Towards a Comprehensive Model of the Marginal Ice Zone · 2011-03-11 · The objectives are to...
1
Figure 2. Map of the Fram Strait area showing the simulated (blue line) and observed (red line) ice edges, the zonal transect at 79 o N used for the experiment (black line), the model domain (dotted line) and the area where wave data are extracted from the ECMWF re-analysis (gray box). Figure 3. Left: Daily peak period T m and significant wave height H s taken from the ECMWF re-analysis and used to produce an incident wave spectrum. Center: Predicted MIZ extent according to floe size (pale green) and ice concentration (dark green). The gray line shows the observed position of ice edge (AMSR-E). Right: Minimum and maximum floe sizes allowed in the MIZ. Towards a Comprehensive Model of the Marginal Ice Zone Dany Dumont a* , Alison Kohout b and Laurent Bertino a a Nansen Environmental and Remote Sensing Centre / Mohn-Sverdrup Center,Thormøhlensgt. 47, N-5006 Bergen, Norway b National Institute for Water and Atmospheric Research, Christchurch, New Zealand Background and Objectives Swell generated in the open ocean propagates remarkable distances into ice covered seas and can significantly affect sea ice dynamics by breaking large sheets in many smaller floes. Waves-in-ice and information about floe size are absent from today’s sea ice–ocean models. However, understanding and predicting wave-ice interactions is very important for the safety of marine operations in the short-term and for better climate modeling at longer time scales. The objectives are to layout the architecture of a comprehensive model of the marginal ice zone (MIZ) that includes waves, and to demonstrate the concept using a one-dimensional model. Representing dynamical processes associated with the MIZ in a sea ice-ocean coupled model requires: a specific rheology for the MIZ; a representation of waves-in-ice; and a parameterization of ice breaking due to waves. Each of these are briefly described and preliminary results are presented. Figure 1. Flowchart diagram of the integrated MIZ model to be developed and validated during the Waves-in-Ice Forecasting for Arctic Operators Project (WIFAR 2010-2013). A σ = 2πμσ fc h 2 3gρ w λ 2 ∝ h 2 T 4 Waves-ice-ice Model The wave scattering model of Kohout and Meylan (2008) is used to compute wave attenuation in the MIZ. For more details, see poster of Kohout et al. (#0021). Floe Breaking Parameterization Acknowledgments This work is part of the MIZ project funded by Total E&P and of the Waves-in- ice Forecasting for Arctic Operators (WIFAR) project collaboratively funded through the Research Council of Norway (PETROMAKS) and Total E&P. References Kohout, Dumont and Bertino (2010) Modelling Ocean Waves in a Marginal ice Zone. Poster #0021, IPY Oslo Conference. Kohout and Meylan (2008) A model for wave scattering in the Marginal Ice Zone based on a two-dimensional floating elastic plate solution. J. Geophys. Res., 113: C09016. Langhorne, Squire, Fox and Haskell (1998) Break-up of sea ice by ocean waves. Annals Glaciol., 27: 438-442. Shen, Hibler and Leppäranta (1987) The role of floe collisions in sea ice rheology. J. Geophys. Res., 92: 7085-7096. Timco and O'Brien (1994) Flexural strength equation for sea ice. Cold Reg. Sci. Tech., 22: 285-298. Toyota, Haas and Tamura. (In press) Size distribution and shape properties of relatively small sea ice floes in the Antarctic marginal ice zones in late winter. Deep Sea Res. II. Atmospheric Forcing Sea Ice Model if D < D c MIZ rheology else EVP rheology end Waves-in-ice Model Floe Breaking Parameterization Ocean Model waves-in-ice spectrum ocean wave spectrum air-sea fluxes ice conditions floe size distribution The TOPAZ Fram Strait Sea Ice-Ocean Model Proof of concept using a one-dimensional numerical implementation (T m , H s ) L MIZ D max < D c marginal ice zone D max > D c ice pack open ocean wave spectrum An ice plate of constant thickness will fail whenever: 1) the flexural stress reaches the flexural strength σ fc or; 2) the flexural strain reaches a maximum value ε fc . A critical wave amplitude is derived for both failure modes: If A ≥ min(A σ ,A ε ) the plate breaks up in many floes with a maximum size D max = λ/2. Fatigue is parameterized as σ fc μσ fc with μ < 1 following Langhorne et al. (1998). Flexural strength is taken from Timco et al. (1994). We present here a one-dimensional implementation of the framework described above as a way of testing components, parameterizations, and