Transcript Surface Exchange Processes

Air-Sea Exchange : 2
SOEE3410 : Lecture 5
Ian Brooks
Gas Fluxes
• Gas fluxes depend upon:
– Strength of mixing (u*, z0)
– Concentration in ocean
– Concentration in
atmosphere
– Molecular diffusion in
interfacial sub-layer.
• Concentration in ocean
depends upon:
– solubility of the gas in
water (temperature
dependent)
– Reactions with other
chemical species
– Biological processes
• Gas flux is driven by the
difference in partial
pressure of the gas
between atmosphere and
ocean.
• Direct measurement of
gas fluxes can be difficult:
– need for fast, accurate
measurement in hostile
environments.
– turbulent fluctuations often
very small compared to
mean value (eg CO2)
SOEE3410 : Atmosphere & Ocean Climate Change Processes
2
CO2 in Ocean Water
The concentration of CO2 dissolved in ocean waters is
dependent upon both chemical and biological processes.
– Dissolved (non-ionic) CO2 makes up only ~1% of Dissolved
Inorganic Carbon (DIC). This can be exchanged with CO2 in the
atmosphere until the (local) partial pressures in atmosphere and
ocean are equal.
– Bicarbonate ion, HCO3-, makes up 91% of DIC.
– Carbonate ion, CO32-, makes up 8% of DIC.
These three components are linked via the equilibrium of:
CO2 + H2O + CO32-  2HCO3-
SOEE3410 : Atmosphere & Ocean Climate Change Processes
3
The partial pressure of CO2 dissolved in water is strongly
temperature dependent, increasing with temperature
(solubility decreases). Thus regions of warming surface
waters tend to be regions of out-gassing, while cooling
waters are regions of uptake.
Regions of warming and cooling are linked by large-scale
ocean circulation, thus CO2 taken up in a cooling region
can be released back to the atmosphere at a later time in a
region of warming.
SOEE3410 : Atmosphere & Ocean Climate Change Processes
4
Dissolved inorganic carbon is removed from surface waters
via the production of CaCO3 by the formation of shells and
corals. These ultimately form sediments on the sea bed.
Ca2+ + CO32-  CaCO3
Note that the removal of a carbonate ion drives the balance
of DIC towards the production of dissolved non-ionic CO2,
increasing its partial pressure in water.
The formation of calcium carbonate thus both removes
some carbon from the global CO2 cycle, and locally tends
to force CO2 back into the atmosphere from the ocean.
SOEE3410 : Atmosphere & Ocean Climate Change Processes
5
• Direct measurements of pCO2 can be made only at spot
locations.
• Global ocean models successfully reproduce the main
features of ocean carbon content:
– Vertical gradient in DIC
– Spatial patterns of surface pCO2: outgassing in tropics, uptake at
higher latitudes
– Models that include marine biology (DOC, plankton dynamics)
also roughly reproduce seasonal cycle of surface pCO2,
atmospheric O2 and surface chlorophyll.
– Rough reproduction of phase and amplitude of interannual
variability in equatorial Pacific
SOEE3410 : Atmosphere & Ocean Climate Change Processes
6
• There remain many aspects of marine carbon cycle that
are not well simulated – poor understanding and
representation of physical or biological processes
– Spatial structure in deep ocean is poorly reproduced
– Largest discrepancies where there are fewest measurements
• Serious discrepancy in estimates of inter-hemisphere
transport: all ocean models suggest zero transport –
most observations suggest a transport of ~1 PgC/year!
– Some recent observations closer to models…still an open
question.
SOEE3410 : Atmosphere & Ocean Climate Change Processes
7
Estimated pCO2 (ocean-air)
for August and January. –ve
values imply uptake to
ocean, +ve values imply
outgassing to atmosphere.
SOEE3410 : Atmosphere & Ocean Climate Change Processes
8
Estimated annual air-sea CO2 flux
SOEE3410 : Atmosphere & Ocean Climate Change Processes
9
Gas fluxes are usually parameterized in terms of a transfer velocity, kgas,
Fgas  kgas C( air sea )
k gas
 Sc 
 kheat  
 Pr 
n
where, kheat, the transfer velocity for heat is given by…
k heat 
Fheat
c p T
Sc is the Schmidt number: the ratio of kinematic viscosity to the
molecular diffusivity of the gas (Sc = /D  660 for CO2), Pr is the Prandtl
number, the analageous quantity for heat (~7).
