Conceptual Models and Parameterizations of Air-Water Gas Transfer Coefficients

1

• Conceptual Models and Parameterizations of
  Air-Water Gas Transfer Coefficients
• Proposal for Advancement to Candidacy
• Applicant Damon Turney
• Committee Chair Jeff Dozier
• Bren School of Environmental Science and
  Management
• Committee Members
• Jeff Dozier, Sanjoy Banerjee, Sally MacIntyre,
  Jordan Clark

2

• Where and How My Research Plan is Going To Make
  An Improvement to Env. Sci. Man.
• What Exactly The Research Plan Is
• What Ive already done
• What I plan to do
• Major Impediments, Time Schedule, and Financial
  Situation of the Research Plan

3

• Air-water gas transfersounds boringbut its
  important for many environmental issues.

4

• Reareation of water with inputs of BOD or COD
• Pathway for loss of toxic (or potentially toxic)
  chemicals
• Potentially a pathway for loss of nutrients
• Pathway for loss of carbon from terrestrial
  ecosystems
• Method of measuring community respiration

• On the average, the ocean absorbs carbon from
  atm.
• In tropics and subtropics the ocean losses carbon
  while in the mid or polar latitudes the ocean
  gains carbon
• Loss of dimethylsulfide to atmosphere
• Reareation of water with inputs of BOD or COD
• Potentially a pathway for loss of nutrients
• Pathway for loss of toxic (or potentially toxic)
  chemicals

• Reareation of water with inputs of BOD or COD
• Pathway for loss of toxic (or potentially toxic)
  chemicals
• Pathway for loss of carbon from terrestrial
  ecosystems
• Potentially a pathway for loss of nutrients

5

• Global Carbon Cycle
• Terrestrial regions often show that a significant
  amount of carbon is lost to the atmosphere
  through wetlands and lakes. The amount can reach
  up to 50 of net annual ecosystem carbon
  production.Richey, J. E., J. M. Melack, et al.
  (2002). "Outgassing from Amazonian rivers and
  wetlands as a large tropical source of
  atmospheric CO2." Nature 416(6881)
  617-620.Kling, G. W., G. W. Kipphut, et al.
  (1991). "Arctic Lakes and Streams as Gas Conduits
  to the Atmosphere - Implications for Tundra
  Carbon Budgets." Science 251(4991) 298-301.
• Oceanic regions are absorbing 30 of the
  anthropogenic CO2 emitted annually. Since the
  Industrial Revolution the oceans have absorbed
  about half of the total anthropogenic
  CO2.Sabine, C. L., R. Feely, et al. (2004). "The
  Oceanic Sink for Anthropogenic CO2." Science 305
  367-371.

6

• Loss of toxic (or potentially toxic) chemicals
  from water bodies
• USEPA http//www.epa.gov/OGWDW/dwh/t-soc/pcbs.ht
  ml http// www.epa.gov/OGWDW/dwh/t-soc/dioxin.htm
  l
• Community Action http//www.copa.org/library/art
  icles/bv/volatile.htmhttp//www.foxriverwatch.com
  /volatilization_pcbs_1.html
• Scientific Literature Shannon, J. D. and E. C.
  Voldner (1995). "Modeling Atmospheric
  Concentrations of Mercury and Deposition to the
  Great-Lakes." Atmospheric Environment 29(14)
  1649-1661.Dewulf, J. P., H. R. Van Langenhove,
  et al. (1998). "Air/water exchange dynamics of 13
  volatile chlorinated C1- and C2-hydrocarbons and
  monocyclic aromatic hydrocarbons in the southern
  North Sea and the Scheldt Estuary." Environmental
  Science Technology 32(7) 903-911.etc

7

• Reaeration of anoxic water
• Gulf of Mexico dead zone
• This is also a problem in highly eutrophic lake
  and river waters, particularly downstream of
  sewage outfall.

8

• The methods used to determine an air-water gas
  transfer rate could use improvement.

