Acidification of the Arctic Ocean EPOCA Kickoff Meeting, Gijon, 11 June 2008 Funding: EU (GOSAC, NOCES), NASA, DOE, Swiss NSF, CSIRO James C. Orr 1, Sara Jutterström 2, Laurent Bopp 3, Leif G. Anderson 2, Victoria J. Fabry 4, Thomas Frölicher 5, Peter Jones 6, Fortunat Joos 5, Ernst Maier-Reimer 7, Joachim Segschneider 7, Marco Steinacher 5 and Didier Swingedouw 8 1 MEL/IAEA, Monaco 2 Dept. of Chemistry, Götenborg University, Sweden 3 LSCE/IPSL, CEA-CNRS-UVSQ, Gif-sur-Yvette, France 4 Dept. of Biological Sciences, California State University San Marcos, USA 5 Climate & Environmental Physics, University of Bern, Switzerland 6 Ocean Sciences Div., Bedford Inst. of Oceanography, Dartmouth, Canada 7 Max Planck Institut für Meteorologie, Hamburg, Germany. 8 Université Catholique de Louvain, Institut d’Astronomie et de Geophysique Georges Lemaitre, Louvain-La-Neuve, Belgium
Decline of surface pH and [CO 3 2- ] during the 21 st century pH reduced by 0.3 to 0.4 by 2100 under IS92a (i.e., a 100% to 150% increase in [H + ]) [CO 3 2- ] decline results in surface undersaturation ( A < 1) in S. Ocean: down to 55+/-5 mol/kg (in 2100, IS92a) Aragonite Saturation Calcite Saturation s 2100i Orr et al (Nature)
Present state of ocean saturation w.r.t. aragonite: [CO 3 2- ] A = [CO 3 2- ] - [CO 3 2- ] A sat By 2100… Large changes in subsurface saturation state ( [CO 3 2- ] A ) [in mol kg-1] Surface ocean is supersaturated everywhere –For at least 400 kyr –& probably 25Ma Aragonite saturation horizon (where [CO 3 2- ] A = 0) –Southern Ocean (down to ~1000 m) –North Atlantic (down to ~3000 m) Surface undersaturation ( [CO 3 2- ] A < 0) –Southern Ocean –Subarctic Pacific Shoaling of the aragonite saturation horizon (i.e., [CO 3 2- ] A = 0) –Southern Ocean (by ~1000 m) –North Atlantic (by ~3000 m) Pacific Atlantic
Uncertainty due to Emissions Scenario (IS92a vs. IPCC SRES scenarios) *From Bern “reduced complexity” model (G.-K Plattner & F. Joos)
Aumont & Bopp (2006) Models: Euphotic Layer ( m) BGC model: PISCES Coupled climate model: IPSL/CM4.1 Atmosphere: LMD Ocean: OPA/ORCA-LIM Model - Resolution: 2° nominal (½° tropics) - Isopycnal Diffusion & GM - TKE Model (prognostic K z ) - Sea ice model (LIM) PO 4 3- Diatoms MicroZoo POM DOM DSi DFe Nano-phyto Meso Zoo NO 3 - NH 4 + Small Part. Big Part.
IPCC Scenarios used for 4 th Assessment Report (AR4) With sulfate aerosols Without sulfate aerosols Year Ctl now Ctl preind
Atmospheric CO 2 Atmospheric CO 2 from 3 coupled carbon-climate models Three fully coupled atmosphere-ocean models (IPCC AR4 WG1 contributors), including ocean & terrestrial carbon modules (C4MIP, Friedlingstein et al., 2006) IPSL.CM4 LOOP (OPA/ORCA2, PISCES) MPIM (MPIOM, HAMOCC5.1) NCAR CSM1.4 (NCOM, OCMIP2+ prognostic) 2xCO 2 Year
Changes differ between 2 Polar Oceans: pH & [CO 3 2- ] Southern Ocean Arctic pHCarbonate
Surface Arctic projected to reach “Ω A < 1” from 10 to 32 years sooner than Southern Ocean (on average), i.e., lower atmospheric pCO 2 by μatm Year Arctic (> 70N)S. Ocean (<60S)Arctic - S. Ocean IPSL MPIM NCAR Atmospheric pCO2 (uatm) Arctic (> 70N)S. Ocean (<60S)Arctic - S. Ocean IPSL MPIM NCAR Model-only projections under SRES A2 scenario
Two “trans-Arctic” sections: (1) Combined AOS-94 + ODEN91 & (2) Beringia 2005 Chukchi Sea East Siberian Sea Laptev Sea Kara Sea Barents Sea Canada Basin Amundsen Basin Nansen Basin Fram Strait Makarov Basin
Trans-Arctic Model vs. Data Evaluation: Temperature ( o C) Salinity
Trans-Arctic Model vs. Data: arag Data Model Model – Data MLD too deep Surface [CO 3 2- ] too high Overall pattern, but less structure
Model minus Data: [CO 3 2- ] along AOS94-ODEN91 IPSL1 IPSL2 NCARMPIM
Model minus Data: [CO 3 2- ] along Beringia 2005 IPSL1 IPSL2 NCAR MPIM
Models vs. Data: mean profile (distance-weighted) AOS94-ODEN91 Beringia 2005
AOS94-ODEN91 Beringia 2005
Projected [CO 3 2- ] A : saturation w.r.t. Aragonite projections from model only (under A2 scenario)
Projected [CO 3 2- ] A : Saturation w.r.t. Aragonite * Beringia (2005) baseline + model perturbations (A2)
