32 years of sea ice physics and biogeochemistry s.f. ackley

32 Years of Sea Ice Physics and Biogeochemistry

S.F. Ackley

Sea Ice Scales

Antarctic Sea Ice Profile

Antarctic Sea Ice Freeboards

Conclusions through Halftime ~1994

• Algal incorporation in Frazil Ice exceeded incorporation in Columnar Ice

• Nearly all Antarctic sea ice contains measurable Chl content but low concentrations

• Primary mechanism of incorporation is wave pumping in frazil-pancake ice

• Concentrations in ice exceeded that in underlying water column by 10x to 100x

Conclusions through Halftime (~1994)

• Algal growth led to nutrient drawdown, so high algal concentrations required renewal of nutrients from surface sea water

• Top surface freezing in autumn led to convective overturning and fueled a fall bloom of algae (Fritsen et al 1994 from ISW)

• Warming leads to porous sea ice, while cooling can cause brine rejection and upwelling (Golden et al 1997)

Krill under Sea Ice

Credit: Klaus Meiners

Krill swarms

First evidence of extensive krill swarms under ice

Krill under sea ice

2nd Half (2007 to Present)

• SIPEX-E.Antarctic Ice—A Seasonal Progression

• SIMBA-Bellingshausen Sea Ice—Temperature Cycling

• DMS and DMSP Production

• A Sea Ice CO2 Pump

ICESat 0166

ICESat 0055

ICESat 1305

ICESat 1297

ICESat 1290

ICESat 0025

Background: AMSR-E ice conc. 10 Oct. 2007; courtesy of G. Spreen

SIPEX Ice temperature and salinity (iron site)

Photo: Mats Granskog

SIPEX Fe biogeochem stations

ice temperature (°C)

-16 -14 -12 -10 -8 -6 -4 -2 0

mea

n i

ce d

epth

(cm

)

-20

0

20

40

60

80

100

120

station 1 station 3station 5station 6station 8station 10station 11station 13station 14

SIPEX Fe biogeochem stations

ice salinity

0 5 10 15 20m

ea

n i

ce

de

pth

(c

m)

-20

0

20

40

60

80

100

120

station 1 station 5station 6station 8station 10station 11station 13station 14

Figure: Delphine Lannuzel

SIPEX - Ice algal biomass

Time →

Chl

orop

hyll a

(mg

m-2)

Rel

ativ

e ic

e co

re d

epth

(%

)

Most of the algal biomass was at the bottom of the cores,except towards the end of the experiment

Time Series Sampling, How does the sea ice at one site change with time? Occupied one floe for 27 days.

Mixture of Ice Types at Ice Station Belgica

Brussels Site - level first year ice- no flooding

Deformed thick ice with snow cover- Fabra Site (60 to 80% neg. freeboard)

open water lead

Liege Site(off photo)

Icebergs

Brussels site- level ice, thin snow Liege site- mod. roughness, thicker snow

Fabra sitehighly deformed, thick snow cover

Brussels site-smooth, thin snow

C. Fritsen C. Fritsen

M. Lewis

SIMBA Brussels SiteTextureG = granularC = columnarFS = froz. snowf = finem = mediumc = coarses = smalll = large

M. Lewis

SIMBA Liège SiteTextureG = granularC = columnarFS = froz. snowf = finem = mediumc = coarses = smalll = large

Brussels Site IMB

Thinner snow cover allows cold pulseto penetrate sea ice

No flooding at snow/ice interface

Radiometer Buoy (Brussels 1)

Ice thickness loss

Ice Temperature (°C)

-7 -6 -5 -4 -3 -2 -1 0

De

pth

(c

m)

-80

-60

-40

-20

0

Brussels 1

Ice Salinity (-)

0 2 4 6 8 10 12 14

De

pth

(c

m)

-80

-60

-40

-20

0

Brussels 1

Brine Salinity (-)

0 20 40 60 80 100 120

De

pth

(c

m)

