Jacobson M.Z. Atmospheric Pollution

Подождите немного. Документ загружается.

288 ATMOSPHERIC POLLUTION: HISTORY, SCIENCE, AND REGULATION

his invention to the American Chemical Society in April 1930 by inhaling CFC-12,

then blowing it over a candle ﬂame, extinguishing the ﬂame. The invention was not

disclosed previously because the Frigidaire department of General Motors needed time

to ﬁle patents on a family of compounds related to the invention (Bhatti, 1999).

In 1931, CFC-12 was produced by the DuPont chemical manufacturer under the trade

name Freon, a name chose by Midgley and his assistants. Its ﬁrst use was in small ice

cream cabinets. In 1934, it was used in refrigerators and whole-room coolers. Soon after,

it was used in household and automotive air conditioning systems. In 1932, CFC-11

[CFCI

(g), trichlorﬂuoromethane], was ﬁrst produced. Its ﬁrst use was in large air condi-

tioning units. CFCs became airborne only when coolants leaked or were drained.

In 1943, Goodhue and Sulli

van of the U.S. Department of Agriculture developed a

method to use CFC-11 and -12 as a propellant in spray cans. CFCs ﬂowed out of a spray

can’s nozzle, carrying with them a mist containing other ingredients. Spray cans were

used to propel hair sprays, paints, deodorants, disinfectants, polishes, and insecticides.

CFC-11 and 12 have also been used as blowing agents in foam production. Foam

is used in insulation, disposable cups and cartons, and ﬁre extinguishers. CFCs are

released to the air during foam-production, itself. CFCs in the air spaces of foam are

usually conﬁned and not an important source of atmospheric CFCs.

Figure 11.11 shows the reported sales of CFC-11 and -12 in 1976 and 1998. The ﬁg-

ure shows that in 1976, almost 58 percent of CFC-11 and -12 were sold for use as

propellants in spray cans. Secondary use for CFC-11 was as a blowing agent and for

CFC-12 was as a refrigerant. Regulation of CFCs (Section 11.10) resulted in large

decreases in the use of CFC-11 and -12 between 1976 and 1998 and increases in replace-

ment compounds. By 1998, total sales of CFCs were about 6 percent of those in 1976.

Several other CFCs were developed, some of which are listed in Table 11.2. CFC-113,

for example, was ﬁrst produced in 1934 and used in air conditioning units. It has since

been used primarily as a solvent in the microelectronics industry and in the dry-cleaning

industry. It has also been used as a spray-can propellant and a blowing agent in foam

production.

100

200

300

400

500

600

700

1976 1998 1976 1998

Total

Propellant

Blowing agent

Refrigerant

Other

CFC sales (1,000 metric tons/yr)

Year

CFC-12CFC-11

57.4%

28.5%

7.9%

57.6%

6.2%

11.0%

25.9%

5.5%

57.7%

3.7%

31.1%

7.5%

8.9%

3.6%

80.8%

6.7%

Figure 11.11. Reported sales of CFC-11 and -12 in 1976 and 1998. Percentages are of the

total for the year.

Source: AFEAS (2001).

11.5.1.2. Other Chlorine Compounds

Chloroﬂuorocarbons are a subset of chlorocarbons, which are compounds con-

taining carbon and chlorine. Hydrochloroﬂuorocarbons (HCFCs) are another subset

of chlorocarbons. HCFCs are similar to CFCs except that HCFCs have at least one

hydrogen atom. The hydrogen atom allows HCFCs to be broken down in the tropo-

sphere by reaction with OH(g). OH(g) does not readily break down CFCs. Because

HCFCs break down more readily than do CFCs, a smaller percentage of emitted

HCFCs than CFCs reaches the stratosphere. Nevertheless, because HCFCs contain

chlorine and some HCFCs reach the stratosphere, HCFCs are still a danger to stratos-

pheric ozone. HCFC-22, ﬁrst produced in 1943, is the most abundant HCFC in the air

today. HCFC-22 has been used as a refrigerant, spray-can propellant, and blowing

agent in foam production.

Other chlorocarbons include carbon tetrachloride [CCl

(g)], methyl chloroform

[CH

CCl

(g)], and methyl chloride [CH

Cl(g)]. Carbon tetrachloride is used as an

intermediate in the production of CFCs and HCFCs and as a solv

ent and grain fumi-

gant. Methyl chloroform is used as a degreasing agent, a dry-cleaning solvent and an

industrial solvent. Methyl chloride is produced synthetically only in small quantities

and used in the production of silicones and tetramethyl lead intermediates (Singh,

1995). Most methyl chloride in the air is produced biogenically in the oceans.

