Turbulent Mixing During an Admiralty Inlet Bottom Water Intrusion

Turbulent Mixing During an Admiralty Inlet Bottom

Water Intrusion

Philip Orton

Hats off to the A-Team:

Sally, Erin, Karin and Christie!

Profs extraordinaire: Rocky and Parker!

Motivation - Why Study Mixing/ Dissipation

sigma-t (kg m-3)

Echo Sounder Backscatter, 120 kHz, 04-Aug-2006, 11:28h

• Power/ importance • Difficulty for modeling

sorted profile raw profile

Plan-of-Attack

• Methods - dissipation/mixing estimation• Along- and across-channel comparisons• Consistency check: Observed dissipation vs

Expected?• Dynamical explanation for weak mixing

H0: Mixing during our study was spatially uniform

test: Compute buoyancy flux at many locations in along- and across-channel surveys

Field Program

88W

300kHz ADCP

Seabird 19 CTD

Echo Sounder

Full transect

Two half-transects

Cross-channel survey

Bush Point

Fine-Structure Instability Turbulence Analysis

A “Thorpe scale” analysis of ~138 CTD density profiles

The Thorpe scale (LT) is the rms re-sorting distance of all points in an overturning “patch”.

Method gives comparable results to microstructure instrumentation (e.g. Klymak and Gregg, JPO 34:1135, 2004).

Matlab mixing toolbox for CTD fine-structure and Lowered-ADCP

sorted profile raw profile

Mixing & Dissipation from Thorpe Scales

322 NLa T

NLaK T22

where a ≈ 1 (Klymak and Gregg; Peters and Johns, 2004)

We assume a mixing efficiency, ≈ 0.22, reasonable for stratified conditions (discussion in Macdonald and Geyer, JGR 109: C05004, 2004).

buoyancy frequency, N = [(g/d/dz)]0.5, is computed over overturn patch heights.

Dissipation of turbulent kinetic energy:

eddy diffusivity:

Station 16, 8/4 15:17h, slack after greater flood

Assume: (a) LO = LT, (b) LO is length-scale for TKE, (c) N is time-scale for dissipation.

Richardson Number, Ri = N2/Shear2

Ricrit= 0.25

Transect #1

FLOOD!

Transect #2

weak ebb

Transect #3

weak flood

Buoyancy Flux, B = N2K

Transect #1

FLOOD!

Transect #2

weak ebb

Transect #3

weak flood

Along-Channel Variability?

W/kg

Across-Channel Variability?

W/kg

Consistency Check: Tidal Dissipation

• Dissipation mean (away from bed) over entire study was 6.4 x 10-4 W/m3

• Hudson has mid-water column values of 10-2 (spring) to 10-3 W/m3 (neap; Peters, 1999)

• NOAA study (Lavelle et al., 1988) showed total tidal dissipation averages ~500 MW

• I estimate the total dissipation during our study as overturns + loglayer = 12 + 112 = 124 MW– assumed log layer dissipation ( ~ U*

3)– quad drag law: CD = 0.002 for velocity at 5-10m height

• This is reasonable, as our tidal range was ~3/4 the mean, U ~ range, ~ U3, and (3/4)3 = 0.4

Why Weak Mixing in Most Places?

Results suggest low mixing because tidal straining is overcoming mixing

horizontal

Richardson

(Stacey)

number, Rix

ebb EBB

Summary

• Was mixing during our study spatially uniform?– Cross-channel variability: results were inconclusive– Along-channel variability: No -- mixing was elevated

by a factor of O(10) in at least one hotspot

• Tidal dissipation estimates were consistent with a prior study, downscaled for below avg. tidal range

• Tidal straining can explain the low mixing that occurred in most of the estuary

• Excellent conditions for a bottom water intrusion!

Overturn Analysis: Quality Control

To avoid mistaking noise for overturns, each “resorting region” must pass various tests:

1) the rms (t,sort - t,raw) in a patch must be greater than the instrument noise ( = 0.002 kg m-3)

2) the T-S space tests of Galbraith and Kelley (J-Tech, 13:688, 1996)

a) near-linearity in the T- relationshipb) near-linearity in the S- relationship

3) rms run-length of overturn patch must be longer than 7 points total

Ambient Conditions

• Tides - end of a ~5 day period of weaker than normal tidal currents– Semidiurnal tidal range near annual low

– Diurnal tidal range on the rise, but below average

• Winds light• Riverflow into Puget Sound - [likely had an above

average summertime flow]