Mechanisms generating spatial variability in Ross Sea biogeochemical fields

Transcription

1 Mechanisms generating spatial variability in Ross Sea biogeochemical fields Matthew C. Long National Center for Atmospheric Research Climate and Global Dynamics Division 15 November 1 1

2 Global carbon cycle Atmospheric CO 2 38 CO 2 atm [ppmv] Antarctic Ice Core Mauna Loa (flask) Watt invents steam engine (1769) Year Motivation 2

3 Global carbon cycle Atmosphere Anthropogenic CO Pg C a 1 45% Natural CO 2 reservoir 6 Pg C Land biosphere 3. Pg C a 1 29% 23 Pg C 4 Atm Ocean 2.3 Pg C a 1 26% for 8 38, Pg C 6 Atm 16 Land globalcarbonproject.org Motivation 3

4 Efficiency of natural CO 2 sinks CO 2 uptake, fraction of emissions.6.4 Atmosphere Terrestrial biosphere Ocean Canadell et al. 7, PNAS Motivation 4

5 Global carbon cycle Glacial/interglacial cycles T [ C] CO 2 [ppmv] Time B.P. [kyr] Petit et al Motivation 5

6 Outline 1. Motivation Global carbon cycle: Nutrient utilization in Southern Ocean deepwater formation regions controls ocean-atmosphere partitioning of CO Introduction to the Ross Sea An important deepwater formation region, characterized by strong seasonality in surface forcing. 3. Mechanisms generating variability in biogeochemical fields Spring: Ekman restratification at ocean fronts; Summer: Ice melt, stirring, and algal community composition. Introduction 6

7 Carbon in seawater z Atmosphere Solubility pump [CO 2 ] sat = f (T, S) CO 2,g CO 2 + H 2O HCO 3 + H+ DIC z = pco atm 2 Ocean CO H + Biological pump Jex Dissolved inorganic carbon: DIC Corg DIC [CO 2 ] + [HCO 3 ] + [CO2 3 ] where [ ] denote concentrations in solution. HCO 3 bicarbonate CO 2 3 carbonate Export atm J ex = (1 A ice)kγ `pco2 pco sw 2 Introduction 7

8 Importance of the Southern Ocean Anthropogenic CO 2 in the ocean 4% entered through the Southern Ocean Column inventory (mol m 2 ) Cumulative flux up to 8 (mol m 2 ) based on tracer time distributions Khatiwala et al. 9 Introduction 8

9 Southern Ocean: meridional circulation Divergence poleward of the Polar Front brings deepwater to the surface. South SAMW: Subantarctic Mode Water AAIW: Antarctic Intermediate Water NADW: North Atlantic Deep Water CDW: Circumpolar Deep Water AABW: Antarctic Bottom Water Marinov et al. 6 North Upper cell: Subduction forms northward-flowing Antarctic Intermediate Water and Subantarctic Mode Water. Lower cell: Antarctic Bottom Water is formed along the continental margin. Introduction 9

10 CORSACS cruises Controls on Ross Sea Algal Community Structure Ross Sea 1

11 9 o E 9 o W CORSACS cruises Controls on Ross Sea Algal Community Structure 15 o E 18 o W 15 o W Depth [m] 1 o E 1 o W o E 6 o W 3 o E o 3 o W Ross Sea 1

12 Ross Sea 11

13 The most productive region in Southern Ocean Ross Sea 12

14 Annual polynya formation Top: SSM/I ice, fraction of ice free time (1995 6) Middle: SeaWiFS chlorophyll a(1997 6) Bottom: pco sw 2 climatology from all R/V N. B. Palmer data (1995 6) Ross Sea 13

15 Antarctic continental shelves Ice Shelf CO 2,g CO 2,aq Heat loss Ice export Sea ice Brine rejection Sea floor CDW AABW Ross Sea 14

