Meteorology 465- The Middle
Atmosphere
The Ozone Hole.
Some
history of the ozone hole.
In
1984, Joe Farman and the British Antarctic Survey noticed yet another year
where the springtime ozone was lower than what they had usually measured,
starting about 1997. They submitted a
paper to Nature, and it was published in 1985.
Look
at the webpages and the chapter on Stratospheric Chemistry to find examples.
Ozone
is supposed to have a 100-year lifetime in the polar lower stratosphere. This rapid ozone loss in the lower
stratosphere really took everyone by surprise.
The UARS satellite was being built at the time to examine the
possibility of ozone loss by chlorine, but at 40 km, when chlorine catalysis is
the major catalytic cycle.
Many
mechanisms were proposed, usually according to the scientists’ specialties:
·
Dynamicists: upward motion of ozone poor air into the
lower stratosphere (N2O was small.)
·
Ionospheric
scientists: ions spiraling into the
polar region, causing ozone loss (NOx was very small.)
·
National
Inquirer: Aliens stealing our ozone (No
evidence to disprove.)
·
atmospheric
chemists: A new chlorine chemistry
(Never seen before.)
Chemists
immediately zeroed in on chlorine as being the culprit because it is the only
chemical that has changed substantially over the last 50 years.
Problems:
1.
Normal
chlorine chemistry won’t work because at that time of year in that location O
atoms are not produced.
2.
Most
of the chlorine would be in the forms of HCl and ClONO2 (a
partitioning problem). No mechanism was
known to shift HCl and ClONO2 to ClO and then keep it there.
3.
The
polar air would need to be mostly isolated from the midlatitude air so that
these processes could have the time to work.
Both
of these problems had to be solved.
Let’s
start with the second of the three problems.
It was well known that every winter in the polar lower stratosphere, the
temperature drops to ~190 K, more in the Antarctic than the Arctic (show
figures). At these temperatures,
the ~6 ppmv of water vapor condenses on existing sulfate aerosols, which are
composed of sulfuric acid and water. As
temperature decreases, these aerosols take on water.
Once
the temperature drops to ~200 K, the particles begin taking on nitric acid as
well as water. At around 195 K, these
particles are either super-cooled liquids or frozen, and have a size of about 1
micon. A large percentage of the nitric
acid can be taken up this way.
If
the temperature drops to ~190 K, significant water can accumulate on the
particles, making them large enough to fall kilometers in days – weeks.
Effects
of PSCs on chlorine chemistry.
Recall
the number of molecules striking a surface persecond is given by F = ¼ <v>A [Molecule], where A is the surface
area density (mm2/cm3). Typically, A is 1 mm/cm3 in the lower stratosphere,
but for PSCs, it can be 10 times larger.
A
reaction does not occur for every collision.
The ratio of reactions (or sticking) to collisions is usually denoted as
g.
First order rate = ¼ g<v>A.
Several
important heterogeneous reactions occur:
NO2 + O3 ® NO3 + O2
NO3 + NO2 + M ® N2O5 + M
This
occurs in the gas-phase. N2O5
is a nighttime reservoir with a lifetime of hours to day in sunlight.
However,
a heterogeneous reaction can occur:
N2O5 (g) + H2O (s) ® 2 HNO3 (s)
The
reaction coefficient, g, is about 0.06. The typical aerosols surface area density is
0.6 to 1 mm2 /cm3. If <v>ClONO2 ~ 200 m s-1,
then the first order rate coefficient for this process is about ¼ x 200x102
x 0.06 x .6x10-8 ~ 2x10-6 s-1.
·
HNO3(g)
® HNO3 (s)
·
HCl
(s) + ClONO2 (g) ® Cl2 (g) + HNO3 (s)
·
ClONO2
(g) + H2O (s) ® HOCl (g) + HNO3
(s)
·
HOCl
(g) + HCl (s) ® Cl2 + H2O
(s)
The
reaction coefficients are strongly dependent on both temperature and
composition of the aerosol.
Chlorine
species are liberated from their reservoirs and converted into gas-phase Cl2;
This can be nearly 100%. Nitrogen
species are converted into HNO3 and sequestered onto the aerosol
particles. If these particles are big,
enough, they fall out of the stratosphere, taking nitrogen species with them.