Note terminology. In previous lecture a generic transfer velocity denoted by UT. K is more common
symbol in gas transfer field
SOEE3410 : Atmosphere & Ocean Climate Change Processes
10
• The transfer velocity is a
function of external forcing
factors
– Usually defined as a function
of wind speed at 10 m (U10)(as
for heat, water vapour)
(remember Kheat = CDU10)
– This simple parameterization
cannot explain the observed
variability in measured gas
fluxes…other factors must be
important
Figure shows KCO2 values derived
from measurements at sea,
plotted against U10. Solid line is a
best fit to the data, dashed lines
are published parameterizations.
From: Frew et al. 2004: Air-sea gas transfer: its
dependence on wind stress, small-scale roughness and
surface films. JGR, 109, C08S17.
SOEE3410 : Atmosphere & Ocean Climate Change Processes
11
Various parameterizations of CO2 transfer velocity as a
function of wind speed. Note large scatter at high wind
speeds.
(Woolf, 2005: Parameterization of gas transfer velocities and sea-state-dependent wave
breaking. Tellus, 57B, 87-94.)
SOEE3410 : Atmosphere & Ocean Climate Change Processes
12
Additional factors shown to
affect gas transfer include:
– Wave fetch
– Boundary layer stability
– Presence of surfactants
(films of biologically
produced organic material)
– Bubble bursting
These all modify both the
surface small-scale wave
field and subsurface
turbulence, and hence the
gas transfer velocity.
Recent studies have shown
that microscale breaking
has a strong influence on
air-sea gas exchange.
Microscale breaking is
associated with waves of
wavelength ~0.1 to 0.5 m.
Small vortices generated
just behind the crest of the
breaking wave enhance
mixing just below the
surface, renewing the
surface sub-layer (i.e. the
layer of water exchanging
gas with the atmosphere)
SOEE3410 : Atmosphere & Ocean Climate Change Processes
13
• Fractional area coverage of
region of thermal disruption
by microbreakers, AB, is
correlated with transfer
velocities for heat (KH) and
a trace gas (KG)
• AB is correlated with mean
square slope – previously
shown to correlate with KG
• Microscale breaking
appeared to dominate gas
transfer at low-moderate
wind speeds
Zappa et al. 2004
Zappa et al. 2004: JGR,
doi:10.1029/2003JC001897
SOEE3410 : Atmosphere & Ocean Climate Change Processes
14
The mechanism by which wave
breaking enhances gas transfer is
still uncertain:
– Bubble mediated gas transfer:
the transfer of dissolved gases
between bubbles entrained
below water surface and
surrounding water is efficient
due to large surface to volume
ratio, and surface tension effects
mean the pressure inside the
bubble is greater than
hydrostatic pressure for its
depth  gas can super saturate.
On bursting at the surface seaair transfer is greatly enhanced.
– Rising plumes of bubbles
enhance turbulence in nearsurface layer, increasing mixing
on ocean side of interface
Gas transfer due to wave
breaking/bubble effects can be
parameterized in terms of the
fractional area of white-capping.
Factors that affect whitecap
fraction :
– Wind speed or wind stress (first
attempts at parameterization
depended only on U10)
– Wave height (a function of seastate…how well developed are
the waves…fetch dependent)
– Swell / waves propagating into
region
SOEE3410 : Atmosphere & Ocean Climate Change Processes
15
One of the most recent proposed
parameterizations (Woolf, 2005) is:
W  4.02107 RH0.96
RH 
u* H
a
W is the whitecap fraction, u* is the
friction velocity, H the significant wave
height, and a the kinematic viscosity
of air. RH is a form of Reynolds
number for wind-waves. The bubble
mediated gas transfer is expressed
as the sum of transfer velocities for
non-breaking (K0) and breaking
processes (Kb)
K gas  K0  Kb
Kb  850 W
K0  1.57104 u* (600/ Sc)0.5
Modelled fractional whitecap cover as a
function of wind speed and fetch (solid
lines). (Woolf, 2005)
The form of the breaking contribution is
based on theoretical considerations, with
constants derived from extensive
observations both in laboratory wave tanks
and in the field.
SOEE3410 : Atmosphere & Ocean Climate Change Processes
16
Models
LM : Liss & Merlivat (1986)
N : Nightingale et al. (2000)
W92 : Wanninkhof (1992)
WM : Wanninkhof & McGillis (1999)
Fetch (km):
10, 30, 100, 300, 1000
Woolf, 2005: Parameterization of gas transfer velocities and sea-state-dependent wave breaking.