9

• The current options
• Assume a rate that is within previously accepted
  values.
• Measure a rate locally and then assume that it
  applies over the period/location of interest.
• Use some kind of an empirical parameterization,
  e.g. F k ( cb - ceq ) where k is
  determined from the empirical fit.
• Each of these will benefit from a solid
  conceptual and mechanistic model of the process.

10

• Typical Parameterizations
• from Nightingale et al. (2000) from Clark JF
  et al. online at http//www.ldeo.columbia.ed
  u/res/pi/Hudson/articles/article1b.html

11

• both plots from NOAA website http//www.aoml.noaa.
  gov/
• Testing

12

• Wanninkhof et al. 1992 Banerjee and MacIntyre,
  2004

13

• For a given wind speed there seems to be
  uncertainty by a factor of 3 to 4.
• Wind may not be the dominant forcing of the
  process at low wind speeds, and even at
  intermediate wind speeds there can exist
  complexities such as surfactants.
• Recent work has pointed to convection as a
  significant forcing.Eugster, W., G. Kling, et
  al. (2003). "CO2 exchange between air and water
  in an Arctic Alaskan and midlatitude Swiss lake
  Importance of convective mixing." Journal of
  Geophysical Research-Atmospheres 108(D12).
• McGillis, W. R., J. B. Edson, et al. (2004).
  "Air-sea CO2 exchange in the equatorial Pacific."
  Journal of Geophysical Research-Oceans 109(C8).

14

• Takahashi et al. (2002)

15

• What conceptual model will accurately describe
  all the complexity?
• How do we make the transition to large space and
  time scales?

16

• The most recent geophysical research is turning
  to surface divergence models of Soloviev and
  Schlussel (1994), Reynolds number models of
  Banerjee (1990), and surface divergence models of
  Chan and Scriven (1970) for conceptual models.
• The surface divergence models are probably the
  most accurate, but this has not rigorously been
  proven and they have not caught on in the
  geophysical community.

17

• Exactly what is the surface divergence model and
  why should we focus on it?

surface renewal surface divergence
18

• Where and How My Research Plan is Going To Make
  An Improvement
• What Exactly The Research Plan Is
• What Ive already done
• What I plan to do
• Major Impediments, Time Schedule, and Financial
  Situation of the Research Plan

19

• Advection-diffusion equation
• No sources/sinks on a time scale fast enough to
  be in this equation
• The equations are linear in c, and so we can
  recast them as
• Where cn is non-dimensionalized using
  (c-ceq)/(cb-ceq)

20

• If we take a typical air-water gas transfer rate
  (say 1x10-2 mol/m2/s from McGillis et al. 2004)
  and use the idea that advection is non-operative
  near the interface we see that 10-2 mol/m2/s F
  -D dcn/dx (cb-ceq) and so dcn/dx lt
  104 m
• So the concentration boundary layer is less than
  100 microns. But this close to the water
  interface the fluid motions are dominated by
  viscosity, since the viscous layer is
  10(viscosity)/(velocity scale) 10-3 m or
  about a centimeter
• This ensures that 1) in the concentration
  boundary layer we can model the fluid motion as
  stagnation flow, 2) we can neglect surface
  deformation in most locations

21

• The only physical scales are D, (cb-ceq), and
  we can non-dimensionalize the
  advection-diffusion equation to get
• The natural velocity scale that now arises is
• MATLAB demo Here is a 1-D example.
• Note that for steady state the solution is the
  error function in the variable zn/2 and therefore
  dcn/dzn0.5

22

• Turney et al. 2005

23

• 3.6 mps channel wind speed. 125 fps or 250 fps

24
(No Transcript)
25

• Tests of the surface divergence theory

• Banerjee et al. 2004
  Turney et al. 2005
  and McKenna and McGillis 2004

26

• My PhD Research Plan
• Objective 1 Test the surface divergence model
  under a variety of physical conditions.
  Characterize the time and space scales of the
  fluid motions. Determine which features of the
  wind waves are producing the important fluid
  motions. Search for a relationship between wave
  properties (i.e. morphology, dynamics) and
  important fluid motions. Assure the validity of
  the assumptions constituting the model.
• Objective 2 Connect the fluid motions that drive
  the surface divergence model with common
  meteorological observables.
• Objective 3 Connect the fluid motions that drive
  the surface divergence model with common
  satellite observables.