Another chlorine-containing gas in the troposphere is hydrochloric acid

[HCl(g)]. HCl(g) has larger natural than anthropogenic sources. Natural sources

include evaporation of chloride from sea-spray and volcanic emissions. Although

some anthropogenic emissions of HCl(g) are from waste incineration, about 98 per-

cent are from coal combustion (Saxena et al. 1993).

11.5.1.3. Bromine Compounds

Although chlorine-containing compounds are more abundant than are bromine-

containing compounds, the latter compounds are more efﬁcient, molecule for molecule, at

destroying ozone. The primary source of stratospheric bromine is methyl bromide

[CH

Br(g)], which is produced biogenically in the oceans and emitted as a soil fumigant.

Other sources of bromine are a group of synthetically produced compounds termed

Halons, which are used in ﬁre extinguishers and as fumigants. The most common Halons

are H-1301 [CF

Br(g)], H-1211 [CF

ClBr(g)], and H-2402 [CF

BrCF

Br(g)]. Methyl

bromide and Halons are bromocarbons because they contain both bromine and carbon.

11.5.1.4. Fluorine Compounds

Compounds that contain hydrogen, ﬂuorine, and carbon but not chlorine or

bromine are hydroﬂuorocarbons (HFCs). HFCs were produced in abundance only

recently as a replacement for CFCs and HCFCs. Because the ﬂuorine in HFCs has lit-

tle effect on ozone, production of HFCs may increase in the future. Unfortunately,

because they absorb thermal-IR radiation, HFCs will enhance global warming if their

use increases. The most abundantly emitted HFC to date has been HFC-134a

[CH

FCF

(g)]. Related to HFCs are perﬂuorcarbons (PFCs), such as perﬂuoroethane

(g)], and sulfur hexaﬂuoride [SF

(g)].

11.5.2. Lifetimes and Mixing Ratios of Chlorinated Compounds

Once emitted, CFCs take about one year to mix up to the tropopause. Because they are

chemically unreactive and cannot be broken down by solar wavelengths that reach the

GLOBAL STRATOSPHERIC OZONE REDUCTION 289

troposphere, CFCs are not removed chemically from the troposphere. Instead, they

become well mixed in the troposphere and slowly penetrate to the stratosphere. Today, the

tropospheric mixing ratios of CFC-11 and CFC-12, the

two most abundant CFCs, are about 270 and 550 pptv,

respectively (Table 11.2 and Fig. 11.12).

11.5.2.1. Lifetimes of CFCs

Because the stratosphere is one large temperature

inversion, vertical transport of ozone through it is slow.

About 10 Mt of chlorine in the form of CFCs reside in

the troposphere, and the transfer rate of CFC-chlorine

from the troposphere to the middle stratosphere is about

0.1 Mt per year. In this simpliﬁed scenario, the average

time required for the transfer of a CFC molecule from

the troposphere to the middle stratosphere is about 100

years.

CFCs are broken down in the stratosphere only

when they are exposed to far-UV radiation (wave-

lengths of 0.01 to 0.25 m), and this exposure occurs at

an altitude of 12 to 20 km and higher

. At such altitudes,

far-UV wavelengths photolyze CFC-11 and CFC-12 by

CFCl

(g)  h C

•

FCl

(g)  C

•

l(g)   250 nm

CFC-11 Dichloroﬂuoro- Atomic (11.21)

methyl radical chlorine

(g)  h C

•

Cl(g)  C

•

l(g)   230 nm

CFC-12 Dichloroﬂuoro- Atomic (11.22)

methyl radical chlorine

At 25 km, e-folding lifetimes of CFC-11 and CFC-12 against photolysis under maximum-

sunlight conditions are on the order of 23 and 251 days, respectively. Average lifetimes

are on the order of two to three times these values.

In sum, the limiting factor in CFC decomposition in the stratosphere is not trans-

ported from the surface to the tropopause or photochemical breakdown in the

stratosphere, but transported from the tropopause to the middle stratosphere. Table

11.2 indicates that the overall lifetimes of CFC-11 and CFC-12 between release at the

surface and destruction in the middle stratosphere are about 55 and 116 years, respec-

tively. The lifetime of CFC-12 is longer than that of CFC-11, partly because the

former compound must climb to a higher altitude in the stratosphere before breaking

apart than must the latter. Because of their long overall lifetimes, some CFCs emitted

in the 1930s through 1950s are still present in the stratosphere. Those emitted today

are likely to remain in the air until the second half of the twenty-ﬁrst century.