Tellus, 57B, 87-94.
SOEE3410 : Atmosphere & Ocean Climate Change Processes
17
Although the basic processes that affect airsea gas transfer are known, the detailed
physics is only poorly understood. Most of
the processes have yet to be measured in
sufficient detail over a wide enough range of
conditions to parameterize them in a manner
usable in large-scale climate models.
SOEE3410 : Atmosphere & Ocean Climate Change Processes
18
Aerosol Fluxes
• Remote ocean regions have relatively clean air – low
aerosol concentrations – compared to continental air
masses. Remain important to climate because of very
large spatial extent.
• There are two primary sources of aerosol over the
oceans:
– Sea-salt : derived from evaporation of sea water droplets
• Droplets formed by bubble bursting and ‘spume’ production –
ripping of water droplet off wave crests by the wind
– Non-sea-salt sulphates : derived from the oxidation of
biologically produced gases, primarily from phytoplankton
SOEE3410 : Atmosphere & Ocean Climate Change Processes
19
Bubble Bursting
Bubbles injected into the ocean by
wave-breaking (U > 4ms-1) rise to
the surface and burst. The bubble
film breaks up into 100s or 1000s
of droplets <2m diameter.
Following this, the bubble cavity
collapses, as water flows in to fill
the cavity a jet forms at its centre –
this ejects a few (up to 7) larger
drops ~2-100m in radius.
A few 10s to 100 times as
many film drops as
jet drops form
SOEE3410 : Atmosphere & Ocean Climate Change Processes
20
Spume production – water droplets are ripped from wave
crests by the wind when U10 exceeds about 7 ms-1.
Droplets range from ~40m to ~1mm in diameter.
SOEE3410 : Atmosphere & Ocean Climate Change Processes
21
• Droplets immediately begin evaporating. Largest
droplets fall quickly back to surface and do not fully
evaporate; smaller droplets can completely evaporate
leaving behind a particle sea-salt, with mixed chemical
composition – various salts and organic material.
• Turbulence mixes the aerosol rapidly through the
boundary layer, where they act as efficient Cloud
Condensation Nuclei.
SOEE3410 : Atmosphere & Ocean Climate Change Processes
22
Courtesy of Gerrit deLeeuw
The uncertainty in sea spray aerosol generation as a function of wind speed is at
least a factor of 10. Measurement to date has used indirect methods. Only
recently has the technology to make direct measurements of the aerosol flux
become available.
SOEE3410 : Atmosphere & Ocean Climate Change Processes
23
U10 = 8 m s-1
U10 = 12 m s-1
SOEE3410 : Atmosphere & Ocean Climate Change Processes
24
MWS > 0.035 : undeveloped
0.03 < MWS < 0.035 : near developed
MWS < 0.03 : well developed
Mean fluxes (m-2 s-1 μm-1)
1.67×105 (undeveloped)
2.45×105
5.08×105 (well developed)
Mean fluxes (m-2 s-1 μm-1)
37.2 (undeveloped)
55.5
42.5 (well developed)
SOEE3410 : Atmosphere & Ocean Climate Change Processes
25
Biological Sources of Aerosol
Over the remote ocean biological processes are a
significant source of aerosol:
– Phytoplankton emit the gas dimethylsuphide (DMS)
– DMS is oxidized in the atmosphere via a complex chain of
reactions with the hydroxyl radical (OH•), ozone (O3), hydrogen
peroxide (H2O2),…to produce sulphuric acid, methane sulphonic
acid and sulphur dioxide. These gases can condense to form
aerosol particles (generically refered to as non-sea-saltsulphate) that eventually grow large enough to act as CCN.
• Newly formed nns-sulphate particles ~0.001m in size – too small
to act as CCN. Further condensational growth or heterogeneous
reactions required to produce larger particles.
SOEE3410 : Atmosphere & Ocean Climate Change Processes
26
Surface
Heat Flux
Atmospheric
Stability
Coherent
Structures
Wind
Direction
Wind Stress
Fetch
Surface
Films
Microbreaking
Waves
Near
Surface
Turbulence
Transfer
Velocity
k
Bubbles
Remote
Sensing
Air-Sea Gas Flux
F = kC
Surface
Films
Modelling
C
Biogeochemistry
Hydrography
Sea Surface
Temperature
SOEE3410 : Atmosphere & Ocean Climate Change Processes
27
SOEE3410 : Atmosphere & Ocean Climate Change Processes
28