Interactions between ocean biogeochemistry and climate Guest presentation for AT 762 Taka Ito How does marine biogeochemistry interact with climate? What can we learn from past climate changes? What are the implications for future climate change?

Outline Introductions - Coupling mechanisms - Two-way interactions Insights from past climate changes - Climate of the past 60 million years - Glacial-interglacial climate change Present and future climate - The future of oceanic carbon uptake - The Southern Ocean

How does marine biology impacts on climate? Modified from Sarmiento and Gruber (2006) Air-sea CO 2 exchange SW absorption DMS emission Greenhouse effect Clouds SST Ocean carbon cycle Photosynthesis by phytoplankton

How does climate impact on marine biology? Ocean circulation, nutrient transport Eddy, frontal scale ~ 100km Cold, nutrient- rich coastal water Warm open ocean High biological production Low biological production Coastal upwelling

Ocean currents control marine biology Basin, planetary scale ~ 10,000 km Large-scale atmospheric wind patterns Upwelling Downwelling Upwelling

Marine Biology impacts on CO 2 fluxes 90+% of the global carbon is stored in the deep ocean Each year, ~ 2 PgC of fossil fuel CO 2 is taken up by ocean Each year, ~ 11 PgC is sequestered into the deep ocean by marine organism Biological CO 2 uptake N. Gruber Air-sea CO 2 flux

Ocean carbon pumps Vertical gradient of DIC in the ocean (Volk and Hoffert, 1985) Simple 1D model (single column view) Vertical temperature gradient (solubility pump) Sinking organic material (biological pump) N. Gruber Current oceanic inventory of biologically sequestered DIC ~ 1700 PgC CaCO 3 burial is the ultimate sink of CO 2 from weathering, volcanism and human emission (10K+ years)

Evolution of the Earth’s climate for the last 60 million years Zachos et al. (2001) Science

Long-term trend and abrupt transitions Long-term CO 2 sequestration - Gradual cooling of the climate - Silicate weathering: inbalance between volcanic CO 2 emission and CaCO 3 burial Abrupt transitions - Ocean bathymetry (gateway hypothesis) - Climate threshold and feedbacks ex. ice sheet and ice-albedo feedback - Ocean circulation and biogeochemistry are involved in both hypothesis

Evolution of the Earth’s climate for the last 400,000 years Vostok ice core About 50% of glacial cooling is due to the change in CO 2 and the rest comes from planetary albedo Orbital pacing of climate

Condition during glacial maximum  pCO2 (ppmv) Terrestrial carbon loss (500 PgC) +13 Ocean cooling (-3 o C) -30 Ocean salinity increase (3%) +7 Subtotal -10 Known perturbations in the global carbon cycle The net effect is O(10 ppm) and is not enough to account for the 100 ppmv change Sigman and Boyle (2000) Nature

ProcessesIssueCitation Southern Ocean productivity SO prod; oxygen Sarmiento and Toggweiler (1984) Iron deposition on Southern Ocn. SO prod Martin (1990) Iron deposition and N 2 fixation Magnitude Falkowski (1997) Iron deposition and diatom MagnitudeBrzezinski et al (2002) Rain ratio (alkalinity increase) Lysocline Archer and Maier-Reimer (1994) Sea-ice coverage MagnitudeStephens and Keeling (2000) Deep ocean stratification MechanismToggweiler (1999) Southern Ocean upwelling MechanismFrancois et al (1997); Sigman and Boyle (2000) Hypotheses for the glacial CO 2 problem

Southern Ocean productivity Idealized box model study (Toggweiler and Sarmiento, 1984) –Atmospheric pCO 2 is proportional to the high lat surface nutrient (P H ) –50% reduction of surface P can reproduce glacial pCO 2 Large amount of unutilized surface nutrient (P) - Why phosphorus is not used up in the Southern Ocean? Surface phosphate climatology (WOA01)

Iron limitation in the Southern Ocean Iron is necessary nutrient for marine phytoplankton –Major source is atmospheric dust deposition –Southern Ocean, far from continental dust sources, is depleted in Fe SOIREE (Boyd et al. 2001) –Purposeful iron addition experiment –Increased marine productivity after iron addition –It is not yet clear its long-term impact on the oceanic carbon uptake

