1. Introduction

Ground-based Fourier spectroscopy is a powerful tool for carrying out routine monitoring of trace gases in the stratosphere, but it is known that inaccuracies in the knowledge of the instrument line shape of Fourier spectrometers can lead to errors in retrieved data.

This thesis describes the results of an investigation of the type and extent of the errors in retrieved vertical profiles that may be expected to arise as a result of common line shape defects (chapter 4).

The development of a portable laser system for measuring instrument line shapes, simultaneously with the acquisition of atmospheric and low pressure cell spectra, is reported (chapter 5) and the line shape measurements obtained using this equipment are presented (chapter 6) and discussed (chapter 7), and recommendations made for further work (chapter 8).

The importance of stratospheric ozone to Earth shall be briefly considered below, followed by an overview of the problem of stratospheric ozone loss in polar regions and at mid latitudes, and a brief discussion of the capabilities of Fourier spectrometry as a means of making ground-based measurements of the stratosphere.

1.1 Large Scale Circulation

The subject of this section is covered in greater depth in the review of stratosphere-troposphere exchange by Holton et al., [1995] and references therein.

1.1.1 Mid Latitude Wave Driven Circulation

That the large scale atmospheric circulation in the atmosphere is not driven by radiative processes alone has been known for some time. Observations of the stratosphere show that, especially at the winter pole, temperatures are substantially elevated with respect to the radiative equilibrium values (Andrews et al., [1987]) suggested by models such as those reviewed by Fels (Fels [1985]).

In the late sixties Dickinson (Dickinson [1968], [1969]) showed that in certain cases this discrepancy could be the result of dynamical factors. Dickinson showed that if planetary wave activity was included in circulation calculations this could result in air being forced down and north at a greater rate than would result from the radiative effects alone, with the result that the polar air temperatures would be warmer that than previous predictions, and hence closer to the observed values.

The importance of wave activity was emphasised by Haynes et al. in the formulation of their ‘downward control’ principle (Haynes et al. [1991]) which may be qualitatively summarised as follows.

Fundamental to the theory of mid latitude wave driven circulation is the fact that as the Earth is a rapidly rotating sphere there will be a gradient in the angular momentum of the atmosphere between the equator and the poles; this has its greatest rate of change occurring in mid latitudes. Gravity waves and Rossby waves¹, which have their source in the troposphere in the mid latitudes, propagate into the stratosphere where, if their amplitude is sufficiently great, they will break introducing a net drag on the mean flow. The effect of this retardation on the mean flow in an angular momentum gradient is to force the flow down and polewards while drawing air up to mid latitudes from the tropics. Air forced to high latitudes by this wave breaking will cool and subside at the poles but it is important to note that although the subsidence is associated with radiative cooling this is not the reason for it. Similarly, the vigorous convection that occurs in the tropics has little effect on the mean mass flow into the stratosphere, this is again governed by mid latitude wave activity sucking air out of the tropics. It follows from this that there will be significant variations with latitude and season as this will affect the amount and type of wave activity. Mesospheric gravity waves, for example, tend to be more important in the southern hemisphere in winter than in the north, and those in the middle and upper stratosphere in the summer hemisphere play an important role in the seasonal variation of the rising tropical air (Garcia & Boville [1994]).

1.1.2 Transport in the Tropics

Mid latitude wave-driven circulation is the main effect of interest here as it is this that governs the transport of the relevant chemical species but, although it is virtually independent of the thermally driven tropospheric circulation, transport to the stratosphere cannot be totally explained in terms of the larger scale circulation.

While vertical transport within the lower stratosphere proceeds on a time scale of months, transport up to the tropical tropopause may occur in a matter of hours; this difference in transport rate is partly responsible for change in atmospheric composition that is seen across the tropopause.

Convection in the tropics plays a significant role in dehydrating the air entering the stratosphere. Thermally heated air rises rapidly, forcing the air above it up or outward from its path while below air will converge and be entrained behind. The rising air undergoes adiabatic expansion which causes a lowering of the local temperature, thus causing much of the water vapour in the air mass to condense or freeze. The latent heat released in by these processes warms the air, increasing its buoyancy and causing it to accelerate upwards. As a result of this, such cumulonimbus turrets may overshoot their radiative equilibrium positions and penetrate the tropopause into the relatively warm lower stratosphere.

