1 Marginal Thermobaric Stability in the Weddell Sea Miles McPhee McPhee Research Company
Jan 21, 2016
1
Marginal Thermobaric Stability in the Weddell Sea
Miles McPhee
McPhee Research Company
2
Thermobaric Instability
Following Loyning and Weber, JGR, 102, p. 27875
),,(),,( 00 zLLzUU pSTpST
1 2
z
z = 0UU ST ,
LL ST ,
ambient
2-layer system, upper layer colder and less saline
)())((10 mm SSTTz Linearized equation of state:
STz )(
0
021
2
1
TT
TTT
TTT
mL
mU
0)( 2110
21
TTz
zz 10)( Thermal expansion coefficient increases with depth
3
Greenland Sea
Weddell Sea
4
Marginal stability line
222011 / HS Strength of thermobaric tendency must exceed the background stratification
1020
2
2
12 1
HS
SSH
Hdeepplumedeepplume
deepplume SHH 12
5
6
Microsoft PowerPoint
Presentation
7
Microsoft PowerPoint
Presentation
8
tionstratificahalinebackgroundnormalizedS
citythermobariofstrengthnormalized
deep
plume
:
:
9
10
ANZFLUX Ship CTD station 50, linearized
11
2.4 hour average of turbulence measurements centered at time 206.35 (Warm Regime drift). Circles are averages; lines are twice the std dev of the 15-min samples.
''TwcH pf 4/122
* '''' wvwuu
12
13
Increase Sml
Two-layer (Type II) stability diagram
following Akitomo (1999) for idealized
Ship Station 50
14
The “thermobaric barrier” calculation:
(1) Calculate the actual density (pressure included), subtract density of a water column with mixed layer properties. Determine the level (zmax) of the maximum difference: max.
(2) Determine the sensible heat that must be vented to reduce water temperature above zmax to ml.
(3) Add the latent heat loss required to increase salinity (by freezing) enough to eliminate at zml
(4) Htot is the total heat loss.
15
22 571/150 mWmonthmMJ
16
Pentagrams indicate Hto t < 100 MJ/m2
17
18
22 181/47 mWmonthmMJ
19
20
4.7plume
21
Microsoft PowerPoint
Presentation
22McPhee, Kottmeier and Morison, JPO, 1999
27.4 W m-2
23
Ice temperatures from the AWIBuoy thermistor string
24
Mean values almost identical: 30 W m-2
25
Microsoft PowerPoint
Presentation
Friction velocity prescribed from buoy results. Heat flux also from buoy ice measurements but scaled by the calculated ice thickness
Ocean heat flux calculated prognostically, ice thickness determined by enthalpy balance at the interface.
Dynamic mixed layer depth based on buoyancy frequency (pot density). Scalar based on difference from near surface value.
The model neglects thermobaric effects but calculates thermobaric barrier parameters at each time step.
26
The 1-D model forced with buoy data and initialized with YU075No thermobaricity effect considered.
27
The 1-D model forced with buoy data and initialized with YU075Eddy viscosity set to 2000 cm2/s across vertical domain after 217.75.
28
Horizontally Homogeneous Model Results
• Initialize model with every profile with Htot < 100 MJ m-2 forced by buoy time series (38)
• 27 1-D profiles became unstable by the end of August
29
Is there a simple way of getting a handle on eddy viscosity and scalar diffusivity when thermobaric mixing is occurring?
1. Parameterize entrainment process in terms of conversion of PE to TKE
2. Base the mixing length on a fraction (of the entrained layer depth
3. Then
massunitTKE
E
q
qHK
2
30
31
32
33
Summary
• Thermobaric instability was not observed directly during the ANZFLUX 94 project but there is a strong inference that it occurred shortly after.
• About ¼ of the profiles observed with the yoyo CTD system during the Maud Rise drift went thermobarically unstable by the end of winter in a simple 1-D model forced with drifting buoy data.
• In the model, preconditioning of the initial density profile to include distinct step-like structure in the upper pyncnocline was necessary for instability.
• Steps were found mostly in the “halo” region surrounding Maud Rise (2500-3000 m isobaths)
34
Summary (cont)
• There may be a good chance of encountering episodes of Type II convection near Maud Rise in late winter.
• Measuring turbulent dissipation rates and turbulent fluxes directly during a Type II episode is feasible based on ANZFLUX experience. Such data would be of great value in evaluating and guiding numerical model development.
• Even in the absence of direct Type II convection, studying processes that maintain the step structure and “pycnocline weather” in the Weddell would add significantly to our understanding of the system.