5-88 May 23 Washington, D.C. Constraining Fluxes at the Top: Advances in Quantifying Air-Sea Carbon Dioxide Fluxes during the JGOFS Decade Speaker: Rik Wanninkhof NOAA/AOML Commentator: Richard A. Feely NOAA/PMEL Outline: New techniques for determining gas transfer velocity Parameterization of pco 2 from SST, SSS, Chl and nutrients Estimation of global CO 2 air-sea fluxes F av = (k s pco 2 ) av Gas transfer velocity Function of: Surface turbulence (wind speed) Physical properties of gas and water k = (Sc) -1/2 = (n/d)( -1/2 Thermodynamic component Function of: Temperature, Salinity, TCO 2 Biology (photosynthesis/respiration) Transport (horizontal/vertical) k, 6 (cm/hr) 1 8 6 4 Tracer N-2 Tracer Wat-91 Tracer-Wann Rn C-14bomb C-14 nat k, 6 W-92 k, 6 L&M S-86 W-99 2 5 1 15 2 U 1 (m/s) Takahashi et al 22 1
CO 2 transfer velocity & exchange coefficient (K=ks): Monitoring using satellite wind speed (Geosat, SSM/I, ERS, QSCAT ) 1985-22 K over Global ocean (k-u Wanninkhof(1992) relationship) 1985 22 Geosat altimeter SSMI Microwave radiometer ERS1 and ERS2 scatterometer QSCAT scatterometer (Boutin et al., 23) Gas Transfer Velocity: Wind Speed and Mean Square Slope Dependence 3 25 25 r 2 =.81 (n = 14) 2 r 2 =.92 (n = 67) k 66 (cm hr -1 ) 2 15 1 k 66 (cm hr -1 ) 15 1 5 5 2 4 6 8 1 U 1 (m s -1 ).5.1.15.2.25 <S 2 > for 4-8 rad m -1 waves Data from 1997 NSF CoOP Coastal Air-Sea Exchange Experiment [Frew et al., 23] N. Frew / D. Glover WHOI 23 2
Jason-1 1 Altimeter Product N. Frew / D. Glover WHOI 23 CO 2 gas transfer velocity: summary summary and issues Monitoring of k using satellite wind speed (Geosat, SSM/I, ERS,, QSCAT ) 1985- present and a given k-u k U relationship: Strong k variability (including interannual) Issue: Need for intercalibration of U retrieved from various instruments K deduced from various k-u k U relationships differ (Boutin et al., GRL, 22): Issue: Calibration of k-u k U relationships still needed Use of altimeter measurements and k-mss k relationship (Glover et al., 22) k-mss estimates close to k-u k U Liss and Merlivat estimates (k-altimeter relationship calibrated with laboratory (wind/wave tank) measurements). Issue: Calibration of k-mss k relationships still needed -Boutin, Etcheto et al., JGOFS 23 3
Joint Global Ocean Flux Program Major scientific question for the JGOFS program: 1. How much carbon is sequestered by the open- oceans? Takahashi et al. CO 2 Air-sea Flux Climatology based on a 4-yr database of over 1,4, measurements of surface seawater pco 2 Global Oceanic Uptake -1.5 ±.4 Pg C yr -1 pco 2 versus Temperature in the Equatorial Pacific 89 Data Sets Collected Between March 1992 and July 21 Cosca et al., (in press) 4
Large-Scale Observational Results: 1992-22 22 El Niño:.2-.4.4 Pg C year -1 Non El Niño:.7-.9.9 Pg C year -1 La Niña:.8-1. Pg C year -1 Average:.6 ±.2 Pg C year -1 from Feely et al. JGR (submitted) Influence of Chl and SST on pco 2 observed during AESOPS and Astrolabe campaigns pco2 campaigns (nov 97 to dec 99) - Seawifs chl a (mar 98) 4S 5S PF 6S PF 13 14 15 16 17 18-17 -16 see POSTER Session 1 Theme 2, Boutin et al.: Air-sea CO2 fluxes in the Southern Ocean inferred from satellite data ) Boutin, Etcheto et al., JGOFS 23 5
CARIOCA pco 2 and CO 2 Flux deduced from QSCAT winds and K-U K U Wanninkhof (1992) relationship DP and FLUX CARIOCA 2 2 pco 2 (µatm atm) -2-4 -6-8 -1-12 -14 DP CARIOCA FLUX (U QSCAT and DP) -14 1/1 2/1 3/1 4/1 5/1 6/1 7/1 8/1 time -1-12 -2-4 -6-8 Flux (mmol m -2 day -1 ) Air sea flux from CARIOCA pco 2, atmospheric pco 2 derived from atmospheric pressure measured onboard the buoy, satellite (QSCAT) wind speed and K-U K U Wanninkhof relationship. From January to July 22: Mean P P = -12.6 µatm Mean Air-sea flux = -3.8 mmol m - 2 day -1 POSTER 1/16/23 Session 1 Theme 1, Etcheto et al.: Recent results from CARIOCA drifters in the Southern Ocean pco 2 regressions South of Tasmania and New Zealand 35 34 pco2 versus Chl in high Chl area fits in region A (chl.37 mg.m 3 ) 38 37 pco2 versus SST in low Chl area by seasons fits in region B (chl <.37 mg.m 3 ) spring december january february fall winter 36 pco 2 (µatm) 33 32 pco 2 (µatm) 35 34 31 33 3.2.4.6.8 1 1.2 1.4 1.6 1.8 chlorophyll (mg.m 3 ) 32 2 4 6 8 1 12 14 16 SST ( C) see POSTER Session 1 Theme 2, Boutin et al.: Air-sea CO 2 fluxes in the Southern Ocean inferred from satellite data ) Boutin, Etcheto et al., JGOFS 23 6
MLR Regression pco 2 versus SST, SSS, Chla,, NO3 and SiO4 in the Equatorial Pacific using 89 Data Sets Collected Between March 1992 and July 21 Cosca et al. (in press) Existing lines Planned lines Global map of existing and planned near-surface pco 2 measurements 7
Current Satellite Sensors Wind speed: (2 scatterometers in the air) -Scatterometer: QSCAT 1999-TBD Seawinds on ADEOS2 23-TBD Sea Surface Temperature: -Visible/IR radiometer: AVHRR 1982-TBD GOES -TBD Meteosat 2nd generation 22 -Microwave radiometer: TMI (4S-4N) 4N) 1997-TBD AMSR-E E on AQUA 22-TBD AMSR on ADEOS2 22-TBD Ocean Color: (6 radiometers in the air) -Visible/IR radiometer: Seawifs 1997-23 MODIS on Terra 21-25 25 MERIS on ENVISAT 22-27 27 MODIS on AQUA 22-27 POLDER 2 & GLI on ADEOS2 23-TBD Sea Surface Height anomalies: (3 altimeters) -Altimeter: Topex-Poseidon 1992 -TBD Jason 1991-TBD RA on ENVISAT 22-TBD Conclusions Remote sensing can be a powerful tool to monitor time and space variations of several parameters influencing CO 2 distribution and air-sea fluxes (wind speed, SSH, SST, Chl). Remote sensing can help interpret and extend in space and time in situ measurements Remote sensing can provide constraints for biogeochemical modelling In situ measurements are essential to: Validate remotely sensed and parameters derived from remote sensing measurements covering various oceanographic provinces at various time scales. Determine the processes contolling variations of parameters observed by remote sensing: measurements of parameters not accessible from space. 8