Moum_IWISE_POReview2013 - Ocean Mixing Group

Report
Direct Mixing Measurements using χpods in IWISE
Profiling Dissipation Measurements using
χpods on Moored Profilers (Moum/Nash)
Shipboard LADCP/χpod profiling
of Internal Wave Structure (Nash/Moum)
Jonathan D. Nash & James N. Moum
College of Earth, Ocean and Atmospheric Sciences
Oregon State University
with help from: Byungho Lim (OSU), Andy Pickering and Matthew Alford (APL-UW)
and thanks to Ming-Huei Chang, Maarten Buijsman, Luc Rainville, Alexander Perlin,
Ray Kreth, Mike Neeley-Brown, John Mickett, Eric Boget, Amy Waterhouse, Zoe
Parsons, Jen MacKinnon, Harper Simmons
Objectives
General
• quantify turbulence dissipation where large amplitude internal waves
are generated
Particular
• capture the energetics of the largest scales that directly extract energy
from the barotropic tides
• while simultaneously measuring mixing associated with the
turbulence that occurs at millimeter and millisecond scales.
• through direct observation, to assess the means by which waves form
and break, elucidate the structure/evolution of the wave breaking, and
quantify the dissipation that induces irreversible mixing
Methods
• χpods on moorings
Stablemoor
Methods
• χpods on moorings
• χpod-like devices on moored profilers
Methods
χpod -LADCP
• χpods on moorings
• χpod-like devices on moored
• χpod on shipboard CTD for full
ocean depth turbulence profiling
Methods
• χpods on moorings
• χpod-like devices on moored profilers
MP χpods
• χpod on shipboard CTD for full
ocean depth
turbulence profiling
• fabricated and deployed 5-component array of
moorings to capture the 2D evolution of the largerscale dynamics
Profiling Dissipation Measurements
using χpods on Moored Profilers
data example
fast thermistors
on APL MP
Kraichnan
theoretical
spectrum
{
εχ=N2 χ/(2 Γ Tz2)
Products →
Testedprofiling
– Puget experiment
Sound 2009
1st continuous
deep-ocean
LT Thorpe
(overturn)
scales
Refined for 2010 (MPN)
ndrate
χ
temperature
variance
dissipation
mesoscale
dominates
half
T current (Kuroshio?)
Deployed
2011
on N1,2 N2
KT turbulence
diffusivity
(Osborn-Cox)
•
elevated
turbulence
at base
of
current
all
data
returned
ε TKE dissipation rate (indirect)
•
is this friction on a western boundary
Km turbulence viscosity
current?
New mean speed / dissipation sensor for
use on χpods and in general on
moorings
inexpensive,
lightweight, low power,
standalone velocity
sensor
χpod
mooring N1 – χpod at 2000 m
1 day time series
compensated
pitot tube
leading to a new GustT
combination probe
speed sensor at low f
dissipation sensor at high f
ε=2x10-7 m2s-3
not acoustic, hence
requires no scatterers
quiet
characterization of
sensor includes tests
in
• wind tunnel
• tidal channel
2nd continuous deep
turbulence profiling time
series
N2 (1830m water depth)
2011
2 units constructed and
deployed – both worked
– only 1 MP profiled
Chipod-LADCP-CTD
fast-T
3-axis accel
3-axis gyro
compass
USB-data
Nash & Moum
inferred
turbulence
(blue)
direct
turbulence
(green)
OSU Ocean Mixing χ-pod/LADCP
direct measurements of abyssal
turbulence from standard shipboard
CTD.
 permits rapid deep profiling
 direct turbulence differs from that
inferred from overturns
low noise-floor (but N2-dependent)
Above: TKE dissipation rate from
LADCP/chipods (green) and Thorpe analyses
(blue) at one of the most energetic stations
sampled during IWISE.
contrasting structures from detailed measurements at 2 ridges
1) broadly-distributed dissipation on the east ridge
2) big breaking lee waves on the west ridge
west
MP
chipods
east
A1 – mid-column dissipation at the generation site
1440 m water depth
mid-column dissipation
not dominated by a single
breaking wave…
Byungho Lim
A1 – mid-column dissipation at the generation site
observation / model comparison (MITgcm / Buijsman)
similar tidal fields, but
water-column instabilities
are not captured by MITgcm
and model dissipation is
mostly near the bottom.
Byungho Lim
T-Chains on the
West Ridge
eastward
westward
west
Jonathan Nash
T1
T2
N2
T3
Buijsman et al
T4
