Transcript Synopsis LOICZ Project

The dynamics of estuarine
turbidity maxima
Stefan Talke * Huib de Swart * Victor de Jonge
Groningen Workshop
March 3, 2006
Overview
• Experiments
• Analysis
• Modeling
•Ultimate Goal: Understand the effect of biology and physical
processes on each other and morphology.
•Preliminary Goal: Describe and analyze physical and
biological processes separately
Overview of Experiments
• Measurements at Longitudinal and Cross-Sectional Transects
– Boats from RWS, WSA Emden, and NP GmBH used (Thanks!)
– Both fixed station and continuous measurements
Germany
Netherlands
Longitudinal Transects: 10 times since Feb. 2005
Cross-sectional Transects: Mar. 2005, Feb. 2006, Summer 2006?
Measurements
ADCP (Acoustic Doppler Current Profiler)
Velocity measured continuously in water column (~0.5 Hz)
Backscatter used to estimate sediment concentration
Bottom tracking used to estimate boat velocity
But, in turbid water, signal disappears!
Water
Fluid Mud
Consolidated Bed
600 kHz ADCP measures velocity
and backscatter (turbidity) in 0.25 m
increments
Measurements
Solution:
Use external echo-sounders and differential GPS
Boat Velocity from GPS
Water
Fluid Mud
Consolidated Bed
210 kHz echosounder penetrates to
the fluid mud layer
Measurements
Solution
Use external echo-sounders and differential GPS!
Boat Velocity from GPS
Water
Fluid Mud
Consolidated Bed
15 kHz echosounder penetrates to the
bed
Measurements
On-board Flow-thru system
Pump water into a bucket continuously
Measure: Turbidity, Fluorescence, Salinity,
Temperature, Oxygen
Take care to prevent light, bubbles!
Water
Fluid Mud
Consolidated Bed
Measurements
Fixed Point Measurements
CTD Casts with OBS + Oxygen sensor
 measure Salinity, Temperature, Turbidity, Depth, Oxygen
Water Samples (surface and water column)
 Analyzed for SSC, Organic Carbon, Nitrates, Silicates,
Phosphorous, pH, Algal counts and types
CTD Casts
Water Samples
Consolidated Bed
Measurements
Long Term Fixed Point Measurements: “X”
Monitored by NLWKN and WSA Emden
Measure: Tidal Stage, Salinity, Turbidity, Oxygen, pH,
Velocity, Temperature, Sediment Concentration
CTD Casts
Germany
X
X
Netherlands
XX X
X
X
X
X
Cross-Sectional Data at Pogum
Germany
Netherlands
Longitudinal Transects
Cross-sectional Transect: March 2005
Echosounder Data
Transect at Pogum
Fluid
Mud
8:27 am
500 m
Echosounder Data
Transect at Pogum
Fluid
Mud
8:35 am
500 m
Echosounder Data
Transect at Pogum
Fluid
Mud
9:21 am
500 m
Echosounder Data
Transect at Pogum
Fluid
Mud
9:36 am
500 m
Echosounder Data
Transect at Pogum
Fluid
Mud
10:03 am
500 m
Echosounder Data
Transect at Pogum
Fluid
Mud
10:25 am
500 m
Echosounder Data
Transect at Pogum
Fluid
Mud
10:36 am
500 m
Echosounder Data
Transect at Pogum
Fluid
Mud
10:47 am
500 m
Echosounder Data
Transect at Pogum
Fluid
Mud
11:22 am
500 m
Echosounder Data
Transect at Pogum
Fluid
Mud
11:32 am
500 m
Echosounder Data
Transect at Pogum
Fluid
Mud
15:11 pm
500 m
Echosounder Data
Transect at Pogum
Fluid
Mud
15:14 pm
500 m
Echosounder Data
Transect at Pogum
Fluid
Mud
15:56 pm
500 m
Echosounder Data
Transect at Pogum
Fluid
Mud
16:00 pm
500 m
Question: What does sediment concentration from ADCP backscatter look like?
Research Questions
• What are the sediment concentrations
– Analysis of ADCP data
• What does mixing and turbulence look like
over a tidal period
– Research project of Robbert Schippers (Msc):
• Analyze field data to estimate turbulent mixing
• Apply GOTM 1-D vertical turbulence model
Sediment Calibration from
Backscatter
• Need to estimate attenuation of sound due to
water and sediment
Water
attenuation
(range 0.050.2)
Sediment Calibration from
Backscatter
• Highly dependant on grain size, density,
frequency
Sediment attenuation
coefficient, 600 kHz
With 2 ADCP’s of
differing frequency,
can estimate mean
grain size
Maximum
at ~ 2
microns
Calculate Absolute backscatter
 Tt R 2
S v  C  10 log 10 
 LP
t