-80

-60

-40

-20

0

Brussels 1


-7 -6 -5 -4 -3 -2 -1 0

De

pth

(c

m)

-80

-60

-40

-20

0

Brussels 1Brussels 2


-7 -6 -5 -4 -3 -2 -1 0

De

pth

(c

m)

-80

-60

-40

-20

0

Brussels 1Brussels 2Brussels 3


-7 -6 -5 -4 -3 -2 -1 0

De

pth

(c

m)

-80

-60

-40

-20

0

Brussels 1Brussels 2Brussels 3Brussels 4


-7 -6 -5 -4 -3 -2 -1 0

De

pth

(c

m)

-80

-60

-40

-20

0

Brussels 1Brussels 2Brussels 3Brussels 4Brussels 5

Ice Salinity (-)

0 2 4 6 8 10 12 14

De

pth

(c

m)

-80

-60

-40

-20

0


Ice Salinity (-)

0 2 4 6 8 10 12 14

De

pth

(c

m)

-80

-60

-40

-20

0


Ice Salinity (-)

0 2 4 6 8 10 12 14

De

pth

(c

m)

-80

-60

-40

-20

0


Ice Salinity (-)

0 2 4 6 8 10 12 14

De

pth

(c

m)

-80

-60

-40

-20

0


Brine Salinity (-)

0 20 40 60 80 100 120

De

pth

(c

m)

-80

-60

-40

-20

0


Brine Salinity (-)

0 20 40 60 80 100 120

De

pth

(c

m)

-80

-60

-40

-20

0


Brine Salinity (-)

0 20 40 60 80 100 120

De

pth

(c

m)

-80

-60

-40

-20

0


Brine Salinity (-)

0 20 40 60 80 100 120

De

pth

(c

m)

-80

-60

-40

-20

0


Relative brine volume (%)

0 10 20 30 40

De

pth

(c

m)

-80

-60

-40

-20

0

Brussels 1


0 10 20 30 40

De

pth

(c

m)

-80

-60

-40

-20

0



0 10 20 30 40

De

pth

(c

m)

-80

-60

-40

-20

0



0 10 20 30 40

De

pth

(c

m)

-80

-60

-40

-20

0



0 10 20 30 40

De

pth

(c

m)

-80

-60

-40

-20

0


Thermodynamics at Brussels Site (unflooded)Thermodynamics at Brussels Site (unflooded)


-7 -6 -5 -4 -3 -2 -1 0

Dep

th (

cm

)-80

-60

-40

-20

0

Brussels 1

[ Chlorophyll a ] (µg / l)

0 5 10 15 20 25

Dep

th (

cm)

-80

-60

-40

-20

0

[ DMS ] (nmol / kg ice )

0 500 1000 1500 2000 2500 3000 3500

[Chla] - Brussels 1[DMS] - Brussels 1


0 5 10 15 20 25

Dep

th (

cm)

-80

-60

-40

-20

0

[ DMSP ] (nmol / kg ice )

0 1000 2000 3000 4000 5000 6000

[Chla] - Brussels 1[DMSP] - Brussels 1


-7 -6 -5 -4 -3 -2 -1 0

Dep

th (

cm

)-80

-60

-40

-20

0



0 5 10 15 20 25

Dep

th (

cm)

-80

-60

-40

-20

0


0 1000 2000 3000 4000 5000 6000



0 5 10 15 20 25

Dep

th (

cm)

-80

-60

-40

-20

0


0 500 1000 1500 2000 2500 3000 3500



-7 -6 -5 -4 -3 -2 -1 0

Dep

th (

cm

)-80

-60

-40

-20

0



0 5 10 15 20 25

Dep

th (

cm)

-80

-60

-40

-20

0


0 1000 2000 3000 4000 5000 6000



0 5 10 15 20 25

Dep

th (

cm)