11.5.2.2. Lifetimes of Non-CFCs

Lifetimes of non-CFC chlorinated compounds are often shorter than are those of

CFCs. The lifetimes of CCl

(g), HCFC-22(g), CH

CCl

(g), CH

Cl(g), and HCl(g)

between emission and chemical destruction are about 35, 12, 5, 1.3, and less than 0.1

year, respectively. Non-CFCs generally have shorter lifetimes than do CFCs because

290 ATMOSPHERIC POLLUTION: HISTORY, SCIENCE, AND REGULATION

0 100 200 300 400 500 600

Altitude (km)

CFC-11

HCFC-22

CFC-12

CCl

(g)

Mixing ratio (pptv)

Tropopause

Figure 11.12. Variation of CFC-11, CFC-12,

HCFC-22, and CCl

(g) with altitude at 30N

latitude. Smoothed and scaled from

Jackman et al. (1996) to present-day near-

surface mixing ratios.

they react faster with OH(g) than do CFCs and are often more water soluble than are

CFCs. The beneﬁt of a shorter lifetime for a chlorine-containing compounds is that, if

breakdown occurs in the troposphere, the chlorine released can be converted to

HCl(g), which is highly soluble and can be removed readily by rainout. Because the

stratosphere does not contain clouds, except for ice-containing clouds that form sea-

sonally over the poles, HCl(g) cannot be removed from the stratosphere by rainout.

Some non-CFCs, such as HCFC-22, photolyze slower than do CFCs, so once HCFC-

22 reaches the middle stratosphere, its concentration builds up there to a greater extent

than do concentrations of several CFCs, as seen in Fig. 11.12.

Of non-CFC chlorine compounds, CH

Cl(g), and HCl(g) have the largest natural

sources. The tropospheric e-folding chemical lifetime of CH

Cl(g) against reaction by

OH(g) is about 1.5 years; that of HCl(g) against reaction by OH(g) is about 15 to 30

days. HCl(g) is also soluble in w

ater and is absorbed by clouds. Volcanos,

which emit

water vapor and hydrochloric acid, produce clouds and rain that remove HCl(g), pre-

venting most of it from reaching the stratosphere (Lazrus et al., 1979; Pinto et al.,

1989; Tabazadeh and Turco, 1993). The facts that the two major natural sources of

chlorine, CH

Cl(g) and HCl(g), have short chemical lifetimes against destruction by

OH(g) and that HCl(g) is soluble in water, whereas CFCs have long chemical lifetimes

and are insoluble, support the contention that CFCs and not naturally emitted chlorine

compounds are responsible for most ozone destruction in the stratosphere.

11.5.2.3. Emissions of Chlorine Compounds to the Stratosphere

Table 11.3 summarizes the relative emissions of anthropogenic and natural chlorine-

containing compounds into the stratosphere in 1994. About 82 percent of chlorine

entering the stratosphere originated from anthropogenic sources. Of the remainder, about

15 percent was methyl chloride, emitted almost exclusively by biogenic sources in the

oceans, and 3 percent was hydrochloric acid, emitted by volcanos, evaporated from sea

spray, and otherwise produced naturally. The relatively large anthropogenic versus natural

GLOBAL STRATOSPHERIC OZONE REDUCTION 291

Table 11.3. Relative Emissions of Selected Chlorine

Compounds into the Stratosphere

CFC-12 CF

(g) 28

CFC-11 CFCl

(g) 23

Carbon tetrachloride CCl

(g) 12

Methyl chloroform CH

CCl

(g) 10

CFC-113 CFCl

Cl(g) 6

HCFC-22 CF

ClH(g) 3

Contribution to

Trade Name or Stratospheric

Chemical Name Chemical Formula Emissions (Percent)

Source: WMO (1995).

Methyl chloride CH

Cl(g) 15

Hydrochloric acid HCl(g) 3

Total 100

Anthropogenic Sources

Natural Sources

source of chlorine into the stratosphere supports the contention that stratospheric ozone

reductions result primarily from anthropogenic chlorine, not natural chlorine.