  2 R  K c  E  E r



• Loss is due to spreading of beam and attenuation
• R= distance along beam to adcp bin (20 degrees offset
from vertical)
• E= measured backscatter
• Alpha= combined, integrated attenuation
• C,L,Kc,Er are instrument constants
Fit regression line sed-conc to abs.
backscatter
Since attenuation depends on sediment
concentration in profile…
1. make initial estimate of sed conc to
calculate backscatter
2. Use backscatter to make linear
regression
3. Re-estimate sediment concentration,
recalculate attenuation, and repeat.
•
What about changes in floc size?
Non Linear Range
Linear Range
What about non-linear range?
Future Steps:
1. Calibrate non-linear range with Feb. 2006 data
2. Estimate mean grain size using backscatter from 2
ADCP’s on board one ship (Friesland)
Cross Section Sediment
Concentration Profiles at Pogum
(March 8,2005)
Germany
Water level (m)
Netherlands
Pogum
Water level (m)
Time (hours)
High turbidity
evident
mg/L
Note structures
in sediment
profile
Non-linear
Range:
> 5 g/L
8:30 am
flood
Water level (m)
Time (hours)
Turbulence collapses,
Sediment settles
mg/L
Sharp gradient
between water and
fluid mud
Fluid mud pools in
channel and shoal
11:30 am
slack
Non-linear
Range:
> 5 g/L
Water level (m)
Time (hours)
Very high turbidity
and fluid mud
Closer to ETM
than earlier
measurements
mg/L
8:30 am
16:00
Ebb
Non-linear
Range:
> 5 g/L
Vertical Observations: Interesting
Salinity Profiles
Flood (morning)
Salinity often (but not always) decreases towards bed.
Need to investigate density profiles…
Vertical Observations: Interesting
Salinity Profiles
Ebb (afternoon)
As fluid mud is approached, salinity goes down. Measurment artifact?
Or real physics?
Why the low salinity? Perhaps not mixed with rest of water column?
Are low salinities evidence of turbidity currents?
How does mixing change between flood and ebb?
Vertical Observations—density
profiles
Including sediment
concentration essential
for water column stability
Note again sharp
transition to fluid mud
Vertical Observations—density
gradients
Positive means unstable
Salinity profile dominates
upper water column
Sediment profile dominates
lower water column
Vertical Observations—Richardson
number
9:00 am (end
of flood tide)
Richardson # measures
ratio of shear (turbulence)
to density gradient
(buoyancy)
 >0.25 density dominates
 < 0.25 shear dominates
 < 0 Unstable
 Turbulence highly damped
Summary vertical and crosssection measurements
• High sediment concentrations observed
– How and at what tidal phase is sediment
being mixed into upper water column?
• Periodic formation of fluid mud layer
– Collapse of turbulence, formation of flocs
Longitudinal Data
Germany
Netherlands
Longitudinal Transects—ADCP measurements in March, April,
June, July, September 2005, and Feb. 2006
Longitudinal Results—Turbidity and
Salinity
Upstream
Downstream
From NP
Aanderaa Probe
In flow-through
system
Distance downstream from Herbrum (km)
Large horizontal salinity and turbidity gradient (Turbidity not yet calibrated)
Question: Are there density driven currents from both salinity and turbidity?
Longitudinal Results—Oxygen and
Fluorescence
Not yet calibrated—Fluorometer
How are Dissolved oxygen and fluorescence related to the physical
parameters of system?
Longitudinal Results--Salinity
Knock
Pogum
Terborg
Fixed NLWKN salinity measurements (note different scales)
Longitudinal Results--Sediment
Knock
Pogum
Terborg
Fixed NLWKN sediment concentration measurements
Note concentrations of up to 10 g/L; in summer, > 25 g/L measured
Longitudinal Results—Density
Gradients
Knock-Pogum
Salinity gradient
Pogum-Terborg
Residual circulation proportional to density gradient
Note that sediment density gradient changes sign
Combined gradient
(salinity + sed. Conc.)
Thought Experiment
• Consider “Bath Tub”
Fresh
Water (Salinity
= 0)
Circulation
cell
Heavy,
turbid
water flows
bottom
Next,
add turbid
water along
to center
Fresh
waterhappens?
circulates to conserve mass
What
Thought Experiment
• Consider another situation
Salt Water
(Ocean)
Fresh Water
(River)
Now,
consider
which difference
density differences
Circulation
cellcase
fromindensity
due
occur
only (gravitational
from salinity circulation)
to salinity
 What Happens?
Thought Experiment
• Now, consider both together
+
Turbidity induced circulation
=
Salt
Water
Fresh Water
Salinity induced circulation
Salt
Water
Fresh Water
Hypothetical
Combined
Salt + turbidity
circulatation
Thought Experiment
Analysis:
 Possible explanation for observed, asymmetric turbidity profiles?
To be realistic, need freshwater flow Q, bed slope, friction, etc.
 Spread of turbid water critical for understanding
depleted oxygen levels and other biological processes
 Next step: Modeling
Salt
Water
Fresh Water
Fresh Water
Hypothetical
Combined
Salt + turbidity
circulatation
Development of Simple Model
• We make the following assumptions:
– No Tides—Consider only averages
– Constant horizontal salinity gradient
– Salinity well-mixed vertically
– Sediment Concentration is prescribed
– Balance between settling velocity and
turbulent mixing
Development of Simple Model
• Following equations solved analytically:
0 
dp
  g sin  
dx
dp