-80

-60

-40

-20

0


0 500 1000 1500 2000 2500 3000 3500



-7 -6 -5 -4 -3 -2 -1 0

Dep

th (

cm

)-80

-60

-40

-20

0



0 5 10 15 20 25

Dep

th (

cm)

-80

-60

-40

-20

0


0 1000 2000 3000 4000 5000 6000



0 5 10 15 20 25

Dep

th (

cm)

-80

-60

-40

-20

0


0 500 1000 1500 2000 2500 3000 3500



-7 -6 -5 -4 -3 -2 -1 0

Dep

th (

cm

)-80

-60

-40

-20

0



0 5 10 15 20 25

Dep

th (

cm)

-80

-60

-40

-20

0


0 1000 2000 3000 4000 5000 6000



0 5 10 15 20 25

Dep

th (

cm)

-80

-60

-40

-20

0


0 500 1000 1500 2000 2500 3000 3500


DMS(P) and Chla evolution at Brussels SiteDMS(P) and Chla evolution at Brussels Site

[Chla] measurements by I. Dumont, [Chla] measurements by I. Dumont, C. Fritsen, B. SaundersC. Fritsen, B. Saunders

Liège Site IMB

Colder Air Temperatures don’t penetrate thick snow cover

Snow/Ice Interface continuouslyflooded with sea water

Sea Ice relatively isothermal

Irregular ice bottom affects sonarreturns. Bottom pinger reset whenCTD removed for repairs.


-7 -6 -5 -4 -3 -2 -1 0

De

pth

(c

m)

-120

-100

-80

-60

-40

-20

0

Liège 1


0 5 10 15 20

Dep

th (

cm)

-120

-100

-80

-60

-40

-20

0


0 1000 2000 3000 4000

[Chla] - Liège 1[DMSP] - Liège 1


0 5 10 15 20

Dep

th (

cm)

-120

-100

-80

-60

-40

-20

0


0 500 1000 1500 2000

[Chla] - Liège 1[DMS] - Liège 1


-7 -6 -5 -4 -3 -2 -1 0

De

pth

(c

m)

-120

-100

-80

-60

-40

-20

0

Liège 1Liège 2


0 5 10 15 20

Dep

th (

cm)

-120

-100

-80

-60

-40

-20

0


0 1000 2000 3000 4000



0 5 10 15 20

Dep

th (

cm)

-120

-100

-80

-60

-40

-20

0


0 500 1000 1500 2000



-7 -6 -5 -4 -3 -2 -1 0

De

pth

(c

m)

-120

-100

-80

-60

-40

-20

0

Liège 1Liège 2Liège 3


0 5 10 15 20

Dep

th (

cm)

-120

-100

-80

-60

-40

-20

0


0 1000 2000 3000 4000



0 5 10 15 20

Dep

th (

cm)

-120

-100

-80

-60

-40

-20

0


0 500 1000 1500 2000



-7 -6 -5 -4 -3 -2 -1 0

De

pth

(c

m)

-120

-100

-80

-60

-40

-20

0

Liège 1Liège 2Liège 3Liège 4


0 5 10 15 20

Dep

th (

cm)

-120

-100

-80

-60

-40

-20

0


0 1000 2000 3000 4000



0 5 10 15 20

Dep

th (

cm)

-120

-100

-80

-60

-40

-20

0


0 500 1000 1500 2000



-7 -6 -5 -4 -3 -2 -1 0

De

pth

(c

m)

-120

-100

-80

-60

-40

-20

0

Liège 1Liège 2Liège 3Liège 4Liège 5


0 5 10 15 20

Dep

th (

cm)

-120

-100

-80

-60

-40

-20

0


0 1000 2000 3000 4000



0 5 10 15 20

Dep

th (

cm)

-120

-100

-80

-60

-40

-20

0


0 500 1000 1500 2000


DMS(P) and Chla evolution at Liège SiteDMS(P) and Chla evolution at Liège Site

[Chla] measurements by I. Dumont, [Chla] measurements by I. Dumont, C. Fritsen, B. SaundersC. Fritsen, B. Saunders

