Nordic Seas Water Property Evolution - United States Navy · dynamics subject to thermobaricity and...

1
θ 0 and S 0 are cool and fresh θ 1 and S 1 are warm and salty The buoyancy force resulting from thermobaricity is unstable and largest for cold, fresh water. Thermobaric instability is more likely at high latitudes. Nordic Seas Water Property Evolution Robert W. Helber, David A. Hebert, Jay F. Shriver, and Alan J. Wallcraft Naval Research Laboratory, Code 7321, Stennis Space Center, MS Ocean Sciences Meeting | Portland, OR| 11 – 16 February 2018 Abstract Cold fresh water from the Arctic and warm salty water from the Atlantic oceans meet in the Nordic Seas where the vertical structure has been evolving in recent decades (Rudels et al. 2012). Because water properties are changing and seawater non- linearities are stronger at low temperatures and salinities making dynamics subject to thermobaricity and cabbeling, numerical ocean forecasting is more complex. Using the Modular Ocean Model version 6 (MOM6) with both isopycnal layer and z-level coordinates and MIT General Circulation Model (MITgcm) in both hydrostatic and non-hydrostatic mode, we demonstrate a mechanism for watermass formation or modification. Figure 1. Carmack, E. C. (2007). "The alpha/beta ocean distinction: A perspective on freshwater fluxes, convection, nutrients and productivity in high-latitude seas." Deep- Sea Research Part II-Topical Studies in Oceanography 54(23-26): 2578-2598. Area of interest Thermobaric Instability Envision an ocean condition where potential density is constant everywhere, but potential temperature and salinity are different above and below the wave interface. θ 0 and S 0 are constant above interface θ 1 and S 1 are constant below interface σ 0 is constant everywhere cool fresh water accelerates downward due to seawater non-linearities ( ) ( ) 1 1 0 0 0 0 , , 2 , , 2 0 g Sp p S p p ρθ ρθ +∆ +∆ < force/volume θ 0 and S 0 are warm and salty θ 1 and S 1 are cool and fresh system is stabilized by non-linearities ( ) ( ) 1 1 0 0 0 0 , , 2 , , 2 0 g Sp p S p p ρθ ρθ +∆ +∆ > force/volume more unstable at low temperature and high pressure more stable at low temperature and high pressure ( ) ( ) 1 1 0 0 0 0 , , 2 , , 2 g Sp p S p p ρθ ρθ +∆ +∆ Model Initialization We are using the GFDL Modular Ocean Model, version 6 (MOM6; see refs.), in isopycnal layer and z-level coordinates and the MIT General Circulation Model (MITgcm; Marshall et al. 1997) in hydrostatic and non-hydrostatic modes. Initialization profiles (red and blue) are designed to mimic Greenland Sea water masses. Red lines are warm water over cool water Blue lines are cool water over warm water The black lines in b, c, and d correspond to the red line in a. MOM6 and MITgcm low resolution simulations are in two-dimensions, 60 km wide and 4000 m deep, all initializations have the same potential density. References: Carmack, E. C. (2007) The alpha/beta ocean distinction: A perspective on freshwater fluxes, convection, nutrients and productivity in high- latitude seas, Deep-Sea Research Part II-Topical Studies in Oceanography 54(23-26): 2578- 2598. Feistel, R. (2003) A new extended Gibbs thermodynamic potential of seawater, Progress in Oceanography, 58, 43–114. Marshall, J., et al. (1997) A finite-volume, incompressible Navier Stokes model for studies of the ocean on parallel computers, Journal of Geophysical Research-Oceans 102(C3): 5753- 5766. Rudels, B., et al. (2012) The East Greenland Current and its impacts on the Nordic Seas: observed trends in the past decade, Ices Journal of Marine Science, 69(5): 841-851. MOM6 is a layered model with Arbitrary Lagrangian Eulerian (ALE) capabilities originating from the Hallberg Isopycnal Model. See: Dunne, J. P., et al. (2012). GFDL's ESM2 Global Coupled Climate-Carbon Earth System Models. Part I: Physical Formulation and Baseline Simulation Characteristics, Journal of Climate 25(19): 6646-6665. Legg, S., et al. (2006) Comparison of entrainment in overflows simulated by z- coordinate, isopycnal and non-hydrostatic models, Ocean Modelling 11(1-2): 69-97. Hirt, C. W., et al. (1974) Arbitrary Lagrangian-Eulerian Computing Method for All Flow Speeds, Journal of Computational Physics 14(3): 227-253. MOM6 z-levels MITGCM z-levels High Resolution z-level Both MOM6 and MITgcm are run with 256 horizontal grid points and 320 vertical grid points in z-level. MOM6 solves the dynamical equations in layer formulation where as MITgcm solves them in finite difference formulation. MOM6 7-layers High Resolution Layer MOM6 is run with 256 horizontal grid points and 7 isopycnal layers. cool water over warm water warm water over cool water Same Density Low Resolution Layer MOM6 is run with 64 horizontal grid points and 7 isopycnal layers. Low Resolution z-level MOM6 is run with 64 horizontal grid points and 80 z-levels. cool over warm cool over warm warm over cool warm over cool non- linear EOS linearized EOS Conclusions: Present models are missing aspects of non-linear ocean thermodynamics (e.g Feistel 2003) The non-linear effect of thermobaricity can be modeled in present models The impact of the thermobaricity is to weaken (strengthen) stratification when cool (warm) water overlays warm (cool) water In the z-level simulations, thermobaricity and cabbeling create subsurface temperature minimum MOM6 isopycnal layer coordinates can model thermobaricity but not diapcynal cabbeling and produces no temperature minimum Linearized test cases produce no thermobaricity effect Large Scale Context Carmack (Deep Sea Research II, 2007): “‘downhill’ transport of Pacific Water (PW) into the Arctic Ocean and then North Atlantic” provides “supply of freshwater” … “as intermediate water (IW) and deep water (DW) in “ the Nordic Seas (Figure 1). Observations Temperature Salinity Sigma0 cool water over warm water warm water over cool water Thermobaric Effect Temperature Salinity Sigma0

