Sea Ice Motion: Physics and Observations Ron Kwok Jet Propulsion Laboratory California Institute of Technology, Pasadena, CA

Size: px

Start display at page:

Download "Sea Ice Motion: Physics and Observations Ron Kwok Jet Propulsion Laboratory California Institute of Technology, Pasadena, CA"

Audra Rich
5 years ago
Views:

1 Sea Ice Motion: Physics and Observations Ron Kwok Jet Propulsion Laboratory California Institute of Technology, Pasadena, CA 7 th ESA Earth Observation Summer School ESRIN, Frascati, Italy 4-14 August 2014

2 1 Kwok

Buoy Drift (1979-1998): International Arctic

3 Buoy Drift ( ): International Arctic Buoy Programme Untersteiner, Colony, Rigor 2 Kwok

4 Arctic Ocean Sea Ice Some relevant facts: Arctic Ocean coverage: Max: ~8 x 10 6 km 2 Min: 4-5 x 10 6 km 2 Mean winter Ice Thickness: ~2.5-3 m Winter Snow Thickness ~10-30 cm Total Winter Volume: ~15,000 km 3 (~70% is in deformed ice, Melling and Riedel, 1995) Ice Export: ~10% of Volume and Area annually (2000 km 3 ) freshwater/heat Albedo:0.8 (snow covered ice); 0.2 (leads) Ice Salinity: 0-3 psu (old ice) 3-10 psu (first-year ice) QuikScat Ice Cover (Nov) Multiyear ice: survived one summer s melt (generally thicker; residence time < 5 yrs but decreasing) 3 Kwok

5 Sea ice motion ( ) Why does it move? How does it moves? What are the consequences of ice motion? 4 Kwok

6 Air water ice Coriolis tilt Dynamics: Force balance Coriolis and tilt rh u i t = t a i +t iw + s ij +C i + rgh H x j x i Air Water Ice: Internal ice stress Hibler s VP formulation 5 Kwok

7 Dynamics: Why sea ice moves Force Balance: Air water ice Coriolis tilt Seasonal variability of each term in a model (Steele et al., 1997) 6 Kwok

8 7 Kwok

1 circuit: 5yrs Time for ice to grow thick Transpolar Drift Stream: Time to traverse:

9 Large scale sea ice circulation - Climatology from 70-80s Ice Motion: 0-40 km/day: wind-driven at short time scales; ~1-3% of geostrophic wind Beaufort Gyre: Time to make 1 circuit: 5yrs Time for ice to grow thick Transpolar Drift Stream: Time to traverse: 2-3 yrs Significant ice area/vol exported through the Fram Strait (Why is it important?) 8 Kwok

10 Relationship between ice drift and wind: How fast does it move? 1982 u =AG + c Vector quantities u = ice motion (buoy drift) A = scaling factor G = Geostrophic wind C = ocean current Away from the coast (~400 km), more than 70% of the variance in ice motion in central Arctic can be explained by geostrophic wind at daily time scales. 9 Kwok

11 10 Kwok

$in fractures Limit of Seasonal$ Ridges Thickness distribution:

12 Sea ice thickness distribution Sample thickness distribution: ~100 km transect Extremes are Associated with Dynamics New Ice in fractures Limit of Seasonal Growth Thick multiyear ice and Ridges Thickness distribution: Variability due to Thermodynamics and Dynamics 11 Kwok

13 Why are ridges important? Ice volume/air drag 12 Kwok

14 A Framework for understanding the ice thickness distribution g(h) g t = - ( fg) h - Ñ (ug ) + Y 13 Kwok

15 Ice Drift (Different length scales) 2000 km 100 km L ICE is a SOLID! All large scale gradients are concentrated in cracks and fractures 14 Kwok

16 Why are cracks important? Strong discontinuities in the displacement field Heat flux and ice growth Seasonal variation in sensible heat flux as a function of ice thickness (m) (after Maykut, 1978) 15 Kwok

17 Fractures in the ice (An extreme year 2013) 500 km 16 Kwok

18 Ice Deformation from Synthetic Aperture Radar 5 km grid Div - ice prod Conv- ridging Ice production 17 Kwok

19 18 Kwok

20 Observational Basis X(t i+1 ) Displacement: Average velocity: z X(t i ) 19 Kwok

21 Retrieval of ice displacements in satellite imagery Images I 1 and I 2 separated by Δt 20 Kwok

22 Uncertainties in motion estimates Uncertainties in displacement: g: geolocation errors f: tracking errors Uncertainties in spatial differences (strain): Δ f Assuming geolocation errors (g) are correlated. 21 Kwok

