Sea Ice. Martin Vancoppenolle" Laboratoire d Océanographie et du Climat (LOCEAN)" CNRS, Paris, France"

Transcription

1 Sea Ice and Biological Productivity of the Polar Oceans Martin Vancoppenolle" Laboratoire d Océanographie et du Climat (LOCEAN)" CNRS, Paris, France" " with contributions from: K. Meiners, S. Moreau, L. Bopp, G. Madec Bloom in the Barents Sea, Aug 31, 2010, earthobservatory.nasa.gov

2 Intro " Pelagic productivity in ice-covered oceans " Productivity in sea ice " Outlook

3 why should we care about sea ice and marine biogeochemistry?

4 Sea ice is rapidly changing Polar oceans are biologically productive Ocean CO Large uncertainties in the marine bgc processes in the frozen oceans The sign of the future PP change is debated The sign of the present-day CO uncertain in the frozen oceans Better understanding of BGC processes in the sea ice zone is required

5 How does sea ice may interact with marine BGC? Light transmission Ocean vertical mixing and upwellings Active processes within the sea ice Uptake and release of material Feedbacks

6 How does sea ice may affect marine BGC? Light transmission Ocean vertical mixing and upwellings Active processes within the sea ice Uptake and release of material Feedbacks

7 How does sea ice may affect marine BGC? Light transmission Ocean vertical mixing and upwellings Active processes within the sea ice Uptake and release of material Feedbacks

8 Are there any active BGC processes in sea ice? Vancoppenolle et al., 2013

9 How are sea ice processes observed? In situ sampling of sea ice and brines (destructive) Spatial variability New methods, no or little intercalibration Complicated access, expensive logistics 30% of cores have less than 1mg chl- a/m2 ROVs are starting Satellites hardly see anything Miller et al., 2014 in prep

10 How are sea ice processes observed? In situ sampling of sea ice and brines (destructive) Spatial variability New methods, no or little intercalibration Miller et al., 2014 in prep Complicated access, expensive logistics ROVs are starting Analysis Satellites hardly see anything

11 How are sea ice processes observed? In situ sampling of sea ice and brines (destructive) New methods, no or little intercalibration Spatial variability Complicated access, expensive logistics ROVs are starting Satellites hardly see anything Jan Lieser, ACE-CRC, personal communication

12 Primary production in the sea ice zone Open ocean" " Ice edge" " Under the ice" " In the ice

13 Primary production estimates in the sea ice zone Tg C per year Arctic Antarctic Total Open ocean Marginal ice zones Sea ice Under ice?? Sources: sea ice stock estimates from extrapolation of obs ice core chl-a [Legendre et al., 1992] Arctic sea ice model, bottom community only [Deal et al., in press] Antarctic sea ice model, surface community only [Arrigo et al., 1997] marginal ice zones and open ocean Arctic: in situ & ocean color [Hill et al., 2013] Antarctic: satellite data of ocean color [Pabi et al., 2008; Arrigo et al., 2008] Locally, ice algae account for up to 50% of production [Gosselin et al., 1997]

14 In ice-covered oceans

15 Ice edge blooms not fully understood peak within 20 days of ice retreat [Perrette et al., 2011] account for most of Arctic PP [Pabi et al., 2008] Ice edge blooms are promoted by Alexander and Niebauer, 1981 stratification ice-edge upwelling release of light limitation seeding from the ice Perrette et al., 2011

16 Under-ice blooms Snow and ice optical properties seem to control under-ice blooms [Fortier et al., 2002] Under ice blooms can (always?) precede ice-edge blooms Could become more prevalent in the future because of more melt ponds 31 g C / m 2 [Mundy et al., 2009] g C / m2 [Arrigo et al., 2012] into single bloom events Arrigo et al., 2012

17 CMIP5 Earth System Models 1900 historical 2005 RCP

18 An example of modeling, NEMO-PISCES No radiative transfer through the sea ice Material exchange through open water only No exchange at the ice-ocean interface No active process in the sea ice Aumont and Bopp, 2003

19

20 Sea ice and marine productivity in PP [gc m -2 y -1 ] σ = 40% x μ Earth System Models 487 Tg C/yr Sea Ice σ = 17% x μ NO3 [mmol m -3 ] σ = 55% x μ where models agree with each other Vancoppenolle et al., 2013

21 Projections of future marine productivity do not agree on the sign of future marine PP Vancoppenolle et al., 2013

22 Sea ice and NO3 in models Sep SIE [10 6 km 2 ] Sattelite (OSI-SAF) Surface NO3 [mmol m -3 ] WOA consistent mechanisms, spread NO3, timing of oloigotrophy Vancoppenolle et al., 2013

