IN THE INDIAN OCEAN? Marta Álvarez

Transcription

1 IS THERE A HIGHER C ANT STORAGE IN THE INDIAN OCEAN? Marta Álvarez C. Lo Monaco, T. Tanhua, A. Yool, A. Oschlies, J.L. Bullister, C. Goyet, F. Touratier, E. McDonagh and H.L. Bryden. IMEDEA, CSIC UIB, Mallorca, Spain Gijón, May /18

2 C ANT inventory referred to 1994 = 110 ± 13 Pg C Indian Ocean contains - 21% of the total C ANT inventory - half of the Atlantic C ANT inventory despite having only 20% less area Indian Ocean 20.3 ± 3 Pg Pacific Ocean 44.5 ± 5 Pg Atlantic Ocean 44.8 ± 6 Pg 2/18 (Sabine et al, Science 2004)

3 Indian Ocean C ANT vertical distribution μmol/kg RedSea /Persian Gulf Intermediate Water Antarctic Intermediate Water (AAIW) (Sabine et al, Science 2004) AAIW marks the deepest C ANT penetration in the Subtropical gyre No C ANT in deep and bottom waters 3/18

4 Hypothesis: C ANT penetrates deeper than m How to assess this question: compare different C ANT methods -> difficult: every method has high uncertainties help: relation of C ANT with tracers (CFC 12, CCl 4 ) 4/18

5 On board R/V Charles Darwin, CD139 cruise, 1/3 15/4/2002, from Durban to Freemantle P.I.: Harry Bryden (National Oceanographic Centre, Southampton, UK) Physics: temperature, salinity, ADCP, LADCP Chemistry: oxygen, salinity, nutrients, CO 2 (ph & TA, TIC), CFCs (11, 12, 113), CCl 4. 5/18

6 Africa Salinity (psu) Australia SAMW SAMW AAIW NADW CDW AABW 6/18

7 Africa CFC-12 (pmol/kg) Australia SAMW NADW AAIW CDW CDW AABW 7/18

8 C ANT TECHNIQUES 1. C ANT SAB99: method by Sabine et al. (GBC,1999), using their ΔC Dis 2. C ANT LM05: method by LoMonaco et al. (JGR, 2005) 8/18

9 2. Back-calculation technique by Lo Monaco et al. (JGR, 2005): OMP Africa NADW contribution Australia Ice-covered SW contribution 9/18

10 C ANT TECHNIQUES 1. C ANT SAB99: method by Sabine et al. (GBC,1999), using their ΔC Dis 2. C ANT LM05: method by LoMonaco et al. (JGR, 2005) 3. C ANT TrOCA: Touratier & Goyet (Tellus, 2007) 4. C ANT TTD 5. C ANT from OCCAM model 10/18

11 C ANT vertical distributions OCCAM LM05 SAB 99 TTD TrOCA 11/18

12 Specific C ANT inventories Africa Australia 12/18

13 Neutral density layers & salinity Africa Surface SAMW Australia AAIW Deep Bottom 13/18

14 Specific C ANT inventories mol C / m LM05 yields very high C ANT estimates C-methods Deeper C underestimate Cbut not due to AABW ANT penetration below TTD & ANT AAIW Troca & OCCAM give compared to TTD and OCCAM Significant C ANT values similar in deep results, and bottom higher than waters SAB99 SAB99 TTD & Troca give similar results, higher than SAB99 LM05 Troca TTD OCCAM 2 0 Surface SAMW AAIW Deep Bottom 14/18

15 Surface, SAMW waters and upper AAIW 80 TrOca LM 60 SABstr TTD Model Revelle = C ANT umol/kg CANT umol/kg Revelle = pf12 ppt pcfc-12 15/18

16 Deep and bottom waters C ANT CANT umol/kg pf12 ppt 40 pcfc-12 TrOca C CANT ANT umol/kg LM SABstr TTD Model 0 16/ pccl pccl4 4 ppt

17 Summary table Reliability code (subjective) 1- plus 2- medium 3- low Upper SAMW AAIW Deep Bottom Mean low SAB LM medium- 2- TrOCA medium+ 2 TTD medium+ OCCAM medium 17/18

18 Hypothesis: C ANT penetrates deeper than 1000 m (Sabine et al 1999) seems to be hinted by any other method How to assess this question: difficult: every method has uncertainties absolutely true, no method is perfect help: relation of C ANT with tracers (CFC 12, CCl 4 ) questions still arise, saturation, mixing, etc.. community should combine methods and time-evolution studies at specially sensitive regions OPEN DISCUSSION: KEY REGION.. THE SO 18/18

19 19/18

20 Back-calculation technique by Sabine et al. (GBC, 1999) to estimate C ANT. C ANT = ΔC* - ΔC Dis = C AOU/R C ½(ΔTA + AOU/R N ) + 106/104 N* - C 280 ΔC Dis C is the current total inorganic carbon Δ TA = TA - TA 0, current alkalinity - preformed alkalinity TA 0 = Sal PO Tpot AOU is the Apparent Oxygen Utilization, assuming oxygen saturation C 280 is the inorganic carbon in equilibrium with the preindustrial atmosphere. C 280 = f (Tpot, Sal, TA 0, pco 2280 ) from thermodynamic equations pco 2280 = CO 2 fugacity at a 100% of water vapor pressure in uatm = f(tpot, Sal, 280) C 280 in GSS96 => constants by Goyet & Poisson (1989) & a constant pco 2280 = 280 uatm linearilized equation: C 280 = (Tpot - 9) (Sal - 35) (TA ) N* = (0.87 (NO 3-16 PO )) term accounting for the denitrification 20/18

