INTERANNUAL VARIABILITY OF PRIMARY PRODUCTION AND CARBON FLUXES

INTERANNUAL VARIABILITY OF PRIMARY PRODUCTION AND CARBON FLUXES ALONG THE U. S. EASTERN CONTINENTAL SHELF: IMPACT OF ATMOSPHERIC FORCING? Bronwyn Cahill 1, 2, Katja Fennel 3 & John Wilkin 2 1 Informus Gmb. H, Berlin, Germany 2 Institute of Marine and Coastal Science, Rutgers University, USA 3 Dept of Oceanography, Dalhousie University, Canada

COASTAL CARBON FLUXES ALONG THE U. S. EASTERN CONTINENTAL SHELF: U. S. ECOS U. S. ECo. S Team Marjorie Friedrichs (VIMS); Eileen Hofmann (ODU); Bronwyn Cahill (Rutgers University); Cathy Feng (VIMS); Kim Hyde (NOAA NMFS); Cindy Lee (Stony Brook); Antonio Mannino (NASA GSFC) Ray Najjar (Penn State); Sergio Signorini (NASA GSFC) Hanqin Tian (Auburn University); Dan Tomaso (Penn State); Yongjin Xiao (VIMS); Jianhong Xue (VIMS); Qichun Yang (Auburn University); John Wilkin (Rutgers University)

Regional differences in continental shelves’ potential to be a source or sink for atmospheric CO 2 èImportant to view regions as distinct provinces (Cai et al 2006, Borges et al, 2005). Global distribution of annual sea-air CO 2 flux measurements g. C m-2 yr-1 (Cai et al, 2006)

U. S. ECo. S Research Objectives: 1. Evaluate continental shelf carbon cycling processes including: biological processes; air sea exchange of CO 2; exchange at shelf break; exchange at land-ocean interface; burial 2. Examine sensitivity of these processes to variability in: river discharge, nutrient loadings, freshwater inflow, precipitation, ocean/air temperature, winds on n i t a io alu ilat v e sim as Land Ecosystem Model Coupled BGC/Circ Model Satellite Data ev ass alua im tion ila tio n In situ Data Coastal Carbon Fluxes Climate/ Land-Use Changes

OBJECTIVES • Inter-annual variability of primary production and air-sea CO 2 flux in three sub-regions of US east continental shelf. • Investigate sensitivity of air-sea CO 2 flux to perturbations in atmospheric forcing. • Identify the important processes responsible for producing year to year changes in air-sea CO 2 flux.

Coupled Biogeochemical Circulation Model: NENA (North. East North American shelf)

NENA Model Specifications ROMS dx ~10 km horizontal resolution; 30 layers (sigma coord); ~3. 7 min time-step Forcing Bulk formulae (Fairall et al. , 2003) applied to sea surface. NCEP NARR 3 -h fields PAR(0) = 0. 43 SWRAD; PAR(z)= f(chl(z)) BC & IC 5 -day averages HYCOM (Chassignet et al. , 2007) output along boundaries for physics; barotropic tides (Egbert & Erofeeva, 2002); NODC climatology for NO 3; TIC and ALK based on Lee et al. (2000) and Millero et al. (1998) 30 river inputs based on climatology derived from USGS freshwater gauge data and total nitrogen in nitrate pool (Howarth et al. , 1996) Biology Fasham-type (Fennel et al. , 2006; 2008; 2009) nitrogen cycle model with explicit sediment denitrification Oneway coupling Carbon model OCMIP standard for carbonate system Wanninkhof (1992) for gas exchange GOM MAB Spring Chl (mg m -3)

X X Key Biological Model Properties: • • • X X Nitrogen dynamics (Fennel et al. , 2006); Carbon dynamics (Fennel et al. , 2008) DOM dynamics (Druon et al. , 2010; J. Xue) Multiple P/Z (in development Y. Xiao) OCMIP standard for carbonate system Wanninkhof (1992) gas exchange

2 CASE STUDIES Atmospheric Forcing Time Period “Present” “Future” NCEP-NARR 3 -h fields: Added anomalies to NCEPNARR fields. Atmospheric anomalies derived from two 10 -year simulations of Reg. CM 3 model (Chen et al. , 2003) representing present and end of century (doubled) CO 2 levels, forced by 100 year transient run of NCAR climate system model TAIR, PAIR, QAIR, RAIN, SWRAD, LWRAD, UWIND, VWIND 2004 to 2007 Future scenario characterized by ~ 2 o. C air temperature increase

Higher Precipitation and SWRAD in Spring / Summer

S N alongshore decrease in wind speed 2004 2005 FUTURE - PRESENT 2006 2007

WINTER SPRING SUMMER FALL

Model evaluation with satellite data Satellite-model statistical comparisons NENA 1 satellite SST Taylor/Target diagrams evaluation (Jolliff et al. , 2008) Model vs. Satellite SST NENA 2 diffce NENA Subregions: satellite Hofmann et al. , 2011 Hofmann et al. , 2008, 2011, Druon et al. , 2010

But how do we evaluate carbon fluxes? We generally need to examine in situ data NENA annual primary productivity g. C m-2 yr-1 annual PP g. C m-2 yr-1 Model shows reasonable comparison to in situ PP data, considering variability involved in situ data NENA Present NENA Future GOM 220 (Balch et al. , 2008) 355± 36 399± 32 MAB 310 (O’Reilly et al. , 1987) 245± 21 238± 21 SAB 320 (Menzel et al. 1993) 217± 21 214± 16

