Andrews Dawid G. An Introduction to Atmospheric Physics

Подождите немного. Документ загружается.

9 Some atmospheric observations

Fig. 1.4 Typical vertical structure of the mean midlatitude ozone number density (molecules m

−3

). Based

on data from US Standard Atmosphere (1976).

Above the middle atmosphere is the upper atmosphere, where effects of ionisation

become dominant in determining the atmospheric structure and the air becomes so rareﬁed

that the assumption that it can be treated as a continuous ﬂuid starts to break down. In this

book we concentrate on the physics of the lower and middle atmospheres.

From hydrostatic balance, the pressure at any level in the atmosphere is proportional

to the mass of air above that level. From the pressure axis in Figure 1.3 it follows that

approximately 90% of the atmospheric mass is in the troposphere, a little under 10% in the

stratosphere and only about 0.1% in the mesosphere and above.

Despite their relatively low

mass, the stratosphere and mesosphere are not insigniﬁcant, however. For example, ozone

in the stratosphere absorbs ultra-violet solar radiation, thereby protecting the biosphere

from potentially damaging effects.

Figure 1.3 is not representative of all latitudes and seasons. A more comprehensive

plot, of the zonal-mean (i.e. longitudinally averaged) temperature averaged over several

Januaries, as a function of height and latitude, is given in Figure 1.5. It will be seen that

although the general shape of the vertical variation of temperature in midlatitudes is roughly

in accord with that in Figure 1.3, there are signiﬁcant latitudinal variations of the heights

and magnitudes of the temperature extrema. For example, the equatorial tropopause is at

a greater altitude and colder than that at higher latitudes, the summer stratopause is lower

In this book we use the unit hPa for pressure (1 hPa = 10

Pa). This is equivalent to the millibar, which was

formerly in common use in meteorology.

10 Introduction

Fig. 1.5

Zonal-mean temperature (K) for January, from the CIRA (COSPAR International Reference

Atmosphere) dataset. A small region at low levels over Antarctica is omitted. Based on data from

Fleming et al. (1990).

and warmer than the winter stratopause and the summer mesopause is extremely cold.

(In fact the lowest natural terrestrial temperatures are found there.)

Some of these temperature features can be crudely explained in terms of simple physical

mechanisms. For example, the warm stratopause can be attributed to the ozone distribution,

which peaks in the stratosphere; absorption of solar radiation by the ozone leads to heating

of the upper stratosphere and, since in equilibrium this heating is balanced mainly by infra-

red cooling from carbon dioxide, there must be a local temperature maximum, so that heat

can radiate to cooler regions.

If radiative control of temperature were to continue down to the ground, the temperature

in the troposphere would decrease much more rapidly with height than is observed, and this

temperature proﬁle (and its associated density proﬁle) would be statically unstable. Such

a temperature proﬁle could not persist, but might be expected to give rise to convective

activity that would modify the temperature proﬁle, causing it to decrease less rapidly with

height until it was just statically stable again. This process appears to occur in the moist

tropical troposphere, where the temperature decrease with altitude is fairly close to the

11 Some atmospheric observations

saturated adiabatic lapse rate, the rate of decrease with height of the temperature of

a parcel of air saturated with water vapour, which condenses (releasing latent heat) as

the parcel rises; see Section 2.8. At higher latitudes other processes, such as midlatitude

cyclones and anticyclones, may also play a part in determining the temperature proﬁle.

The position of the tropopause depends on an interplay between the processes that cause

the temperature to fall with height in the troposphere and increase with height in the

stratosphere; the precise details are a subject of active current research.

The explanations of the cold equatorial tropopause and extremely cold summer

mesopause are quite complex. It turns out that there is dynamically driven rising motion

in both of these regions; the rising air expands as it moves to lower pressure and cools.

Conversely there is descent over the winter pole, leading to temperatures that are warmer

than would otherwise be the case.

Figure 1.6 shows the zonal-mean zonal (west-to-east) winds for January. These winds are

relatedtothetemperatureﬁeldinFigure 1.5 by thermal windshear balance, as expressed

Fig. 1.6

Zonal-mean zonal wind (m s

−1

) for January, from the CIRA d ataset. Thin solid lines: eastward

winds; thick solid line: zero winds; dashed lines: westward winds. A small region at low levels

over Antarctica is omitted. Based on data from Fleming et al. (1990).

12 Introduction

by equation (4.28b). In the troposphere, the mean zonal winds are generally eastward

at midlatitudes, with two prominent ‘jet streams’, and westward at low latitudes. (The

terms ‘westerly’, meaning eastward, and ‘easterly’, meaning westward, are commonly

used in meteorology, but will be avoided in this book.) In the stratosphere and mesosphere

the mean zonal winds are generally eastward in winter and westward in summer. The

winds at low latitudes shown in Figure 1.6 are not representative of every January, partly

because they do not follow a simple annual cycle in the equatorial lower stratosphere;

there is a prominent quasi-biennial oscillation there, with a period of approximately

28 months.

1.4.2 Gravity waves

Figure 1.7 shows northward and eastward wind components between altitudes of 60 and

80 km, over Alaska, measured by a ground-based radar. The radar transmits radio waves

almost vertically and measures the backscattered signal; from this, wind velocities can

be determined; see Section 7.3.2. Roughly sinusoidal wind ﬂuctuations in the vertical are

seen, with a wavelength of about 15 km and propagating downwards in time with a period

of about 9 h.