23 Increases in PP further enhance sea ice melting BIO-PHYSICAL FEEDBACK [Lengaigne et al., 2009]

24 In sea ice

25 Brine inclusions The crystalline network of H2O hardly tolerates salt => pure ice + saline brine" " Porosity changes with temperature" " Critical permeability transition around 5% porosity" " Fluid transport: percolation and convection Micro-photography of artificial saline ice (Light et al., 2003) Computed using X-ray sea ice tomography (Pringle et al., 2009; Golden et al., 2007)

26 Brine inclusions Brine salinity Brine volume Permeability

27 Brine dynamics Permeability, vertical brine density profile and meltwater production drive brine dynamics" " All feature seasonal variations mostly driven by temperature" " Convection (gravity drainage) and meltwater percolation (flushing) govern desalination T σ melt φ = 5 % T σ melt φ > 5 % CONVECTION (WINTER) STABLE BRINES / PERCOLATION (SUMMER)

28 Primary producers in sea ice Low-light adapted, specific species (diatom, flagellates, ) Live in brine (stable, ventilated, but cold and saline) Excrete EPS to protect themselves [Underwood et al., 2013] Primary production in sea ice is hardly measurable Biomass databases are being developed Large-scale estimates of PP are derived from models C. Krembs

29 Global sea ice databases BEPSII SCOR WG (Stefels et al.) attempts notably to build global sea ice BGC databases First project on Antarctic sea ice cores, but being extended in the Arctic now Chl-a is a widely measured proxy for biomass 1300 cores Methods Ice core extraction Cutting core into sections Melting in at < 5 C in the dark Filtration Fluorometric analysis Get Data Takes time, but it s being done for 30 years

30 Core samples

31 Seasonality FALL SPRING 1 Light 2 Nutrient supply mechanisms, e.g., brine dynamics, vary in time 3 Physical conditions (T and S) are not always optimal for ice algal growth 4 Snow 5 Water column solid = mean dots = std

32 1 Light PAR snow ice Mean seasonal cycle of incoming PAR in the Southern Ocean sea ice zone Incoming PAR computed as a function of latitude, day of year, cloud fraction, humidity [Shine, 1984; Vancoppenolle et al., 2011] Averaged over the entire sea ice zone (using SSMI data) Incoming light takes off in September and shuts off in May Small latitude variations

33 2 Brine dynamics and nutrient supply T σ CONVECTION (WINTER) φ = 5 % T σ φ > 5 % STABLE BRINES / PERCOLATION (SUMMER) If unstable brine gradient in sea ice, nutrient fluxes are possible. Convection starts below C [Jardon et al., in revision] If no unstable gradient in sea ice, no nutrient fluxes are possible, probable limitation by one of the nutrients

34 3 Temperature and brine salinity Photosynthetic efficiency decreases at low temperatures and high salinities. Brine salinity increases fast with temperature and this effect outcompetes temperature. At - 5 C, growth is 10 x smaller than at - 2 C. - 5 C Normalized ice algal growth in mesocosm experiments as a function of solution salinity (Arrigo and Sullivan, 1992)

35 Temperature constraints brine convection possible brine salinity stress Mean seasonal cycle of 2m air temperature based on NCEP- NCAR data ( ) Air temperature provides two constraints on algal growth Air temperature frequently drops below - 3 C from February to November - > nutrient supply by brine convection possible Air temperature does not go above - 5 from April to October - > brine salinity stress on ice algae

36 Chl-a from Antarctic sea ice cores: seasonality brine convection possible brine salinity stress polar night 1300 cores, Consistent chl-a methods Light and air temperature control ice algae Same work is underway for the Arctic Chl-a rich ridged-ice is undersampled This is all that we have Meiners et al., 2013

37 Primary production in Arctic sea ice ice algae phyto Bottom community model ice + phyto ice/phyto Simulations by Dupont et al., 2012

38 Present ice models cannot simulate Antarctic productivity Bottom community model 27% 29% h skel 16% 31% 56% 39% Multi-layer model Δh profile- type classification contribution to integrated chl- a

39 Dependence on ice thickness normalized depth in the ice RIDGING limit of thermodynamic growth 1 2

40 Outlook

41 Bgc processes in the sea ice zone are only being unravelled. We only mentioned aspects regarding PP but there are many other aspects (CO2, iron, DMS, see review paper Vancoppenolle et al., QSR 2013) Large model uncertainties in future PP projections Pelagic big nutrient problem in the Arctic bloom dynamics in the sea ice zone are not perfectly understood (radiation, ocean circulation, light adaptation) under-ice blooms in some models / not in the others potential bio-physical feedback Sea ice Sea ice direct contribution to PP seems small, but large observational errors (spatial variability, no ridged ice) primary production is not well observed Sea ice databases; new observation methods