27

• Novel technique for imaging the surface
  divergence. Possibly useful for field studies.

• Dime under 3mm of water dime under 1mm of
  water
• Both are taken at wavelengths of 1300nm to 2500nm

28

• Unpublished figure, data that I took

29

• The experiments are described best in section 4
  of my qualification exam write up.
• Collect collocated images (with pixel resolution
  of O(.1mm)) of surface velocities and surface
  morphology at frame rates of O(100 fps)
• Independent variables are wind speed, wave
  characteristics, surface heat loss, surface force
  dynamics, water depth (unfortunately).

30

• Can we determine a gas transfer velocity from
  satellite observables?
• The necessary information is a relationship
  between surface morphology and fluid motions, and
  also a calculation of heat loss from the ocean
  surface?
• This sounds like a tall order, but the connection
  between surface morphology and satellite
  observables is well established. Satellite
  methods for calculating energy budgets at the
  ocean surface also exist.Gautier, C., Diak, G.,
  Masse, S., 1980 A Simple Model to Estimate the
  Incident Solar Radiation at the Surface from GOES
  Satellite Data. J. of Appl. Meter. 19 1005-1012.
• Researchers in Japan have already proposed a very
  similar idea for determining wind speed from
  satellite altimeters, scatterometers, and
  sun-glitter (essentially surface morphology). If
  this approach is valid, the extension to gas
  transfer coefficients should follow
  straight-forward.Ebuchi, N. and S. Kizu (2002).
  "Probability distribution of surface wave slope
  derived using sun glitter images from
  Geostationary Meteorological Satellite and
  surface vector winds from scatterometers."
  Journal of Oceanography 58(3) 477-486.
• Zhao, D. L. and Y. Toba (2003). "A spectral
  approach for determining altimeter wind speed
  model functions." Journal of Oceanography 59(2)
  235-244.

31

• Approach obtain gas transfer rate data from the
  Gas-Ex experiment, SOFeX experiment, or elsewhere
  and collocate it with remotely sensed data that
  gives mean square slope, significant wave height,
  surface irradiance, and surface temperature. Use
  the satellite data to calculate the wave age,
  maximum wave height, and also a surface heat
  flux.
• The likely remote sensing platforms will be
• Topex/Poseidon altimeter
• NSCAT or other scatterometer data
• Visible sun-glitter data
• GOES data for surface irradiance
• AVHRR for sea surface temperature
• Look at 5 to 10 cases. Can we estimate surface
  divergence from this data? If so, do we see
  agreement between model and measurements?

32

• Where and How My Research Plan is Going To Make
  An Improvement
• What Exactly The Research Plan Is
• What Ive already done
• What I plan to do
• Major Impediments, Time Schedule, and Financial
  Situation of the Research Plan

33

• Budget

Equipment Needed New Solenoid Switch for Water
Chiller Polyethylene Filters Tanks of
Nitrogen Tanks of SF6 Tanks of He Four Week
Rental of Infrared Camera Thermocouple
probe Silvered Particles Water heater and
insulation Humidity sensor Surface tensiometer (I
can borrow this) Topex/NSCAT/GOES/AVHRR Images??
Potential Money Sources Fannie and John Hertz
Foundation Fellowship NSF oceanography, Jan
15 EPA STAR Fellowship Air Waste Management
Association Robert and Patricia Switzer
Foundation Link Foundation Fellowship in Ocean
Engineering American Water Works Association NASA
Graduate Student Researchers Program NASA Earth
System Science Fellowships
34

• Time Table

35

• end

36

• Testing
• Template