Radiometer indicateschanges in irradiance,snow cover thickness, biological growth

Opening and closing ofLeads in the ice

Breakup of the ice floe where IMBsseparate and driftindependently

SIMBA Accomplishments

Antarctic sea ice is a springtime “Biogeochemical Reactor” Driven by Physical Feedback, where thermal changes drive porosity and convection, leading to:

• Nutrient Flux =>Enhanced Biological Productivity• Biological Degradation=>DMSP+DMS Flux • Porosity Changes=>CO2 Exchanges• Shelf Sediments and Iceberg Melting as Iron

sources

The Sea Ice CO2 Pump

A long lived dogma…

Weiss et al 1979, Gordon et al 1984, Poisson

and Chen 1987

« Weddell Sea pack ice effectively

blocks the air-sea exchange of gases »

No evidence of marked ventilation is

found in deep waters of the Weddel Sea.

Thus sea ice appears to prevent air-sea

exchange.

A long lived dogma challenged?…

Golden et al., 1998

Theoretical and experimental

evidence that sea ice permeability

increases considerably above -

5°C, the so-called “law of fives”

(Golden et al., 1998. Science 282:

2238)

Science 282: 2238

Kelley & Gosink 1970s-80s« unlike ices from pure freshwater, sea ice is a highly permeable

medium for gases »They found rate of CO2 penetration about 60 cm h-1 at -7°C

« gas migration through sea ice is an important factor in ocean-atmosphere winter communication particularly when the surface temperature is > -10° » (Gosink et al., 1976. Nature 263: 41)

Sea Ice Phase Diagram

CaCO3 dissolution/precipitation

Thompson and Nelson 1956 showed that at a temperature just below the freezing point calcium carbonate begins to precipate from the entrained brines in sea ice and remains in the ice.

Weiss et al 1979 observed alkalinity anomalies in the surface water of the Weddell Sea

Jones et al. 1983 observed a transfer of CaCO3 from the ice to surface waters in the Arctic Ocean

Papadimitriou et al. 2004 and Dieckmann et al. observed CaCO3 precipitation in artificial and natural sea ice and identified it as Ikaite

Rysgaard et al. 2007 suggested that CaCO3 precipitation in sea ice might act as a sink for CO2.

2HCO3- + Ca2+ CaCO3 + CO2

2HCO3- + Ca2+

CaCO3 + CO2

CO2

In winter, precipitation of In winter, precipitation of CaCOCaCO33 occurs within sea occurs within sea ice. ice. Produced COProduced CO22 is expelled is expelled with brine, while CaCOwith brine, while CaCO33 is is trapped within brine trapped within brine channelschannels

Brine sink rapidly carrying Brine sink rapidly carrying COCO2 2 (see Brine Drainage (see Brine Drainage Slide)Slide)

Part of COPart of CO22 which passes which passes below the pycnocline is below the pycnocline is « removed » from the « removed » from the system.system.

fall/winterfall/winter

SEA ICE PUMPSEA ICE PUMPA potential abiotic CaCOA potential abiotic CaCO33 Carbon source Carbon source

Is COIs CO22 released to the released to the atmosphere anytime atmosphere anytime new ice forms?new ice forms?(Answer on Next Slide)(Answer on Next Slide)

D. Nomura, 2006D. Nomura, 2006

4545

Rysgaard et al., 2007, Delille et al., in prep.Rysgaard et al., 2007, Delille et al., in prep.

Upward CO2 Flux over New Ice at the SIMBA site

A B C D E-2.5

0.0

2.5

5.0over iceover snow

Air

-ic

e C

O2 f

lux

es

(mm

ol m

-2d

-1)

-9

-8

-7

-6

-5

-4

Inte

rfa

ce t

emp

era

ture

(°C

)

Figure 17:Temperature at the ice interface and CO2 fluxes over ice and snow (positive fluxes correspond to efflux to the atmosphere) at 5 sites of the second "frost flowers" station.