Transcript of Nordic Seas Water Property Evolution - United States Navy · dynamics subject to thermobaricity and...

Page 1: Nordic Seas Water Property Evolution - United States Navy · dynamics subject to thermobaricity and cabbeling, numerical ocean forecasting is more complex. Using the Modular Ocean

θ0 and S0 are cool and fresh

θ1 and S1 are warm and salty

The buoyancy force resulting from thermobaricity is unstable and largest for cold, fresh water. Thermobaric instability is more likely at high latitudes.

Nordic Seas Water Property EvolutionRobert W. Helber, David A. Hebert, Jay F. Shriver, and Alan J. WallcraftNaval Research Laboratory, Code 7321, Stennis Space Center, MS

Oce

an S

cien

ces M

eetin

g |

Port

land

, OR|

11

–16

Feb

ruar

y 20

18

AbstractCold fresh water from the Arctic and warm salty water from the Atlantic oceans meet in the Nordic Seas where the vertical structure has been evolving in recent decades (Rudels et al. 2012). Because water properties are changing and seawater non-linearities are stronger at low temperatures and salinities making dynamics subject to thermobaricity and cabbeling, numerical ocean forecasting is more complex. Using the Modular Ocean Model version 6 (MOM6) with both isopycnal layer and z-level coordinates and MIT General Circulation Model (MITgcm) in both hydrostatic and non-hydrostatic mode, we demonstrate a mechanism for watermass formation or modification.

Figure 1. Carmack, E. C. (2007). "The alpha/beta ocean distinction: A perspective on freshwater fluxes, convection, nutrients and productivity in high-latitude seas." Deep-Sea Research Part II-Topical Studies in Oceanography 54(23-26): 2578-2598.

Area of interest

Thermobaric InstabilityEnvision an ocean condition where potential density is constant everywhere, but potential temperature and salinity are different above and below the wave interface.