21 ΔC Dis is obtained with own CD139 data using CFC 12 ages and limit at 40 years a) Old deep waters, C ANT = 0 => ΔC* = ΔC Dis b) In upper waters, having the age: ΔC Dis = ΔC*t σ=cte ΔC* t = C - C Bio -C t where C t = f (Tpot, Sal, TA 0, pco 2t ) pco 2 in the atmosphere at 2002 age (t) c) In between => weigthed mean ΔC* and ΔC*t σ=cte ΔC Dis is obtained with own CD139 data using CFC 12 and CCl 4 ages a) Tpot<3ºC, (assumed 100% saturation) deep waters, ΔC Dis = ΔC*t CCl4 σ=cte ΔC* t = C - C Bio -C t where C t = f (Tpot, Sal, TA 0, pco 2tCCl4 ) pco 2 in the atmosphere at 2002 age CCl4(t) b) In upper waters, having the CFC age: ΔC Dis = ΔC*t CFC12 σ=cte ΔC* t = C - C Bio -C t where C t = f (Tpot, Sal, TA 0, pco 2t ) pco 2 in the atmosphere at 2002 age CFC12 (t) c) In between => weigthed mean a & b 21/18

22 Back-calculation technique by Lo Monaco et al. (JGR, 2005): C ANT = C T C Bio -C T 0 obs (C T C Bio C 0 obs ) REF C T = measured TIC C Bio = 0.73 (O 20 O 2 ) (TA TA 0 ) => biological activity variation in TIC TA 0 => preformed TA O 20 = O 2 sat α Κ O 2 Sat => αo 2 = 12% undersaturation K = > mixing ratio of ice-covered water (OMP) C 0 obs => preformed TIC currently observed in the formation area water masses REF = reference water where no C ANT should be detected. 22/18

23 Back-calculation technique by Lo Monaco et al. (JGR, 2005) TA 0 = k(s) TA 0 (S) + k(n) TA 0 (NADW) C 0,obs = k(s) C 0,obs (S) + k(n) C 0,obs (NADW) southern relationships: winter surface data from the Atlantic and Indian oceans (WOCE and OISO cruises) TA 0 (S) = PO S θ (± 5.5 µmol/kg, r 2 = 0.96, n = 243) C 0,obs (S) = PO S θ (± 6.3 µmol/kg, r 2 = 0.99, n = 428) northern relationships subsurface data from the North Atlantic and Nordic Seas (WOCE and KNORR cruises) TA 0 (N) = S θ (± 9.3 µmol/kg, r 2 = 0.92, n = 297) C 0,obs (N) = S NO (± 9.2 µmol/kg, r 2 = 0.79, n = 364) mixing ratios of southern and northern waters: k(s) + k(nadw) = 1 determined from OMP analysis 23/18

24 Back-calculation technique by Lo Monaco et al. (JGR, 2005): OMP analysis modified from Lo Monaco et al. (JGR, 2005) Endmembers: AAIW, NADW-E, NIDW, Indian Water, WDS/CDW, ISW Weddell Variables: 6 Sal, Tpot, 4 SiO2 and NO IW Tpot 2 0 AAIW2 NIDW WDW/CDW NADW-E -2 ISW Wed Salinity Tpot AAIW2 NADW-E IW NIDW WDW/CDW Tpot -2 ISW Wed SiO2 4 IW AAIW2 2 NADW-E NIDW 0 WDW/CDW -2 ISW Wed NO 24/18

25 Back-calculation technique by Lo Monaco et al. (JGR, 2005): OMP Tpot, Sal, SiO 2 NO STD Res R (n=1299) Tpot Sal SiO2 NO data Med-Calc weighted & adim Tpot Sal Med-Calc Real SiO2 NO Med-Calc Real 25/18

26 Back-calculation technique by Lo Monaco et al. (JGR, 2005): C ANT = C T C Bio -C T 0 obs (C T C Bio C 0 obs ) REF REF = NADW= reference water where no C ANT should be detected. Applied to samples with more 50% NADW from OMP analysis (C T C Bio C 0 obs ) REF = ± 1.4 umol/kg C ANT = C T C Bio -C T 0 obs (- 54.4) (LoMonaco JGR umol/kg) 26/18

27 C ANT TrOCA, Touratier et al (Tellus B, 2007): TrOCA = O 2 + a (C T -1/2 TA) a = ψ O2 / [ψ CO2 + ½ ( ψ H+ ψ HPO24 )] C ANT (TrOCA) = (TrOCA TrOCA 280 ) / a C ANT = (O (C T 0.5 TA) exp ( θ /TA 2 ))/1.279) 27/18

28 4. C ANT TTD (Waugh et al., 2004; 2006; Tanhua et al., 2008): - each water sample has its own age, i.e. time since it was last in contact with the atmosphere. The sum of all these ages makes the TTD of a water sample - the mean age (Γ ) and the width of the TTD (Δ) are assumed to be of equal magnitude: realistic assumption of the relation between advective and diffusive transport in the Ocean -C ANT is an inert passive tracer where air-sea disequilibrium hasn t changed over time. 28/18

29 5. C ANT OCCAM - global, medium-resolution, primitive equation ocean general circulation model (Marsh et al., 2005). - OCCAM's vertical resolution is 66 levels (5 m thickness at the surface, 200 m at depth), with a horizontal resolution of typically 1 degree. - Advection is 4th order accurate, and the model is time-integrated using a forward leapfrog scheme with a 1 hour time-step. - Surface fluxes of heat and freshwater not specified but are calculated empirically using NCEP-derived atmospheric boundary quantities (Large and Yeager, 2004). - OCCAM incorporates a NPZD plankton ecosystem (Oschlies, 2001; Yool et al., 2007) which drives the biogeochemical cycles of nitrogen, carbon, oxygen and alkalinity. 29/18