Brine Drainage at SIMBA

2HCO3- + Ca2+

CaCO3 + CO2

CaCO3 + CO2

2HCO3- + Ca2+

CO2

CO2

In spring, CaCOIn spring, CaCO33 trapped within sea ice trapped within sea ice dissolves. This dissolves. This process consumes process consumes COCO2.2.

Budget of winter Budget of winter and spring processes and spring processes is a net sink of COis a net sink of CO2.2. It It depends on:depends on:

ratio of CaCOratio of CaCO33 trapped vs COtrapped vs CO22 expelled (?)expelled (?)

quantity of COquantity of CO22 which pass below the which pass below the pycnocline during the pycnocline during the autumn-winter (?)autumn-winter (?)

fall/winterfall/winter springspring

GAS COMPOSITION IN SEA ICEGAS COMPOSITION IN SEA ICEA potential abiotic CaCOA potential abiotic CaCO33 Carbon pump Carbon pump

4646

Rysgaard et al., 2007, Delille et al., in prep.Rysgaard et al., 2007, Delille et al., in prep.

Sea ice exhibits marked CO2 dynamics controlled by (i) salinity (ii) precipitation/dissolution of CaCO3 and (iii) primary production

CaCO3 precipitation occurs within sea ice and can be a very efficient pathway for atmsopheric CO2 uptake

Sea ice in spring exchanges CO2 with the atmosphere. Sea ice acts first as a source of CO2 to the atmosphere then as a sink.

This spring sink of the antarctic sea ice is about -0.025 PgC yr. It would represent an additional sink of 50% to the CO2 sink of the Southern Ocean

Taking into account CaCO3 precipitation and particular sea-ice processes especially discrepency in salt:gas rejection (loose et al. 2009) and possible artefacts in the transcient hallogen tracers, it might deserve to revisite C anthropogenic computation in sea ice covered waters

Conclusions

Where are we?

• Strong Coupling between the biogeochemistry and physics of the ice, related to growth, thermal driving, snow

• Climatically active gases, DMS and CO2, are important new developments

• Time series from IMBs with additional sensors like radiometers, oxygen, pCO2 look important

• Modeling of fluid flow looks critical• Subtle differences in conditions can lead to big

differences in the biogeochemistry

It’s been fun

Thanks to Elizabeth for the invite.

GAS COMPOSITION IN SEA ICEGAS COMPOSITION IN SEA ICECOCO22 : the ”trouble maker” : the ”trouble maker”

SummarySummary

• pCOpCO22 is a is a difficult variable to measuredifficult variable to measure in the sea ice environment in the sea ice environment• a suite of a suite of measurement techniques are still in developmentmeasurement techniques are still in development and and

in the need of proper validationin the need of proper validation• pCOpCO22 is generally is generally supersaturatedsupersaturated in most of the sea ice cover in most of the sea ice cover

during the during the winterwinter• pCOpCO22 is highly is highly undersaturatedundersaturated in the whole sea ice cover during in the whole sea ice cover during

spring and spring and summersummer• Potential and concurrent mechanismsPotential and concurrent mechanisms for the pCO for the pCO22 drawdown drawdown

during the summer are:during the summer are:Dilution of brinesDilution of brinesDissolution of calcium carbonateDissolution of calcium carbonatePrimary ProductionPrimary Production

• There is a potential There is a potential inorganic CaCOinorganic CaCO33 CO CO22 pump pump associated to sea associated to sea

ice growth and decayice growth and decay

4848

From land fast sea ice to multiyear pack ice

Ispol drift experimentR.V. PolarsternNov-Dec 2004First and multi-year pack ice

AA03-V1 cruiseR.V. Aurora AustralisSep-Oct 2003First year pack ice

Simba drift experimentR.V. N.B. PalmerOct 2007First year pack ice

66°3

9' S

66°3

8‘30

S66

°38’

S

Delille al. , 2007

GAS COMPOSITION IN SEA ICEGAS COMPOSITION IN SEA ICECOCO22 : the ”trouble maker” : the ”trouble maker” Balancing the effects…Balancing the effects…

Dumont d’Urville, 1999Dumont d’Urville, 1999It works!...and all three processes contribute!It works!...and all three processes contribute!