numerical schemes before its implementation in the HYCOM TOPAZ system. The goal is to compute the width of the MIZ in Fram Strait based floe size and properties of the floe size distribution. ice concentration viscous-plastic MIZ ice viscosity (kg s -1 ) Comparison between EVP and MIZ rheologies * Contact author: [email protected] Main Results and Conclusion • The model produces reasonable values for the width of the MIZ given the numerous assumptions (deep water wave dispersion, one-point wave source, mean direction instead of directional spectrum, etc.) • Ice concentration does not capture the location of the MIZ. • Maximum floe size < 100m in qualitative agreement with observations (e.g. Toyota et al. 2010). • Preliminary results in one dimension show the relevance of the architecture. Future Work • Implement in two dimensions in HYCOM. • Validate in Fram Strait (WIFAR cruise Sep. 2010). • Refine waves-in-ice models and parameterizations. Parameter Symbol Value Flexural strength σ fc 0.6 MPa Maximum flexural strain με fc 3.5 × 10 -5 Fa2gue μ 0.6 – 1.0 Sea water density ρ w 1025 kg m -3 Gravita2onal accelera2on g 9.81 m s -2 Cri2cal floe size D c 500 m Ice Young’s Modulus E 0.5 MPa Poisson’s ra2o ν 0.3 A ε = με fc λ 2 4πh ∝ T 4 h H s (m) Photo: Courtesy of Prof.V. Squire T m (s) c < 0.98 D < 500m model obs 2007 2008 2009 D min = π 2 Eh 3 3gρ w 1− ν 2 ( ) 14 ST ( ) = 1.25 8π T 5 H s 2 T m 4 e −1.25 TTm ( ) 4 A nested configuration of HYCOM has been set-up in Fram Strait at a 3.5-km resolution. Boundary conditions are provided by TOPAZ system which is the Arctic monitoring and forecasting system of MyOcean. Sea ice dynamics is modeled using a dual rheology composed of the elastic-viscous-plastic (EVP) rheology and a novel rheology for the MIZ developed at the NERSC based on the ideas of Shen et al. (1987). This formulation needs a criterion that will determine the dynamical regime. The current version uses an ice concentration threshold (mainly for testing purposes), but the objective is to include waves and use a criterion based on floe size. Greenland Ice edge D min D max
Transcript of Towards a Comprehensive Model of the Marginal Ice Zone · 2011-03-11 · The objectives are to...
Figure 2. Map of the Fram Strait area showing the simulated (blue line) and observed (red line) ice edges, the zonal transect at 79oN used for the experiment (black line), the model domain (dotted line) and the area where wave data are extracted from the ECMWF re-analysis (gray box).
Figure 3. Left: Daily peak period Tm and significant wave height Hs taken from the ECMWF re-analysis and used to produce an incident wave spectrum. Center: Predicted MIZ extent according to floe size (pale green) and ice concentration (dark green). The gray line shows the observed position of ice edge (AMSR-E). Right: Minimum and maximum floe sizes allowed in the MIZ.
Towards a Comprehensive Model of the Marginal Ice Zone Dany Dumonta*, Alison Kohoutb and Laurent Bertinoa
a Nansen Environmental and Remote Sensing Centre / Mohn-Sverdrup Center, Thormøhlensgt. 47, N-5006 Bergen, Norway b National Institute for Water and Atmospheric Research, Christchurch, New Zealand
Background and Objectives
Swell generated in the open ocean propagates remarkable distances into ice covered seas and can significantly affect sea ice dynamics by breaking large sheets in many smaller floes. Waves-in-ice and information about floe size are absent from today’s sea ice–ocean models. However, understanding and predicting wave-ice interactions is very important for the safety of marine operations in the short-term and for better climate modeling at longer time scales. The objectives are to layout the architecture of a comprehensive model of the marginal ice zone (MIZ) that includes waves, and to demonstrate the concept using a one-dimensional model. Representing dynamical processes associated with the MIZ in a sea ice-ocean coupled model requires: a specific rheology for the MIZ; a representation of waves-in-ice; and a parameterization of ice breaking due to waves. Each of these are briefly described and preliminary results are presented.
Figure 1. Flowchart diagram of the integrated MIZ model to be developed and validated during the Waves-in-Ice Forecasting for Arctic Operators Project (WIFAR 2010-2013).
€
Aσ =2πµσ fch
2
3gρwλ2 ∝
h2
T 4
Waves-ice-ice Model
The wave scattering model of Kohout and Meylan (2008) is used to compute wave attenuation in the MIZ. For more details, see poster of Kohout et al. (#0021).