23 Ice Motion Tracking Block Diagram Estimation and image selection Motion tracking 22 Kwok

24 Observations of ice motion Buoy drift/trajectories (since mid-to-late 70s from the Arctic buoy program) Argos location (Uncertainty: ~300 m) GPS (uncertainty: ~10 1 m) Density: typically ~10 2 km, hourly samples Satellite fields (tracking features in sequence of images) Passive microwave (uncertainty: km) Routine retrievals since late 90s Synthetic Aperture Radar data (uncertainty: 10s of meters) Routine retrievals since early 2000 Time sampling: hours to several days 23 Kwok

25 24 Kwok

26 Large scale ice drift (from Passive microwave rad, QuikSCAT, ASCAT) IPCC WG1-Ch 4 FAQ 25 Kwok

27 Daily ice motion: Dec 3,07 - Feb 15,08 From passive microwave fields (~12.5 km resolution) Alaska Fram Strait Greenland 26 Kwok

Fram Strait Area Outflow Annual and Winter (Oct-May): 1979-2007 Peak (near peak of +NAO index) 10% of

28 Fram Strait Area Outflow Annual and Winter (Oct-May): Peak (near peak of +NAO index) 10% of Arctic Ocean area years Kwok and Rothrock [1999] and Kwok [2009] Variability is high! 27 Kwok

29 What about volume flux (freshwater/heat)? Volume flux (km 3 ) 0.5 m Avg: 2200 km m ~0.07 Sv Based on ice draft from NPI and AWI moorings 28 Kwok

30 Source regions of sea ice by backpropagation Area swept by the trajectories is highly correlated to the area flux 29 Kwok

$fractures in the$ ice cover

31 High-res Satellite mapping of timevarying fractures in the ice cover (Radarsat-1, Envisat, Sentinel-1) SAR imagery 25 m resolution 10 km Grid resolution: 10 by 10 km 10 1 km 10 2 km 10 3 km 30 Kwok

32 Time-varying deformation - Beaufort Sea (Divergence) 31 Kwok

33 Shear patterns density, orientation, persistence Shear 32 Kwok

34 Deformation: Contrast between seasonal and multiyear ice regions Shear Deformation Deformation activity Seasonal ice Multiyear ice (Kwok, 2005) 33 Kwok

such as rafting and pressure ridging Current ridging schemes used in coupled

35 Ridging/rafting (mechanical redistribution) Much less effort has focused on how ice is redistributed among thickness categories by mechanical processes such as rafting and pressure ridging Current ridging schemes used in coupled ice/ocean models are largely heuristic and are difficult to verify empirically 34 Kwok

36 Model vs RGPS ice production Nov00-Apr01 RGPS (Derived from Obs) Models Model simulations produce less ice because deformation is poorly simulated comparison is over limited domain 35 Kwok

37 Upwelling of Arctic Pycnocline associated with shear motion of ice (SHEBA) Kinematic stress curl From satellite ice drift McPhee et al Kwok

38 37 Kwok

39 Clockwise Rotation of ice motion vectors over 100 minute intervals over a region of relatively thick ice 2002 Ice Motion G. Wind DOY 2003 Ice Motion G. Wind DOY Blue-green-orange-red show progression in time Kwok, Cunningham and Hibler (2004) 38 Kwok

40 Velocity Gradients and Divergence 39 Kwok

41 40 Kwok

42 Satellite Ice Drift and MY ice coverage MY ice coverage trend MY ice coverage trend Trend in drift speed 41 Kwok

43 Vector Trend Kwok

44 CMIP3 sea ice motion (Kwok, 2011) 43 Kwok

45 Summary Remarks Brief highlights of the role of ice motion in the shaping the character of the Arctic Ocean sea ice cover Ice motion is complex and plays a critical role in the time-varying behavior of the ice cover Controls the extremes in the ice thickness distribution Very challenging observationally because of space and time scales of variability In most models, small-scale ice motion seems to modeled poorly Mechanics and sub-grid scale processes are not represented correctly Need better ice treatment of ice behavior Lack observations of short time scale processes 44 Kwok

46 45 Kwok

Spectral Albedos. a: dry snow. b: wet new snow. c: melting old snow. a: cold MY ice. b: melting MY ice. d: frozen pond. c: melting FY white ice

Spectral Albedos. a: dry snow. b: wet new snow. c: melting old snow. a: cold MY ice. b: melting MY ice. d: frozen pond. c: melting FY white ice Spectral Albedos a: dry snow b: wet new snow a: cold MY ice c: melting old snow b: melting MY ice d: frozen pond c: melting FY white ice d: melting FY blue ice e: early MY pond e: ageing ponds Extinction