Cl2
is not reactive with O3.
However, it is easily photolyzed into two Cl atoms with near visible
sunlight (not UV). As a result, when
the airmass with Cl2 is illuminated, we have ppb levels of Cl.
What would Cl do?
Still
we do not have any O to speak of. We
need other reactions to destroy ozone.
Molina
and Molina proposed the following:
ClO + ClO + M ® ClOOCl + M
ClOOCl + hv ® Cl + ClOO
ClOO + M ® Cl + O2 + M
2{Cl + O3 ® ClO + O2}
net: 2O3 + hv ® 3 O2
It
turns out that ClOOCl also thermally decomposes, but only significantly at
temperatures above ~ 210 K, well above the temperature of the springtime polar
vortices.
Can
we explain the ozone loss by this mechanism?
Do the calculation.
d[O3]/dt = - 2 kClO+ClO{[ClO]}2
Go to near the peak in ozone at 50 hPa. Assume T = 195 K. [M] = 2x1018 cm-3 .
cClO = 1.3 ppbv; cO3 = 2 ppmv initially
kClO+ClO = 1.3 x 10-13 cm3
molecule-1 s-1
d[O3]/dt = - 2 1.3x10-13
(2.6x109)2 = - 2x106 moelcules cm-3
s-1
If we assume that each day is 40,000 long, then 8x1010
molec cm-3 are lost each day.
For [O3] = 4 x 1012 intially, it takes 50 days. Ozone loss starts in August and continues
through September into October. We have
about 50 days.
We
have a second reaction:
ClO + BrO ® BrCl + O2
® Br + ClOO
® Br + Cl + O2
Br + O3® BrO + O2
Cl + O3 ® ClO + O2
net: 2 O3 ® 3 O2
For these conditions, kClO+BrO = 1.3x10-12
cm3 moelcule-1 s-1. for 7 pptv of BrO, the O3 loss rate is about 1x105
molecules cm-3 s-1, or about 10% of ClO+ClO for this much
chlorine.
A
plot of ClO and ozone shows this anti-correlation that develops over time.
We
can summarize the reactions with a single plot, or two.
We
have a number of issues to resolve:
1. reactive
chloine becomes essentially all of the available chlorine, around 1 to 1.5 ppbv
2. NO and NO2
go into HNO3 and HNO3 goes into PSCs.
3. If PSCs
fall out, < 10% of the NOy can be left. Water vapor can also be reduced.
4. If PSCs do not fall out, eventually they will re-evaporate into
the air. This HNO3 can photolyze
to produce OH and NO2. Then
almost immediately, ClO + NO2 + M ® ClONO2 + M. The balance with HCl takes longer to
re-establish.
5. Frequent PSC events over the course of a spring might be
necessary for large ozone loss. So the
vortex must remain relatively isolated and cold into well into the spring for
ozone loss to be large.
6. Differences between the Arctic and Antarctic are the temperatures
and the length of time in the spring that the vortex is intact. See figure.
Other low temperature
regions?
The
processes that cause the rapid ozone loss made scientists think about what
might be happening globally. Look for
regions of low temperature.
However,
they were aware of another reaction that had implications:
NO2 + O3 ® NO3 + O2
NO3 + NO2 + M ® N2O5 + M
This
occurs in the gas-phase. N2O5
is a nighttime reservoir with a lifetime of hours to day in sunlight.
However,
a heterogeneous reaction can occur:
N2O5 (g) + H2O (s) ® 2 HNO3 (s)
The
reaction coefficient, g, is about 0.06. The typical aerosols surface area density is
0.6 to 1 mm2 /cm3. If <v>ClONO2 ~ 200 m s-1,
then the first order rate coefficient for this process is about ¼ x 200x102
x 0.06 x .6x10-8 ~ 2x10-6 s-1.
This
reaction has a significant impact on the nitrogen species in the lower
stratosphere. NOx is reduced
by a factor of 4, or more.
Two
questions:
1. How does
this influence ozone loss?
2. What do
you think a volcano would do?