z
Az
  g
du
dz
Horizontal Momentum
dz
Vertical Momentum
   o   ( s  s 0 )  C
Prescribed Density variation
 
C 
wsC  (K z
) 0

z 
 z 
Sediment Mass Balance
H
 ubdz
Q
Water Mass Balance
0
• Basically, classical gravitational circulation model with
longitudinal sediment gradients as a forcing mechanism
Development of Simple Model
• Preliminary Results:
– Presribed Salinity gradient: 1 psu/km
– Sediment Concentration:
Preliminary Results
• Downstream of ETM at maximum gradient:
Salinity driven flow
Sal. + Sed. driven flow
Fresh Water out
Flow reversed
at bottom for large
sediment gradients
Saline Water
into estuary
Flow reversal
Not reversed when
sediment gradient not
large
Saline flow shifted
upwards
Preliminary Results
• Upstream of ETM at maximum gradient:
Salinity driven flow
Sal. + Sed. driven flow
Flow out of estuary
Flow enhanced
at bottom because
salinity and sediment
gradients in same
direction
Flow into estuary
Could turbidity currents
explain upstream shift
of turbidity zone and
asymmetrical profile?
What about other processes?
• Tidal asymmetry
Could this be the cause of upward migration of turbidity zone?
--Suggested by C. Habermann, others but never tested
--2D Analytic model of de Swart and Schutelaars will test
this hypothesis (currently being worked on)
And, havn’t forgotten biology…
• Modeling scalars
– Model of H.M. Schuttelaars being adapted to model algae
– Important processes—interaction of sediment concentration
with light availability
No Growth
Growth
Growth
Light availability:
Io depends on time of day,
season, cloud cover, etc.
Attenuation coefficient ‘k’:
kw = water attenuation
kb = self shading by
algae
ks = shading by
sediment (proportional to
SSC)
kd = shading by detritus
Summary
• Questions to come out of measurements
– Why the funny vertical salinity profiles?
– Can we calibrate non-linear range of ADCP
backscatter?
– Are there turbidity currents being driven by
high sediment concentrations? How does this
interact with longitudinal salinity gradient?
– What controls the position and longitudinal
extent of the ETM plume?
Summary—Numerical Modeling
Algae
Fresh(er) Water
Periodic Stratification
Saline Water
Gravitational Circulation
Flocculation and
settling
Turbulence vs.
Bouyancy
Exchange
Fluid Mud
Bed friction
Consolidated Bed
Bacteria with
high oxygen
consumption
Turbidity Current
Summary—Next steps
• Continue developing models to investigate
these questions
– Turbidity induced circulation
– Effect of tidal assymmetry on residual
circulation
– Vertical mixing processes near ETM
• Both Salinity and Sediment induced density
differences and stratification
Summary—Next steps
Analyze recent cross-sectional measurements over tidal period
 2 ADCP’s on one ship (Friesland)—estimate mean grain size
 Calibrate non-linear range
Analyze vertical mixing and turbulence
Analyze residual currents, fluxes over tidal period
Analyze near bed turbidity currents
Feed results into Models to make more realistic
1200 kHz ADCP
600 kHz ADCP
Water
Fluid Mud
Consolidated Bed
Thanks for listening!