11.5.3. Catalytic Ozone Destruction by Chlorine

Once released from their parent compounds in the stratosphere, chlorine atoms from

CFCs and non-CFCs react along one of several pathways. Chlorine reacts in a catalytic

ozone destruction cycle,

At midlatitudes, the chain length of this cycle increases from about 10 in the lower

stratosphere to about 1,000 in the middle and upper stratosphere (Lary, 1997).

The primary removal mechanisms of active chlorine [Cl(g)ClO(g)] from the

catalytic cycle are reactions that produce chlorine reservoirs, HCl(g) and chlorine

nitrate [ClONO

(g)]. Chlorine reservoirs are called such because they temporarily

store active chlorine, preventing it from destroying ozone. Conversion of Cl(g) to

HCl(g) occurs by

(11.26)

Conversion of ClO(g) to ClONO

(g) occurs by

ClO

•

(g)  N

•

(g)

ClONO

(g)

Chlorine Nitrogen Chlorine (11.27)

monoxide dioxide nitrate

(g)

Methane

(g)

Hydroperoxy

radical

(g)

Molecular

oxygen

(g)

Molecular

hydrogen

(g)

Hydrogen

peroxide

Cl(g) +

Atomic

chlorine

HCl(g) +

(g)

Methyl

radical

H(g)

Atomic

hydrogen

(g)

Hydroperoxy

radical

Hydrochloric

acid

292 ATMOSPHERIC POLLUTION: HISTORY, SCIENCE, AND REGULATION

•

l(g)  O

(g) ClO

•

(g)  O

(g)

Atomic Ozone Chlorine Molecular

chlorine monoxide oxygen

(11.23)

•

lO(g)  •O

•

(g) C

•

l(g)  O

(g)

Chlorine Ground- Atomic Molecular

monoxide state atomic chlorine oxygen

oxygen

(11.24)

•O

•

(g)  O

(g) 2O

(g)

Ground- Ozone Molecular

(11.25)

state atomic oxygen

oxygen

(net process)

At any time, about 1 percent of the non-CFC chlorine in the stratosphere is in the form

of active chlorine. Most of the rest is in the form of a chlorine reservoir. Because CFCs

release their chlorine by photolysis in the middle and upper stratosphere, it follows

that HCl(g) mixing ratios should also peak in the middle and upper stratosphere.

Indeed, observations conﬁrm this supposition.

The HCl(g) reservoir leaks back to atomic chlorine by photolysis, reaction with

OH(g), and reaction with O(g), all of which are slow processes. The e-folding lifetime

of HCl(g) against photolysis, for example, is about 1.5 years at 25 km. HCl(g) also dif-

fuses back to the troposphere, where it can be absorbed by clouds. The ClONO

(g)

reservoir leaks back to atomic chlorine by photolysis with an e-folding lifetime of

about 4.5 hours at 25 km.

11.6. EFFECTS OF BROMINE ON GLOBAL OZONE REDUCTION

Like chlorine, bromine affects stratospheric ozone. The primary source of stratospher-

ic bromine is methyl bromide [CH

Br(g)], which is produced biogenically in the

oceans and emitted as a soil fumigant. Other sources of bromine are Halons, deﬁned in

Section 11.5.1. The tropospheric mixing ratios of the most common Halons,

ClBr(g) (H-1211) and CF

Br(g) (H-1301), are both about 2 pptv, less than 1 per-

cent of the mixing ratios of CFC-11 and -12. Nevertheless, the efﬁciency of ozone

destruction by the bromine catalytic cycle is greater than is that by the chlorine cat-

alytic cycle.

Methyl bromide and Halons photolyze in the stratosphere to produce atomic

bromine. Photolysis of methyl bromide occurs above 20 km by

Br(g)  h C

•

(g)  B

•

r(g)   260 nm

Methyl Methyl Atomic (11.28)

bromide radical bromine

The e-folding lifetime of CH

Br(g) against loss by this reaction is about 10 days at

25 km.

Once atomic bromine is in the stratosphere, it reacts in another catalytic ozone

destruction cycle,

GLOBAL STRATOSPHERIC OZONE REDUCTION 293

•

r(g)  O

(g) BrO

•

(g)  O

(g)

Atomic Ozone Bromine Molecular (11.29)

bromine monoxide oxygen

BrO

•

(g)  •O

•

(g) B

•

r(g)  O

(g)

Bromine Ground- Atomic Molecular

monoxide state atomic bromine oxygen

oxygen

(11.30)

•O

•

(g)  O

(g) 2O

(g)

Ground- Ozone Molecular

state atomic oxygen

oxygen

(net process)

(11.31)

The chain length of this cycle increases from about 100 at 20 km to about 10

at 40 to

50 km (Lary, 1997). The chain length of the bromine catalytic cycle is longer than is

that of the chlorine catalytic cycle because Br(g) is removed more slowly from the

bromine cycle by reaction with CH

(g) and H

(g) than Cl(g) is removed from the chlo-

rine cycle by reaction with the same chemicals.