Above the tropopause the air is no longer buoyant and sinks, but having mixed with warmer stratospheric air its equilibrium level may be in the lower stratosphere. This is, of course, an ongoing process and so further rising air forces that which is already there to spread out leading to the development of a large ‘anvil’ extending beyond the cumulonimbus feeding it. The top of these anvils cools radiatively while the bottom is subject to upwelling radiation resulting in a turbulent reflux in which ice crystals grow in size until they precipitate out, thereby reducing the mixing ratio of water to stratospheric levels (about 3 ppmv) (Danielsen [1982]). Although such cumulonimbus anvils can penetrate into the lower stratosphere this is not a major source of transport into the main part of the stratosphere; this is governed by the aforementioned mid latitude wave activity.

1.1.3 The Polar Vortex

After the autumn equinox, radiative cooling of the air over the poles becomes significant and, in conjunction with the Earth’s rotation and the poleward flow of air (section 1.1.2), results in the development of strong westerly winds, known as the polar night jet, around the poles (Brasseur et al. [1999]). These westerlies, which extend into the mesosphere, form a barrier to meridional transport, thereby separating the mid latitude and polar air masses. The bulk of air arriving from mid latitudes will then descend outside the vortex, although some air may pass over the edge of the vortex at high altitudes and some may enter through the side as a result of various mixing events that shall not be discussed here (Plumb et al. [1994]). Air contained within the vortex may be considered as isolated throughout the winter (especially in the southern hemisphere), until the vortex breaks up in the spring due to radiative heating.

The polar vortex in the southern hemisphere is significantly more pronounced and stable than that in the north due, in part, to topography: the south pole being situated in a large, relatively high landmass surrounded by a large expanse of ocean while the north pole is situated in an almost landlocked sea.

1.2 Stratospheric Ozone and Terrestrial Ultraviolet Levels

The absorption of ultraviolet light by atmospheric ozone was first investigated by Cornu who showed by means of laboratory and field measurements that ozone in the atmosphere could be absorbing sunlight at wavelengths less than 293 nm (Cornu [1879a], [1879b]). Stratospheric oxygen (diatomic and triatomic) plays a crucial role in reducing the intensity of ultraviolet radiation reaching the surface of the planet to levels suitable for life.

The link between ultraviolet radiation and skin cancers in man is well recognised, with the action spectrum showing a peak at 260-270 nm due to absorption by nucleic acid. Exposure to high levels of ultraviolet radiation, especially between 295 and 315 nm, can also lead to the induction of cataract. Acute overexposure can lead to photokeratits (snow blindness/welder’s flash) while gross overexposure can result in vascularisation of the cornea; the action spectrum here has a peak at about 270 nm but vascularisation can occur at wavelengths up to 320 nm.

Diatomic oxygen is responsible for reducing the levels of short wavelength (less than 242 nm) ultraviolet radiation as described by

O2 + hn ® O + O

(1.1)

while ozone can absorb ultraviolet in the wavelength range of approximately 200 - 320 nm according to

O3 + hn ® O2 + O.

(1.2)

1.3 Stratospheric Ozone Chemistry.

1.3.1 Ozone Production and Chapman’s Mechanism.

The first attempt to explain the formation and destruction of ozone in the atmosphere was published by Chapman in 1930 (Chapman [1930]). Chapman assumed that the ozone was “situated in a uniform layer of the atmosphere 10 km. thick, between 40 and 50 km. height.” Chapman postulated that the formation of ozone occurred as follows:

Molecular oxygen is photolysed to produce atomic oxygen which then combines with other oxygen molecules in the presence of another atom or molecule, M:

O2 + hn ® 2O,

(1.3)

O + O2 + M ® O3 + M.

(1.4)

Some of the ozone so produced may then be broken down by combination with atomic oxygen:

O + O3 + M ® 2O2 + M.

(1.5)

Chapman also proposed that ozone might be destroyed by thermal decomposition:

2O3 ® 3O2

(1.6)

or “by dissociation in the manner inverse to” (1.4):

O3 ® O2 + O.

(1.7)

The energy for this reaction being imparted either by ultraviolet radiation or collision with another molecule.

While Chapman’s oxygen-only scheme covers the formation of ozone correctly, the destruction mechanisms proposed provide an insufficient sink. Reactions (1.4) and (1.7) are ‘fast’ while (1.3) and (1.6) are ‘slow’. Equation (1.7) does not actually destroy O₃ as the O formed is mopped but by (1.4), so that the fast reactions conserve odd oxygen but convert it between O and O₃. Equation (1.6) is the only mechanism in Chapman’s scheme whereby ozone may be destroyed but it is now known that this is too slow for the amount of destruction required to explain the observed values (Andrews [1987]) and additional reactions are required (Brasseur et al. [1999]). These shall be considered in the following sections.