16 Deepwater formation Ross Sea temperature and salinity Padman et al. 9 Ross Sea 15

17 CO 2 sink: biogeochemical modeling results Polynya formation (1996 6) Open water area [km 2 ] Oct Nov Dec Jan Feb Mar Apr As ice clears, phytoplankton bloom reduces pco sw 2 ; Sea ice limits winter outgassing; Net CO 2 sink: 13 Tg C yr 1 ; 5% of entire Southern Ocean. Arrigo et al. 8 Ross Sea 16

18 Research question What mechanisms control the spatial and temporal variability of primary productivity and biogeochemical fluxes on the Ross Sea continental shelf? Ekman restratification in spring 17

19 CORSACS stations 71 S Summer Spring Ice free time [%] S 75 S 77 S (a) 66 (b) 16 E 17 E 18 E 19 E E 16 E 17 E 18 E 19 E E Ekman restratification in spring 18

20 Spring cruise: 6 7 Satellite primary productivity and open water area: Primary Productivity [mmol C m 2 d 1 ] Open water area [1 5 km 2 ] Oct Nov Dec Jan Feb Mar Bloom is taking off, temporal rates of change are high! Long et al. 1 Ekman restratification in spring 19

21 Spring pco sw 2 decline mid-november early-december Wind [m s 1 ] pco 2 [µatm] T air [ C] SSS SST [ C] Time [d] Background: gray = ice, white = open water. Long et al. submitted Ekman restratification in spring

22 Nutrient replete stin [µmol kg 1 ] spo 4 [µmol kg 1 ] ssi(oh) 4 [µmol kg 1 ] (a) (b) (c) Summer Depth [m] 8 1 (d) (e) (f) 4 6 Spring 8 1 Ekman restratification in spring 21

23 Low iron variability (& low iron) TDFe [nm] dfe [nm] Depth [m] (a) NBP6 1 Fe Stns (c) NBP6 8 Fe Stns (b) (d) Summer Spring Sedwick et al. in review Ekman restratification in spring 22

24 Restratification can alleviate light limitation PAR PAR [ ( )] µ µ max = 1 exp Iz E k 1 5 (a) µ/µ max (b) Photoacclimation Phaeocystis Diatoms MLD [m] PAR = photosynthetically available radiation Ekman restratification in spring 23

25 Climatological hydrography Zonal salinity gradient ; Orsi & Wiederwohl 9 Ekman restratification in spring 24

26 Shelf topography interacts with mean flow Ross Sea Gyre Regional flow drives confluence of different water masses Flow is largely barotropic, thus shaped by topography Ekman restratification in spring 25

27 Submesoscale variability in biogeochemistry Ekman restratification in spring 26

28 Observations of spring bloom Cruise track and ice Zonal density gradient Sigma θ 72 S 72 S S 74 S S 76 S S 16 E 17 E 18 E 19 E E Long et al. submitted 78 S 16 E 17 E 18 E 19 E E 27.4 Ekman restratification in spring 27

29 27.86 Observations of spring bloom Cruise track and ice 72 S 74 S 76 S 78 S 16 E 17 E 18 E 19 E E Elevated oxygen concentrations found to the west of fronts. Long et al. submitted } Depth [m] (a) (b) (d) (f) Distance [km] (c) (e) Longitude [ E] O 2 [µmol kg 1 ] Ekman restratification in spring 27

30 Observations of spring bloom Buoyancy frequency Depth [m] (a) Distance [km] (c) (e) Longitude [ E] N 2 [s 2 ] Chlorophyll a Fluorescence Distance [km] Flr [arbitrary units] (b) (d) (f) Longitude [ E] Enhanced stratification and elevated biomass associated with fronts. N 2 = (g/ρ)dρ/dz Ekman restratification in spring 28