When atomic bromine is removed from its catalytic cycle, it forms hydrobromic

acid [HBr(g)] by

(11.32)

When BrO(g) is removed, it forms bromine nitrate [BrONO

(g)] by

BrO

•

(g)  N

•

(g)

BrONO

(g)

Bromine Nitrogen Bromine (11.33)

monoxide

dioxide nitrate

The HBr(g) reservoir leaks slowly back to atomic bromine by reaction with

OH(g). The BrONO

(g) reservoir quickly leaks back to atomic bromine by photol-

ysis. The e-folding lifetime of BrONO

(g) against photolysis is about 10 minutes

at 25 km.

11.7. REGENERATION RATES OF STRATOSPHERIC OZONE

The presence of chlorine and bromine has decreased levels of ozone in the stratosphere.

If chlorine and bromine could be removed easily from the stratosphere, the stratospheric

ozone layer could regenerate quickly. The problem is that the overall lifetimes of cer-

tain chlorocarbons and bromocarbons are on the order of 50 to 100 years; thus, the

natural removal rate of chlorine- and bromine-containing compounds from the strato-

sphere is slow.

Suppose that all ozone in the stratosphere were destroyed, all ozone-destroying

compounds were removed, but all oxygen remained. How long would the ozone layer

take to regenerate? An estimate can be obtained from Fig. 11.13. This ﬁgure shows

results from two computer simulations of the global atmosphere in which all ozone in

the present-day atmosphere was initially removed, but oxygen was not. In the ﬁrst sim-

ulation, ozone regeneration was simulated in the absence of chlorine and bromine. In

the second, ozone regeneration was simulated in the presence of 1989 concentrations

of chlorine, but in the absence of bromine. In both simulations, the globally averaged

column abundance of ozone regenerated to relatively steady values in less than a year.

Regeneration during the simulation in which chlorine was initially present was about 2

to 3 percent less than that during the no-chlorine case, consistent with the estimated

global reduction in ozone of 2 to 3 percent between the 1970s and 1989 due to chlorine-

containing compounds.

Atomic

bromine

Br(g) +

HBr(g) +

Hydrobromic

acid

(g)

Hydroperoxy

radical

(g)

Hydrogen

peroxide

(g)

Hydroperoxy

radical

(g)

Molecular

oxygen

294 ATMOSPHERIC POLLUTION: HISTORY, SCIENCE, AND REGULATION

11.8. ANTARCTIC OZONE DEPLETION

Every September through November since 1980, the minimum ozone column abun-

dance over the Antarctic decreases below its yearly average. Figure 11.14 shows that

in 2000, the lowest measured column abundance over the Antarctic was about 94 DU

(occurring on September 29,

2000), which was 68 percent less than 293.4 DU, the

globally (90S to 90N) and year-2000 averaged ozone column abundance. Between

1981 and 2000, the area over which ozone depletes (the Antarctic ozone hole)

increased, as shown in the ﬁgure. The ozone hole area is deﬁned as the area of the

globe over which the ozone column abundance decreases below 220 DUs. The ozone

hole in 2000 covered nearly 30  10

, an area larger than the size of North

America. Most Antarctic ozone depletion occurs between 14 and 18 km in altitude.

Beginning in 1992, springtime ozone decreases were observed up to 24 km and down

to 12 to 14 km.

Figure 11.15 shows the extent of the Antarctic ozone hole on October 1, 2000.

Whereas the minimum on that date was near 94 DU within the polar vortex, an ozone

maximum of 478 DU was observed just outside the vortex. The proximity of the max-

imum to the minimum is one factor that allows the ozone hole to replenish itself from

October to December.