Antarctic ice core data Increased dust deposition during cold and arid ice age climates (Vostok ice core data) Iron hypothesis (Martin, 1990)

Climate, iron and N 2 fixation Nitrogen cycle (NO 3, NO 2, NH 4, Organic nitrogen) With NO 3 :PO 4 ratio is slightly less than 16, ocean is slightly depleted in N Some organisms can utilize N from dissolved N 2 gas (N 2 fixers) Increased Fe input to the ocean can stimulate N 2 fixation (Falkowski 1997)

Outstanding issues Ocean GCM coupled to marine ecosystem model (with explicit Fe cycle) still cannot reproduce glacial pCO 2 - Small response in atmospheric CO 2 -Regional compensation Other mechanisms? -Ocean circulation / seaice changes -Phytoplankton community structure changes and its impact on CaCO 3 burial -Total nutrient inventory (N 2 fixation) and its response to the increased iron deposition

What controls the air-sea CO 2 flux in the present climate? Takahashi et al. (2002)

What controls the air-sea CO 2 flux? Patterns of uptake and outgassing for natural CO 2 CO 2 uptake: cooling and net biological carbon sink CO 2 outgassing: heating and upwelling of regenerated DIC Thermal flux Biological flux

Simulated anthropogenic CO 2 fluxes Highlights the importance of the Southern Ocean Forward model: OCMIP-2 Orr et al. (2002) Inverse model: Mikaloff-Fletcher et al. (2006) Difference between two simulations for preindustrial and contemporary conditions

Mean CO 2 fluxes out of ocean into ocean (mol m -2 yr -1 ) Extratropical SH is a region of net CO 2 uptake Is the SH carbon flux changing? How? NaturalAnthro Lovenduski et al. (2007)

Extratropical atmospheric variability Month-to-month variability of atmospheric pressure Contraction and expansion of polar vortex Shift of westerly wind belt From NWS CPC NAMSAM

SAM impacts on SH circulation Rintoul et al. (2000) During positive phase of SAM, stronger wind over ACC increases zonal transport and upwelling in the Antarctic region Hall and Visbeck (2002)

Observed biological response Chlorophyll responds to SAM (Lovenduski and Gruber 2005) Antarctic region - Chl increases with SAM index Subantarctic region -Chl decreases with SAM index Regression of SeaWIFS chlorophyll anomaly onto SAM index ( ) Mechanisms? Impact on carbon fluxes?

Variability of SH carbon fluxes Positive-phase of SAM leads to anomalous outgassing - Atmospheric CO 2 budget: Butler et al. (2007) - GCM simulation: Lovenduski et al. (2007) Driven by increased wind-driven upwelling of deep waters enriched in DIC Regression of CO 2 flux onto SAM index

Multi-decadal trends Linear trend ( ) based on satellite observation Relatively small temperature change in the SH - Large heat capacity of the Southern Ocean Observed temperature trend : IPCC (2007) chap 3

Observed and modeled SAM trends Positive trend in SAM All of the IPCC models predict positive trend Driven by ozone depletion and global warming Stronger westerly wind over ACC G. Marshall (2003)R. Miller (2006) ObservationIPCC models

Carbon flux trends driven by SAM Positive trend in SAM leads to increased upwelling of deep waters enriched in DIC –Outgassing of natural CO 2 Atmospheric inversion ( ) (Le Quéré et al. 2007) Carbon uptake in the Southern Ocean may decline over time…

The future: the effect of Southern Ocean stratification Increased precipitation under global warming Potential melting of Antarctic ice sheet SST warming Precipitation change ( ) NCAR CCSM SRES A1B scenario Can stratification counteract wind stress changes?

Oceanic variability Explicitly resolved eddies impact on MOC structure and its sensitivity to the surface winds Hallberg and Gnanadesikan (2006)

Toward realistic Southern Ocean carbon cycle simulation Southern Ocean State Estimate (Mazloff, Heimbach and Wunsc) OCMIP / ecosystem model (Dutkiewicz et al. 2005)

Challenges for future modeling Model evaluation and improvements Process-level improvements Testing models against observational metric Statistical analysis Determine modes of carbon flux variability Attribution experiments Hierarchical modeling Repeat calculations taking out one process at a time Simple models help to interpret complex simulations