(from Solubility equations)(from Solubility equations)

(from Oxygen production)(from Oxygen production)

(from Total alcalinity anomaly)(from Total alcalinity anomaly)

4747

Key processes ? (Sal brine > sal seawater)

-8 -7 -6 -5 -4 -3 -2 -10

250

500

750

1000 AA 2003/V1 Ispol

atmospheric

concentration

sea ice temperature (°C)

pC

O2 (

pp

m)

-8 -7 -6 -5 -4 -3 -2 -10

250

500

750

1000 AA2003 Ispol

atmospheric

concentration


pC

O2 (

pp

m)

-8 -7 -6 -5 -4 -3 -2 -10

250

500

750

1000 spring summer

dilution effect


pC

O2 (

pp

m)

Key processes ? (Sal brine > sal seawater)

-8 -7 -6 -5 -4 -3 -2 -10

25

50

75

100

125


Sa

linit

y

-8 -7 -6 -5 -4 -3 -2 -10

500

1000

1500

2000

2500

3000

3500


TA

35

(µm

ol k

g-1

)

-8 -7 -6 -5 -4 -3 -2 -175

100

125

150

saturation


O2 (

% s

atu

rati

on

)

Air-ice gas transfer

PermeableImpermermeable

Scaled using sea ice temperature derived from the NEMO-LIM model

Spring air-ice fluxes of COSpring air-ice fluxes of CO22 for the Antarctic sea ice cover is assessed to - for the Antarctic sea ice cover is assessed to -

0.025 PgC from October to December0.025 PgC from October to December

mmol m² d-1

CO2 fluxes

Spring antarctic sea ice cover would represent an additional sink of about Spring antarctic sea ice cover would represent an additional sink of about 50% of the overall CO50% of the overall CO22 sink of the Southern Ocean sink of the Southern Ocean

Air-ice fluxes:Air-ice fluxes:-0.025 Pg-0.025 Pg

Air-sea fluxes south of 50°S:Air-sea fluxes south of 50°S:-0.05 Pg yr-1-0.05 Pg yr-1

Independent assessment

related CO2 transfer

from the atmosphere

(mmol m-2)

temperature increase and related dilution

-60

Primary production -25

CaCO3 dissolution -57

Estimates of potential air-ice CO2 fluxes related to spring and summer physical and biogeochemical processes observed during the 2003/V1 and ISPOL cruises. Flux representative of a 4 months period.The overall CO2 fluxes reach 142 mmol m-2.

Taking into account a mean value for the Antarctic sea ice edge surface area of 16×106 km2, the corresponding overall CO2 fluxes account for 0.029 PgC.

This compares favourably with our previous estimate of an additional sink of 0.025 PgC.

Sea ice exchanges CO2 with the atmosphere

Semiletov et al. 2004


Zemelink et al. 2006




Nomura et al. T2-017


Heineschet al.

Heinesch et al. This study

This study

Back of the envelope computation









Heineschet al.

Heinesch et al. This study

This study

-1 gC m-2 month-1

Back of the envelope computation

-1 gC m-2 month-1Raw mean of spring air-ice CO2 fluxesSpring surface of antarctic sea ice cover

20* 106 km²

Time length of fluxes 2 months

Overall spring antarcticair-ice CO2 fluxes - 0.04 PgC yr-1

Overall S.O. open water fluxes (Takahashi et al. 2009)

- 0.05 PgC yr-1

32 years of sea ice physics and biogeochemistry s.f. ackley

Documents

sea watersea ice

porous sea ice

frazil ice

ice floe

year ice

isothermalirregular

years of sea ice physics

floodingdeformed thick