The Antarctic ozone hole now appears every year during the Southern

Hemisphere spring. A smaller Arctic ozone dent (reduction in ozone to 240 to 260

DU) appears during the Northern Hemisphere late winter and spring

(March–May). In 1997, when ozone reductions over the Arctic were the most

severe on record, the minimum ozone value in the dent on April 1 was only 247

DU, appearing at 78.5N, 5.625W. This minimum was not low enough to qualify

as a depletion (less than 220 DU). Nevertheless, the average ozone column abun-

dance 60 to 90N in March 1997 was 22 percent lower than that in March 1979

[Fig. 11.9(a)], indicating that Arctic loss in 1997 was nontrivial. The ozone dent

seemed to disappear in 1999, when the ozone column abundance 60 to 90N in

GLOBAL STRATOSPHERIC OZONE REDUCTION 295

100

150

200

250

300

350

0 100 200 300 400

Avgerage global ozone column

(DU)

Day and date of simulation

10/1 1/7 4/17 7/26 11/4

No chlorine

With chlorine

Figure 11.13. Change in ozone column abundance, averaged over the globe, during two global

model simulations in which chlorine was present and absent, respectively. In both cases,

ozone was removed initially from the model atmosphere on October 1, 1988. Bromine was

not included in either simulation.

Source: Jacobson (1999a).

March was only 2.2 percent lower than it was in March 1979. The dent reappeared

in March 2000, when the column abundance was 16 percent lower than that in

1979. The hole and dent are caused by set of interlinked factors. One factor linking

global ozone reductions to polar ozone depletion is the presence of chlorine and

bromine in the stratosphere.

296 ATMOSPHERIC POLLUTION: HISTORY, SCIENCE, AND REGULATION

-90

-60

-30

-180 -120 -60 600 120 180

Latitude (degrees)

Longitude (degrees)

50 100 150 200 250 300 350 400 450 500

Ozone column abundance (DU)

Figure 11.15. Ozone column abundance (in DU) on October 1, 2000. The ﬁgure shows an

ozone hole over the Antarctic (with a minimum value of 94 DU at 86.5S, 64.4W) and an

ozone maximum of 478 DU at 58.5S, 83.125E. No data were available over the North Pole.

Data were obtained from the Total Ozone Mapping Spectrometer (TOMS) satellite and were

made available by NASA Goddard Space Flight Center, Greenbelt, Maryland.

100

150

200

250

300

1980 1985 1990 1995 2000

Ozone minimum (DU)

Ozone hole area (10

)

Year

Area of North America

Area of Antarctic

continent

Figure 11.14. Minimum ozone column abundances and areal extent of the ozone hole over

the Antarctic region from 1979 to 2000. Data from NASA Goddard Space Flight Center. For

comparison, the area of the Antarctic is about 13  10

and the area of North America

is about 24  10

GLOBAL STRATOSPHERIC OZONE REDUCTION 297

Figure 11.16. Polar stratospheric clouds, photographed in spring 2000 in the Arctic. Courtesy

NASA.

11.8.1. Polar Stratospheric Cloud Formation

The ozone hole over the Antarctic appears in part because the

Antarctic winter

(June–September) is very cold. Temperatures are low because much of the polar

region is exposed to 24 hours of darkness each day during the winter, and a wind sys-

tem, the polar vortex, circles the Antarctic. The vortex is a polar front jet-stream wind

system that ﬂows around the Antarctic continent, trapping cold air within the polar

region and preventing an inﬂux of warm air from outside this region.

Because temperatures are low in the Antarctic stratosphere, optically thin clouds,

called polar stratospheric clouds (PSCs) form (Fig. 11.16). These clouds have few

particles per unit volume of air in comparison with tropospheric clouds. Two major

types of clouds form. When temperatures drop to below about 195 K, nitric acid and

water vapor grow on small sulfuric acid–water aerosol particles (Toon et al. 1986).

Initially, it was thought that nitric acid and water molecules deposited to the ice

phase in the ratio 1:3. Such ice crystals have the composition HNO

• 3H

O and

are called nitric acid trihydrate (NAT) crystals. More recently, it was found that

these particles contain a variety of phases. Some contain nitric acid dihydrate

(NAD) (Worsnop et al. 1993), and others contain supercooled liquid water (liquid

water present at temperatures below the freezing point of water), sulfuric acid, and

nitric acid (Tabazadeh et al. 1994). Together, nitrate containing cloud particles that

form at temperatures below about 195 K in the winter polar stratosphere are called

Type I polar stratospheric clouds.

When temperatures drop below the frost point of water, which is about 187 K for

typical polar stratospheric conditions, a second type of cloud forms. These clouds con-

tain pure water ice and are Type II polar stratospheric clouds. Usually, about 90

percent of PSCs are Type I and 10 percent are Type II (Turco et al. 1989). Typical