31 Observations of spring bloom Zonal density gradient Sigma θ S 27.8 Time [h] Wind speed [m s 1 ] Wind Direction [ azimuth] 74 S 27.7 Chlorophyll a [mg m 3 ] S S 78 S S mean wind direction 1 16 E 17 E 18 E 19 E E Predominantly southerly winds: tendency for Ekman transport to advect light water over dense. 76 S 78 S Fronts E 17 E 18 E 19 E E Ekman restratification in spring 29

32 Hypothesis: Ekman restratification (Northern Hemisphere example) Thomas & Ferrari 8 Ekman restratification in spring 3

33 Numerical Experiments Modified version of ROMS, using biogeochemical model of Fennel et al. [6]. Vertical mixing parameterized using KPP [Large et al. 1994]. Density structure and biogeochemical fields initialized using spring cruise data. τ y w F atm b EXP 1: 1D surface buoyancy fluxes only; EXP 2: 2D uniform lateral buoyancy gradient (3 1 8 s 2 ); and EXP 3: 2D variable lateral buoyancy gradient initialized using S i = s(z) + S 3X «2nπx x x + a n sin + φ n, L n=1 fit to salinity observations along 76 3 S. Ekman restratification in spring 31

34 Salinity evolution in 1D and 2D experiments Salinity EXP1 (1D) Depth [m] d 5 d 1 d 15 d d 23 d 1D case: mixed layer deepens continually, entrainment increases salinity and density; 16 EXP2 (2D) Depth [m] (a) D case: Ekman advection of buoyancy limits mixed layer depth; pycnostads are are formed, capping relic mixed layers (b) Ekman restratification in spring 32

35 Defining the Ekman buoyancy flux (EBF) Buoyancy equation: b ( ) t = ub + F b Integrated over depth h (2D experiments): bdz = F atm b + F ent b u b t h h x dz Ekman buoyancy flux (h > δ ek ): EBF = h u b x dz = M e h b s Ekman restratification in spring 33

36 Ekman advection provides stabilizing buoyancy flux Fb [m 2 s 3 ] 1 8 b dz [m s 2 ] (a) (b) Ekman Surface Entrainment Time [d] 2D 1D Positive surface flux indicates cooling (i.e. ). F atm b During initial high winds, EBF dominates over destabilizing F atm b. Except in strongest winds, the EBF is greater than buoyancy losses associate with entrainment at the base of the mixed layer (F ent b ). Ekman restratification in spring 34

37 Deep and shallow mixed layers in close proximity EBF [m 2 s 3 ] 1 8 Depth [m] (a) Day (a) (b) Day 1 (c) (f) (e) N 2 [s 2 ] EBF [m 2 s 3 ] 1 8 (c) Day 3 Day 15 (g) Depth [m] (b) (d) (d) (h) Longitude [ E] Ekman restratification in spring 35

38 Mixed layer variations are reflected in biogeochemical fields O 2 and phytoplankton biomass Depth [m] Oxygen [µmol kg 1 ] Phyto [% of max] Longitude [ E] pco sw 2, b/ x, and EBF pco 2 [µatm] (a) (b) Day 3 Day Longitude [ E] b [m s 2 ] <EBF> [m 2 s 3 ] 1 8 Ekman restratification enhances primary productivity; Biomass and oxygen accumulation, as well as CO 2 drawdown are enhanced where EBF <. Ekman restratification in spring 36

39 Divining the signature of the EBF pco sw 2 observations (zonal anamolies) pco 2 [µatm] 4 76 S b [m s 2 ] S 5 5 pco sw 2 is correlated with buoyancy field and the EBF on about 1 km scales; 4 77 S Phasing of relationship complicated by space-time convolution inherent in observations Longitude [ E] <EBF> [m 2 s 3 ] Ekman restratification in spring 37

40 Summer cruise: 5 6 Satellite primary productivity and open water area: Primary Productivity [mmol C m 2 d 1 ] Open water area [1 5 km 2 ] Oct Nov Dec Jan Feb Mar After bloom peak, substantial portion of seasonal NCP already occurred, rates of change are low. Long et al. accepted, JGR Summer mixed layer 38