AIR-SEA CO 2 FLUXES “PRESENT” SOME CHARACTERISTICS: • Generally acts as a sink “FUTURE” • • Clear alongshelf gradient Interannual variability Regional differences “Future” – shift in position of alongshore gradient Positive ocean is a sink of CO 2 Negative ocean is a source of CO 2

GOM NENA 1. 4 sea-air CO 2 fluxes 2004 to 2007 & VDK et al. , 2011, observations from 2004 to 2008 2004 2005 2006 2007 NENA NET ANNUAL FLUX: -1. 77 MOL C M-2 Y-1; VDK ET AL. , 2008 NET ANNUAL FLUX: 0. 34 MOL C M 2 Y-1 CAHILL ET AL. , IN PREP

GOM NENA 1. 4 p. CO 2 2004 to 2007 & VDK et al. , 2011, observations from 2004 to 2008 2004 2005 2006 2007 SPRING AUTUMN CAHILL ET AL. , IN PREP

GOM NENA 1. 4 & NENA 4. 1 Sea-Air CO 2 Flux & p. CO 2 2004 to 2007 SPRING PRESENT NET ANNUAL FLUX: -1. 77 MOL C M-1 Y 1 AUTUMN FUTURE NET ANNUAL FLUX: -1. 74 MOL C M-2 Y-1 CAHILL ET AL. , IN PREP

MAB NENA 1. 4 sea-air CO 2 fluxes 2004 to 2007 & Takahashi et al. , 2009 2004 2005 2006 2007 NENA NET ANNUAL FLUX: -1. 2 MOL C M-2 Y-1; TAKAHASHI ET AL. , 2009: -1. 84 MOL C M-2 Y-1 “VARIOUS” OTHER ESTIMATES: -0. 6 to -1. 7 MOL C M-2 Y-1 (Fennel et al. , 2008, Previdi et al. , 2008, De. Grandpre et al. , 2002) CAHILL ET AL. , IN PREP

MAB NENA 1. 4 p. CO 2 2004 to 2007 & Takahashi et al. , 2009 2004 2005 2006 2007 SPRING AUTUMN CAHILL ET AL. , IN PREP

MAB NENA 1. 4 & NENA 4. 1 Sea-Air CO 2 Flux & p. CO 2 2004 to 2007 SPRING PRESENT NET ANNUAL FLUX: -1. 2 MOL C M-1 Y-1 AUTUMN FUTURE NET ANNUAL FLUX: -1. 21 MOL C M-2 Y-1 CAHILL ET AL. , IN PREP

SAB NENA 1. 4 sea-air CO 2 fluxes 2004 to 2007 & Jiang et al. , 2008, observations from 2005/2006 2004 2005 2006 2007 NENA NET ANNUAL FLUX: -0. 51 MOL C M-2 Y-1; JIANG ET AL. , 2008 NET ANNUAL FLUX: -0. 48 MOL C M-2 Y-1 CAHILL ET AL. , IN PREP

SAB NENA 1. 4 p. CO 2 2004 to 2007 & Jiang et al. , 2008, observations from 2005/2006 2004 2005 2006 2007 SPRING AUTUMN CAHILL ET AL. , IN PREP

SAB NENA 1. 4 & NENA 4. 1 Sea-Air CO 2 Flux & p. CO 2 2004 to 2007 SPRING PRESENT NET ANNUAL FLUX: -0. 51 MOL C 1 M-1 Y- AUTUMN FUTURE NET ANNUAL FLUX: +0. 2 MOL C M-2 Y-1 CAHILL ET AL. , IN PREP

DISSOLVED INORGANIC CARBON (DIC) Approximate difference in annually integrated flux using a secondorder Taylor series expansion

PROCESS IDENTIFICATION USING TAYLOR SERIES DECOMPOSITION FUTURE – PRESENT ∆CO 2 FLUX ALL TERMS CO 2 FLUX Schmidt Number = f(T) Solubility = f(T, S) Winds = f(U, V) p. CO 2 = f(TA, TIC, T, S) p. CO 2 Temperature Salinity Biological Effects, NEP TIC/TA mixing pted from Previdi et al. , 2009; Colman et al. , 1997; Wetherald & Manabe, 1

∆CO 2 FLUX ALL TERMS Schmidt=f(T) Solubility=f(T, S)

∆CO 2 FLUX ALL TERMS Winds=f(U, V) p. CO 2=f(TA, TIC, T, S)

∆CO 2 FLUX ALL TERMS Net Ecosystem Production NEP=f(PP, Rem) Rate of organic carbon accumulation (mol C m-3 yr-1) NEP=f(PP, Rem)

∆CO 2 FLUX VS ∆NEP ∆CO 2 FLUX 2004 2005 ∆NEP

∆CO 2 FLUX VS ∆NEP ∆CO 2 FLUX 2006 2007 ∆NEP

CONCLUSIONS • U. S. East Continental Shelf is an overall sink of atmospheric CO 2 • Alongshelf gradient (S-N) in magnitude of flux, regional differences. • Potentially important inter-annual variability in air-sea CO 2 fluxes in all sub regions of U. S. East Continental Shelf. • Winds and p. CO 2 dominate the response of sub-regions to variability in atmospheric forcing. • Regime shifts (sink source) occur in response to “future” perturbations in atmospheric forcing. • Complex picture of sink / source regimes along US East Continental Shelf!