Projected [CO 3 2- ] C : Saturation w.r.t. Calcite * Beringia (2005) baseline + model perturbations (A2)
Data-model approach improves consistency of projected undersaturation in Arctic surface waters A (δpCO2) 1 st signsAverageCalcite 1 st signs IPSL2014 (+22)2046 (+117)2059 (+168) MPIM2014 (+18)2048 (+136)2070 (+244) NCAR2014 (+16)2048 (+126)2060 (+180) “Data-Model” projections under SRES A2 scenario along Beringia section
IPCC Scenarios in use for 4 th Assessment Report (AR4) With sulfate aerosols Without sulfate aerosols Year Ctl now Ctl preind
Undersaturation is strongest in the Arctic: simulation with +1% increase per year *Model approach (model results only) Aragonite undersaturation [CO 3 2- ] Arag at 2xCO 2
Why?: Perturbation in [CO 3 2- ] due only to climate change is large and negative in the Arctic (2xCO2)
Mean Arctic profiles at 2xCO 2 with & without terrestrial ice melt CO 3 2- S T AlkDIC + CO2 & Climate & Ice melt Control +CO 2 & Climate
Mean Arctic profiles at 4xCO 2 with & without terrestrial ice melt CO 3 2- S T AlkDIC + CO2 & Climate & Ice melt Control +CO 2 & Climate
Simulated changes in surface [CO 3 2- ] at 2xCO 2 ArcticSouthern Ocean Preindustrial CO 2 only6564 CO 2 + clim (no land ice)6366 CO 2 + clim + land ice5764 Change (total) Change (CO2) Change (clim + land ice)-80 Change (land ice)-5-2 Fraction (CO 2 ) Fraction (clim + land ice) Fraction (land ice)0.08 2xCO 2
Arctic Marine Calcifiers Pelagic: –Foraminifera [calcite] –Shelled pteropod (Limacina helicina) [aragonite] –Coccolithophores (Coccolithus pelagicus, Emiliana huxleyi) [calcite] not the dominant Arctic primary producer Benthic: –Molluscs dominate, particularly bivalve molluscs [calcite & aragonite] –Gastropods, scaphopods (tusk shells) [aragonite] –Echinoderms (Brittle stars, sea stars, sea urchins, sea cucumbers) [high Mg-calcite in internal ossicles] –Benthic forams [calcite], –Coralline red algae [high Mg calcite] –Bryzoans –BUT, No Cold-water corals yet discovered (perhaps too cold) How will Arctic ecosystems respond to ocean acidification?
Effects on other other Arctic animals?
Conclusions With 2 transArctic data sections & 3 models, we projected changes in [CO 3 2- ] and saturation under SRES A2 scenario –Changes w.r.t. Aragonite: Now - some near-subsurface waters already undersaturated (Canada Basin), due to anthropogenic CO 2 increase in 10 years - some surface waters become undersaturated in 40 years - average surface waters become undersaturated –Changes w.r.t. Calcite: in 10 years - near-subsurface waters become undersaturated in 50 years - some surface waters become undersaturated in 70 years - average surface waters become undersaturated –Changes occur 10 to 30 years sooner in Arctic, relative to the Southern Ocean Uncertainties remain (circulation, climate change, terrestrial ice melt/runoff, sea ice, riverine Alk & DIC delivery) Potential loss of Arctic marine calcifiers by 2100? Need for low-temp undersaturated perturbation studies (bivalves, echinoderms, coccolithophores, cold-water corals,…) Need impact studies in currently undersaturated zones (shelves)
Aragonite Saturation along trans-Arctic sections Future [ CO 3 2- ] computed on section after adding model perturbations to data: DIC, Alk, T, S, SiO 2, & PO 4 3- (Historical + SRES A2) Deep saturation horizons resist change Undersaturation invades from surface –Aragonite: surface undersat. by 2050 Aragonite Calcite *Data-Model approach [CO 3 2- ] ARAG
Calcite Saturation along trans-Arctic sections Aragonite Calcite *Data-Model approach [CO 3 2- ] CALC Future [ CO 3 2- ] computed on section after adding model perturbations to data: DIC, Alk, T, S, SiO 2, & PO 4 3- (Historical + SRES A2) Deep saturation horizons resist change Undersaturation invades from surface –Calcite: surface undersat. by 2100
Simulated changes in surface [CO 3 2- ] at 4xCO 2 ArcticSouthern Ocean Preindustrial CO 2 only3639 CO 2 + clim (no land ice)3539 CO 2 + clim + land ice2638 Change (total) Change (CO2) Change (clim + land ice)-100 Change (land ice)-9 Fraction (CO 2 ) Fraction (clim + land ice) Fraction (land ice)0.10 4xCO 2