1.3.2 Stratospheric Ozone Depletion.

The current concern about decreasing levels of stratospheric ozone stems largely from the results of a study of total ozone columns over Halley Bay, Antarctica (27° W, 76° S), for the period 1957-1984 carried out by Farman et al. (Farman et al. [1985]). This showed that despite the often accepted model predictions that variations in total ozone would be slight (Farman et al. [1985], WMO [1982]) there was a significant decrease in the springtime amounts over Halley Bay. These findings were supported by measurements made by the TOMS and SBUV instruments onboard the Nimbus 7 satellite (Stolarski et al. [1986]) and by balloon borne ozone sonde measurements from McMurdo station (167° E, 76° S) (Deshler et al. [1990]). These balloon soundings also showed that the loss was occurring between approximately 12 km and 22 km. In 1994 over the south pole the region of ozone depletion was shown to extend from approximately 9 km to 23 km with total destruction between 14 km and 19 km (Hofmann et al. [1994]).

The Arctic has been shown to experience pronounced ozone loss in the spring in an analogous manner, although the extent of depletion is rather less, with the seasonally averaged depletion reaching only -1.5% per year in the arctic spring compared with -3% in the antarctic (Stolarski et al. [1991]). While depletions in the mid latitudes are substantially smaller than at the poles, and are roughly comparable in both hemispheres, the high population densities in the northern hemisphere make this trend a matter some concern.

The following sections shall look briefly at the main processes responsible for the destruction of stratospheric ozone.

1.3.3 HO_x Chemistry

O(1D) + H2O ® 2OH,

(1.8)

O(1D) + CH4 ® OH + CH3.

(1.9)

Four catalytic cycles are listed below with their regions of operation (Brasseur et al. [1999]):

1)

H + O3 ® OH + O2

(1.10)

OH + O ® H + O2

(1.11)

Net:

O3 + O ® 2O2

[above ~40 km]

2)																																																		OH + O ® H + O2																																																																(1.11)
																																																		H + O2 + M ® HO2 + M																																																																(1.12)
																																																		HO2 + O ® OH + O2																																																																(1.13)
Net:																																																		2O + M ® O2 + M																																																																[above ~40 km]

3)

OH + O3 ® HO2 + O2

(1.14)

HO2 + O ® OH + O2

(1.13)

Net:

O3 + O ® 2O2

[below ~40 km]

4)

OH + O3 ® HO2 + O2

(1.14)

HO2 + O3 ® OH + 2O2

(1.15)

Net:

2O3 ® 3O2

[below ~30 km]

Cycle 2) does not involve ozone directly but does affect the balance of oxygen species in the stratosphere by converting monatomic oxygen to diatomic and cycle 1) only becomes important above about 60 km. The height dependence of each of these cycles stems from the variation in concentration of each of the relevant chemical species (Wennberg et al. [1994]).

1.3.4 NO_x Chemistry

The main source of stratospheric NO_x species (NO and NO₂) is the photolysis of nitrous oxide. Nitrous oxide is has its source in the troposphere, mainly being released from soils and oceans (WMO [1999]), but the only well quantified sink is stratospheric photochemistry (Brasseur et al. [1999]).

Nitrous oxide is transported across the tropopause, following the general circulation (section 1.1.1) and upward in the stratosphere. The majority of the nitrous oxide is photolysed:

N2O + hn ® N2 + O(1D)

(1.16)

while a small proportion is oxidised in one of the following ways:

O(1D) + N2O ® 2NO

(1.17a)

O(1D) + N2O ® N2 + O2

(1.17b)

with the former being slightly more prevalent than the latter (Seinfeld & Pandis [1999]). The rôle played by NOx was identified around 1970 (Crutzen [1970], Johnston [1971]); two main catalytic cycles exist by which NOx may participate in the destruction of ozone:

1)																																																		NO + O3 ® NO2 + O2																																																																(1.18)
																																																		NO2 + O ® NO + O2																																																																(1.19)
Net:																																																		O + O3 ® 2O2

2)																																																		NO + O3 ® NO2 + O2																																																																(1.18)
																																																		NO2 + O3 ® NO3 + O2																																																																(1.20)
																																																		NO3 + hn ® NO + O2																																																																(1.21)
Net:																																																		2O3 ® 3O2

Cycle 1) is dominant throughout the middle atmosphere. Neither cycle exhibits the degree of height dependence of HO_x chemistry as a result of other reactions, not directly related to ozone, controlling the amounts of NO and NO₂ at various altitudes (Brasseur et al. [1999]).

Summary
Acknowlegements
Contents
Chapters: 1, 2, 3, 4, 5, 6, 7, 8
Appendices
References