41 Deep convection homogenizes spring water column Variance is constructed de novo each summer. Typical profiles: 1 3 sdic [µmol kg 1 ] Oxygen [µmol kg 1 ] Depth [m] Depth [m] Summer mixed layer 39

42 Rate processes decouple from geochemical tracers pcosw 2 v. NPP pcosw 2 v. chlorophyll a pco2 [µmol] pco2 [µatm] pco2 [µatm] Chl a Fluorescence 4 units] 6 [Relative Chl a Fluorescence [Relative units] Summer/Spring int(npp)dz [mmol m 2] NPP = net primary production determined by 14 C incubation Long et al. accepted, JGR Summer mixed layer 4

43 Variance propagates downscale Structure function km 42 km 8 pco sw 2 anomaly (76 3 S) VAR(pCO 2 ).6 VAR(pCO 2 ) γ(v) = 1 VAR {Z(x v) Z(x)} 2 Zonal anomaly pco 2 [µatm] Zonal anomaly pco 2 [µatm] Spring Summer 176 Longitude 18 [ E] Longitude [ E] Distance 4 [km] 6 Distance [km] Front is in the same place! Lateral gradients intensify due to time-integrated biological production; Stirring enhances short-length scale variability in summer. Summer mixed layer 41

44 Variance propagates downscale Evolution of an unstable baroclinic front Initial DIC distribution: DIC = f (T) Mahadevan et al. 4 Summer mixed layer 42

45 Locally intense pco sw 2 drawdown due to ice-melt pco 2 [µatm] As the mixed layer depth is reduced, NPP is integrated over a smaller volume. 72 S 74 S 76 S pco 2 [µatm] S 16 E 17 E 18 E 19 E Salinity Distance [km] Summer mixed layer 43

46 Shallow mixed layers export a higher percentage of C org C org v. sdic Export fraction v. mixed layer depth no export.9.8 POC [µmol kg 1 ] 4 POC [µmol kg 1 ] C org Export/NCP.7 C org Export/NCP sdic [µmol kg 1 ] sdic [µmol kg 1 ] MLD [m] MLD [m] Summer/Spring Summer mixed layer 44

47 Community composition aligns with mixed layer depth Phaeocystis Diatom % 1 72 S 8 74 S 76 S 78 S (a) 16 E 17 E 18 E 19 E E 16 E 17 E 18 E 19 E E (b) 6 4 Export ratio r C:P r C:N ǫ p [ ] Phaeocystis.5 ± ± ±.3.8 ±.1 Diatom.78 ± ± ± ±.3 Summer mixed layer 45

48 Air-sea exchange is proportional to net production Air-sea CO 2 flux v. NCP int(f CO2 )dt [mmol m 2 ] C org left in the water column int z (C org ) dz / int z (NCP) dz int 1 (NCP)dz [mol m 2 ] Jan Oct F co2 dt = f (NCP) NCP = net community production = NPP Respiration Depth [m] Conclusion 46

49 Understanding the fate of carbon on the Ross Sea shelf Observations and box model Overflow/entrainment O 2 [µmol kg 1 ] Atm sat. z < m sdic [µmol kg 1 ] z < m Atm sat. pco 2 [µatm] Atm sat. Oct Nov Dec Jan Feb Mar Apr May Arrigo et al. 8 Conclusion 47

50 Summary Ekman restratification plays an important role triggering the spring bloom, contributing % of annual production (6 7); Solar heating and ice-melt derived freshwater fluxes drive restratification later in the season; Fronts linked to bathymetry explain the location of recurring blooms; interannual variability in wind direction can modulate the magnitude of seasonal primary production; Accurately representing the processes affecting stratification, rates of NPP, and air-sea exchange is important to resolving the global ocean carbon cycle; In the future, biogeochemical fluxes are likely to change due to different physical regimes and shifts in algal community composition. Conclusion 48

51 Questions? Conclusion 49