θ0 and S0 are constant above interface

θ1 and S1 are constant below interface

σ0 is constant everywhere

cool fresh water accelerates downward due to seawater non-linearities

( ) ( )1 1 0 0 0 0, , 2 , , 2 0g S p p S p pρ θ ρ θ + ∆ − + ∆ <

force/volume

θ0 and S0 are warm and salty

θ1 and S1 are cool and fresh

system is stabilized by non-linearities

( ) ( )1 1 0 0 0 0, , 2 , , 2 0g S p p S p pρ θ ρ θ + ∆ − + ∆ >

force/volume

more unstable at low temperature and high pressure

more stable at low temperature and high pressure

( ) ( )1 1 0 0 0 0, , 2 , , 2g S p p S p pρ θ ρ θ + ∆ − + ∆

Model InitializationWe are using the GFDL Modular Ocean Model, version 6 (MOM6; see refs.), in isopycnal layer and z-level coordinates and the MIT General Circulation Model (MITgcm; Marshall et al. 1997) in hydrostatic and non-hydrostatic modes. Initialization profiles (red and blue) are designed to mimic Greenland Sea water masses.

Red lines are warm water over cool water

Blue lines are cool water over warm water

The black lines in b, c, and d correspond to the red line in a. MOM6 and MITgcm low resolution simulations are in two-dimensions, 60 km wide and 4000 m deep, all initializations have the same potential density.

References:• Carmack, E. C. (2007) The alpha/beta ocean distinction: A perspective on freshwater fluxes, convection, nutrients and productivity in high-latitude seas, Deep-Sea Research Part II-Topical Studies in Oceanography 54(23-26): 2578-2598.• Feistel, R. (2003) A new extended Gibbs thermodynamic potential of seawater, Progress in Oceanography, 58, 43–114.• Marshall, J., et al. (1997) A finite-volume, incompressible Navier Stokes model for studies of the ocean on parallel computers, Journal of Geophysical Research-Oceans 102(C3): 5753-5766.• Rudels, B., et al. (2012) The East Greenland Current and its impacts on the Nordic Seas: observed trends in the past decade, Ices Journal of Marine Science, 69(5): 841-851.

MOM6 is a layered model with Arbitrary Lagrangian Eulerian (ALE) capabilities originating from the Hallberg IsopycnalModel. See:• Dunne, J. P., et al. (2012). GFDL's ESM2 Global Coupled Climate-Carbon Earth System Models. Part I: Physical Formulation and Baseline Simulation Characteristics, Journal of Climate 25(19): 6646-6665.• Legg, S., et al. (2006) Comparison of entrainment in overflows simulated by z-coordinate, isopycnal and non-hydrostatic models, Ocean Modelling 11(1-2): 69-97.• Hirt, C. W., et al. (1974) Arbitrary Lagrangian-Eulerian Computing Method for All Flow Speeds, Journal of Computational Physics 14(3): 227-253.

MOM6 z-levels MITGCM z-levels

High Resolution z-levelBoth MOM6 and MITgcm are run with 256 horizontal grid points and 320

vertical grid points in z-level. MOM6 solves the dynamical equations in layer formulation where as MITgcm solves them in finite difference formulation.

MOM6 7-layers

High Resolution LayerMOM6 is run with 256 horizontal grid

points and 7 isopycnal layers.

cool water over warm water

warm water over cool water

Same Density

Low Resolution Layer MOM6 is run with 64 horizontal grid

points and 7 isopycnal layers.

Low Resolution z-level MOM6 is run with 64 horizontal grid

points and 80 z-levels.

cool

ove

r w

arm

cool

ove

r w

arm

war

m

over

coo

l

war

m

over

coo

l

non-linear EOS

linearized EOS

Conclusions:• Present models are missing aspects of non-linear ocean thermodynamics

(e.g Feistel 2003)• The non-linear effect of thermobaricity can be modeled in present models• The impact of the thermobaricity is to weaken (strengthen) stratification

when cool (warm) water overlays warm (cool) water• In the z-level simulations, thermobaricity and cabbeling create subsurface

temperature minimum• MOM6 isopycnal layer coordinates can model thermobaricity but not

diapcynal cabbeling and produces no temperature minimum• Linearized test cases produce no thermobaricity effect

Large Scale ContextCarmack (Deep Sea Research II, 2007): “‘downhill’ transport of Pacific Water (PW) into the Arctic Ocean and then North Atlantic” provides “supply of freshwater” … “as intermediate water (IW) and deep water (DW) in “ the Nordic Seas (Figure 1).

Observations

Temperature Salinity

Sigma0

cool water over warm water warm water over cool water

Thermobaric Effect

Temperature Salinity Sigma0