W

http://spg.ucsd.edu/Satellite_Data/California_Current/

Study Site

Are Baffles a problem?

Aggregates that were lost were ~50% the size of the baffle openings

Horizontal Currents 10-20 cm/sec

17

w settling velocityu

Baffle

K&M 1979. Defined the standard for particle flux sampling in upper oceanImpossible to validate in coastal waters

18

“Each tube had an inside diameter of 7.39 cm and was equipped with a baffle system (SOUTAR et al., 1977) that consisted of 16 smaller tubes (length 7.6cm). The top ends of the baffle tubes had been milled to a wall thickness of 0.06 mm to minimize surface area (about 5% of the cylinder mouth area which is 43 cm 2). We assume that materials hitting these edges fall into the collectors and contribute to the total flux. GARDNER (1977) has shown that open cylinders with a length-to-width ratio of approximately 2 or greater will yield representative fluxes. With our use of a baffle system, an adequate length-to-width ratio (8.4) and density gradients (see below), we assume that our traps sample the vertical flux” of particulate matter with reasonable accuracy. We also have 210pb data (see below) supporting our assumptions. However, like other investigators attempting to measure vertical fluxes, we presently have no way of definitely knowing whether our supposition is correct.

K&M 1979. Defined the standard for particle flux sampling in upper ocean

Tried Adding 2:1 Cylinders

No significant improvement

19

Line P• Corg flux ~ 4.5 to 3 mmol C m-2 d-1

• Significant contamination by Cyprid Barnacle Larve (origin?). Seen in upper 300 m.

• Both CFE002 and CFE003 Saw reproduction event Clio Pteropods at 150 m.

• Most PIC was Aragonite!

-144°W -141°W -138°W -135°W -132°W -129°W -126°W -123°WLongitude

46°N

48°N

50°N

52°N

54°N

56°N

58°N

60°N

Latit

ude

Line P stations June 2013

4000

4000

4000

4000

40002000

2000

2000

200

200

200

200

3000

3000

3000

1000

1000

1000

PAPA20 16

12 0604

CFE DEPLOYED @P16 Bishop/UC-Berkeley

-144°W -141°W -138°W -135°W -132°W -129°W -126°W -123°WLongitude

46°N

48°N

50°N

52°N

54°N

56°N

58°N

60°N

Latit

ude

CFE 002CFE003Deployed

June 13 2013Recovered

June 22 2013

20

-144°W -141°W -138°W -135°W -132°W -129°W -126°W -123°WLongitude

46°N

48°N

50°N

52°N

54°N

56°N

58°N

60°N

Latit

ude

Line P stations June 2013

4000

4000

4000

4000

40002000

2000

2000

200

200

200

200

3000

3000

3000

1000

1000

1000

PAPA20 16

12 0604

CFE DEPLOYED @P16 Bishop/UC-Berkeley

-144°W -141°W -138°W -135°W -132°W -129°W -126°W -123°WLongitude

46°N

48°N

50°N

52°N

54°N

56°N

58°N

60°N

Latit

ude

CFE 002CFE 003Deployed

June 13 2013Recovered

June 22 2013

Monthly composite

CFE 003 at P160

200

400

600

800

162 164 166 168 170 172 174Days since Jan 1 2013

POL

TRA

MOST PIC FLUX IS ARAGONITE

XPOL

21

CFE 003

0

200

400

600

800

162 164 166 168 170 172 174Days since Jan 1 2013

0

100

200

300

400

500

600

700

0 1 2 3 4

whole dive

all dives

mATN-cm2 cm-2 d-1

3 5 7 9 11 13 TEMP

Preliminary POCATN Flux

32.2 34.2 Sal

Multiply by 3 to get mmol C m-2 d-1

CFE MissionSedimentation variability at 3 depths

250

500

0

Depth(m)

Time (hours)0800 1600 0000 0800 1600…

Bishop et al. (2016) Biogeosciences, 13, 3109–3129, 2016

22

May 2011

Aug toSep

2011

JAN 2013MAY 2012

1

0.30.1

0.03

CHL

a

10.3.

13:00 UTC MAY 29 2011

WINDSPEEDm sec-1

Summary: Ocean Known Unknowns• Lack of prediction of Natural Ocean C Pump

which moves 10 Pg C/yr. • No way to assess OBCP stability without

observations. Few last decade.• Remote Sensing biomass to C Export not yet

agreed on – even the sign of the trend with respect to biomass / primary production.

• Coastal Zone: Likely underestimated C export• A Carbon-ARGO program/w. ensemble

deployments of Carbon Explorer/ Carbon Flux Explorer robots in key ocean areas will lead to far better C Prediction.

23

Observations of Inter-annual Variability of C fluxes will lead to prediction of future BIOPUMP changes

Anomalous SSTs have lead to 2nd

year of extreme

drought in CA

http://www.ncdc.noaa.gov/teleconnections/enso/indicators/sea-temp-anom.php?

BIG STEP: “ENSEMBLE” C-Explorer Deployments

Carbon FLUX Explorers to Follow.

0 *2013PAPA

***

*

*

Ensemble Weather Forecast

24

CFE MissionSanta Cruz Basin

250

500

0

Depth(m)

Time (hours)0800 1600 0000 0800 1600…

Bishop et al. (2016) Biogeosciences, 13, 3109–3129, 2016