Floe Breaking Parameterization
Acknowledgments This work is part of the MIZ project funded by Total E&P and of the Waves-in-ice Forecasting for Arctic Operators (WIFAR) project collaboratively funded through the Research Council of Norway (PETROMAKS) and Total E&P.
References Kohout, Dumont and Bertino (2010) Modelling Ocean Waves in a Marginal ice Zone. Poster #0021, IPY Oslo Conference. Kohout and Meylan (2008) A model for wave scattering in the Marginal Ice Zone based on a two-dimensional floating elastic plate solution. J. Geophys. Res., 113: C09016. Langhorne, Squire, Fox and Haskell (1998) Break-up of sea ice by ocean waves. Annals Glaciol., 27: 438-442. Shen, Hibler and Leppäranta (1987) The role of floe collisions in sea ice rheology. J. Geophys. Res., 92: 7085-7096. Timco and O'Brien (1994) Flexural strength equation for sea ice. Cold Reg. Sci. Tech., 22: 285-298. Toyota, Haas and Tamura. (In press) Size distribution and shape properties of relatively small sea ice floes in the Antarctic marginal ice zones in late winter. Deep Sea Res. II.
Atmospheric Forcing
Sea Ice Model
if D < Dc MIZ rheology else EVP rheology end
Waves-in-ice Model
Floe Breaking Parameterization
Ocean Model
waves-in-ice�spectrum
ocean wave�spectrum air-sea fluxes
ice conditions
floe size�distribution
The TOPAZ Fram Strait Sea Ice-Ocean Model
Proof of concept using a one-dimensional numerical implementation
(Tm, Hs)
LMIZ
Dmax < Dc�marginal ice zone
Dmax > Dc�ice pack
open ocean�wave spectrum
An ice plate of constant thickness will fail whenever:�1) the flexural stress reaches the flexural strength σfc or; 2) the flexural strain reaches a maximum value εfc.
A critical wave amplitude is derived for both failure modes:
If A ≥ min(Aσ,Aε) the plate breaks up in many floes with a maximum size Dmax= λ/2. Fatigue is parameterized as σfc
µσfc with µ < 1 following Langhorne et al. (1998). Flexural strength is taken from Timco et al. (1994).
We present here a one-dimensional implementation of the framework described above as a way of testing components, parameterizations, and numerical schemes before its implementation in the HYCOM TOPAZ system. The goal is to compute the width of the MIZ in Fram Strait based floe size and properties of the floe size distribution.
ice concentration
viscous-plastic
MIZ ice
visc
osi
ty (
kg s
-1)
Comparison between�EVP and MIZ rheologies
* Contact author: [email protected]
Main Results and Conclusion
• The model produces reasonable values for the width of the MIZ given the numerous assumptions (deep water wave dispersion, one-point wave source, mean direction instead of directional spectrum, etc.)
• Ice concentration does not capture the location of the MIZ.
• Maximum floe size < 100m in qualitative agreement with observations (e.g. Toyota et al. 2010).
• Preliminary results in one dimension show the relevance of the architecture.
Future Work
• Implement in two dimensions in HYCOM. • Validate in Fram Strait (WIFAR cruise Sep. 2010). • Refine waves-in-ice models and parameterizations.
Parameter Symbol Value
Flexural strength σfc 0.6 MPa Maximum flexural strain µεfc 3.5 × 10-5
Fa2gue µ 0.6 – 1.0 Sea water density ρw 1025 kg m-3
Gravita2onal accelera2on g 9.81 m s-2
Cri2cal floe size Dc 500 m Ice Young’s Modulus E 0.5 MPa Poisson’s ra2o ν 0.3
€
Aε =µε fcλ
2
4πh∝T 4
h
Hs (m)
Photo: Courtesy of Prof. V. Squire
Tm (s)
c < 0.98
D < 500m
model
obs
2007
2008
2009
€
Dmin =π2
Eh3
3gρw 1−ν2( )
1 4
€
S T( ) =1.258π
T 5Hs2
Tm4 e−1.25 T Tm( ) 4
A nested configuration of HYCOM has been set-up in Fram Strait at a 3.5-km resolution. Boundary conditions are provided by TOPAZ system which is the Arctic monitoring and forecasting system of MyOcean.
Sea ice dynamics is modeled using a dual rheology composed of the elastic-viscous-plastic (EVP) rheology and a novel rheology for the MIZ developed at the NERSC based on the ideas of Shen et al. (1987). This formulation needs a criterion that will determine the dynamical regime. The current version uses an ice concentration threshold (mainly for testing purposes), but the objective is to include waves and use a criterion based on floe size.
Greenland Ice edge Dmin Dmax