 30-50 m sensor
spacing to detect
overturns
500 m
overturns
 2-sec sampling to
capture inertial
subrange
 Vertical
synopticity (test
sampling schemes
of other platforms)
 3 months data
east
700 m
waves
T-Chains on the
West Ridge
spring tides / diurnal inequality
T1
T2
N2
T3
Mooring T3 during spring
+ diurnal inequality
T4
T-Chains on the
West Ridge
neap tides / semidiurnal period
T1
T2
N2
T3
Mooring T3 during
neap/semidiurnal period
T4
Time – mean structure / east ridge
Springs
(diurnal)
T1
T2
N2
T3
T4
Dissipation tied to lee waves
Strong spring/neap changes
Isopycnals displaced down
in the mean?
Lee-wave shifts closer to
ridge crest during neaps?
Neaps
(semidiurnal)
Dissipation evolution and scaling
ε~ u3bt
T1
T1
T2
T2
N2
N2
T3
T3
T4
T4
 dissipation scales with u3bt
(nonlinear!)
 consistent with Klymak et al
(2010)’s “recipe” for ε over a
supercritical ridge
… u3 because flux into trapped lee
waves ~( ubt x u2bt ) …
Summary Results
• 1st continuous turbulence profiling away from ship-based upper
ocean measurements
• χpod-CTD measurements have led to beginning of contribution
to Global Repeat Hydrogaphy Program
• NEW VELOCITY SENSOR
- speed + turbulence
leading way to new possibilities
• observational confirmation of Klymak etal (2010) ε scaling
• breaking waves: vertically-integrated ε O(1 W/m2)
comparable to flux divergence 5 kW/5 km
suggests significant local dissipative losses
• vertically-distributed turbulence may be difficult to model but
significant to water mass mixing through vertical flux divergence
Summary Results
continued contributions to NRL field science
MORT Mixing Over Rough Topography
BWE
Breaking Wave Effects in High Winds
technological:
• loan of OSU-developed instrumentation
• technical-level analysis
scientific:
• participation in science-level analysis
• contribution to publications
Moored profiler χpod estimates of turbulence dissipation rate, ε
LT – large-eddy length scale
statistic simply computed
from 1D profiles
- but an imperfect statistic
in an evolving 3D field
Lo – large-eddy length scale
defining buoyancy limit on
turbulence
Lo = √(ε/N3)
if LT = Lo, then ε = LT2 N3
is LT = Lo ?
Moored profiler χpod estimates of turbulence dissipation rate, ε
LT – large-eddy length scale
statistic simply computed
from 1D profiles
- but an imperfect statistic
in an evolving 3D field
Lo – large-eddy length scale
defining buoyancy limit on
turbulence
Lo = √(ε/N3)
if LT = Lo, then ε = LT2 N3
is LT = Lo ?
same data – different
definition of N2
pod
How do we know χpods work?
7 χpods on EQUIX
mooring yields 7 time
series of χ, ε
Equatorial Internal
Wave Experiment 2008
16-day experiment
at 0, 140W
Oct/Nov 2008
24h continuous
profiling of χ, ε
6-10 profiles/h
5 χpods on TAO
mooring yields 5
time series of χ, ε
Perlin & Moum, 2012 JAOTech
pod
How do we know χpods work?
χ
ε
profiler
χpods
χpods
Perlin & Moum, 2012 JAOTech
Comparison of ε
computed from χpod and
from pitot tube
A1 – mid-column dissipation at the generation site
Observation / model comparison at T3
Observation / Model
comparison at T3
Andy Pickering


model does a pretty good job
with the vertical distribution
and daily-averages
details are a little different
Andy Pickering
T1
T1
T2
T2
N2
N2
T3
T3
MITgcm / Buijsmann et al 2013
Dissipation evolution /
compare to MITgcm
T4
T4
First continuous deep
turbulence profiling
time series
MP-N
2010
mesoscale current (Kuroshio?) dominates 2nd half
elevated turbulence at base of current
is this friction on a western boundary current?
T-Chains on the
West Ridge
spring diurnals vs. neap semidiurnals
diurnal composite / spring
semidiurnal composite / neap
Summary Results
Integrated Dissipation from big
breaking waves contributes O(1 W/m2)
vertically-integrated ε
this suggests ΔFε= 5 kW/m in 5km…
ε is significant to FE!
Distributed Mixing
(detached from
bottom) is difficult to
model
 Can models accurately
capture mid-column ε and
its vertical distribution?
ε~ u3bt
Buijsman et al
Conclusions
 Can we assign error
bounds on model ε?
ε~ u3bt
log10 ε
observed

similar documents