These quasi-sinusoidal ﬂuctuations are strongly suggestive of some kind of wave motion,

and further investigation conﬁrms this. The waves are an example of the ﬂuid-dynamical

gravity waves mentioned in Section 1.1. Atmospheric gravity waves are analogous to

horizontally propagating surface waves on water, which depend on the restoring mechanism

provided by the contrast in density between air and water: the water in a wave crest is denser

Fig. 1.7 An example of an atmospheric gravity wave over Alaska, as measured by a ground-based radar.

The northward and eastward wind components (in m s

−1

) are shown as a function of height,

between altitudes of 60 and 80 km, and at each hour over a period of about 17 hours. From

Balsley et al. (1983).

13 Some atmospheric observations

than the surrounding air and tends to fall, while the air in a wave trough is lighter than the

surrounding water and tends to rise. We can imagine the smooth vertical density variation in

the atmosphere to be approximated by a stack of thin ﬂuid layers, whose densities decrease

with height. The surface-wave mechanism operates at each density interface, so now the

wave can propagate vertically as well as horizontally.

The particular type of gravity wave shown in Figure 1.7 is called an inertia–gravity wave

(see Section 5.4); it is actually of large enough horizontal scale (a few hundred kilometres)

and period to be inﬂuenced to some extent by the Earth’s rotation. These measurements

provide an unusually clear example of a sinusoidal oscillation: in most cases, too many

other dynamical processes are occurring in the atmosphere for waves to be very easily

identiﬁed, and careful data analysis must be performed to isolate them.

Perhaps surprisingly, the downward phase progression of the waves in Figure 1.7 indi-

cates upward propagation of ‘information’ by the waves (i.e. an upward group velocity).

Gravity waves will be studied in detail in Section 5.4 and, among other things, it will be

shown that this type of wave is dispersive; the phase and group velocities can therefore be

in different directions. This is just one of the ways in which the propagation characteris-

tics of the atmospheric waves studied in this book differ from those of the more familiar

non-dispersive waves such as electromagnetic waves in a vacuum.

Gravity waves are generated in many different ways, including by air ﬂow over mountains

and by convective activity in the troposphere. Waves generated in the lower atmosphere may

propagate upwards into the stratosphere and mesosphere. As the background air density

decreases, the amplitudes of the wave ﬂuctuations in wind (and associated ﬂuctuations in

temperature and density) will rise. As a result, gravity waves may attain large amplitudes

in the mesosphere and exert a considerable inﬂuence on the mean atmospheric state there.

1.4.3 Rossby waves

Figure 1.8 depicts the temperature in the Northern Hemisphere, at a level near 24 km

altitude, during a period when the stratosphere was disturbed by a vigorous dynamical

event known as a stratospheric warming. A cold region is located on one side of the pole

(roughly along 0

◦

E) and a warm region on the other (roughly along 180

◦

E). Moving around

a latitude circle near the pole, we ﬁnd that the temperature varies roughly sinusoidally with

longitude, one wavelength encompassing 360

◦

of longitude. The fact that the cold part

of the disturbance is smaller and stronger than the warm part shows that the ﬂuctuation

is not exactly sinusoidal. However, the phenomenon is wave-like in many respects and is

an example of the Rossby wave mentioned in Section 1.1. The horizontal wavelength is

several thousand kilometres in extent and the wave’s dynamics are quite different from

those of the gravity wave. The propagation mechanism is rather subtle, depending both on

the rotation and on the curvature of the Earth; the details are given in Section 5.5. Rossby

waves, like gravity waves, are dispersive.

Several types of Rossby wave are observed in the atmosphere. Some are stationary,

that is, their wave patterns are ﬁxed with respect to the Earth; since they are dispersive

14 Introduction

Fig. 1.8

Polar stereographic map of atmospheric temperature (K) near 24 km altitude in the Northern

Hemisphere stratosphere on 9 January 1992, as measured by the Improved Stratospheric and

Mesospheric Sounder (ISAMS) on the Upper Atmospheric Research Satellite (UARS). The North Pole

is at the centre, the 60

◦

Nand30

◦

N latitude circles are shown and the equator is the outer circle;

four longitudes are also shown. (Diagram supplied by Dr A. M. Iwi.)

they may still propagate information, because the group velocity may be non-zero even if

the phase speed is zero. Others have patterns that travel with respect to the Earth. Strong

disturbances in the stratosphere, such as that shown in Figure 1.8, are frequently due to

upward propagation of Rossby waves generated by large-scale weather disturbances in

the troposphere. It is found that, when the background winds are eastward, as in winter

(see Figure 1.6), only the longest-wavelength stationary Rossby waves can propagate

vertically. This accounts for the observed prevalence of larger-scale disturbances in the

winter stratosphere.

1.4.4 Ozone

As mentioned in Section 1.4.1, atmospheric ozone is important because it absorbs solar

ultra-violet radiation, thus protecting human and animal life from potentially harmful

15 Some atmospheric observations

Fig. 1.9

Zonal-mean volume mixing ratio of ozone (parts per million by volume), as a function of latitude

and height, for January, based on the 5-year climatology of Li and Shine (1995). Data provided by

Dr D. Li.

consequences. It is therefore crucial to have good measurements of ozone concentrations

and a good understanding of the processes by which it is produced and destroyed.

A typical vertical proﬁle of the ozone number density was shown in Figure 1.4.An

alternative measure of the concentration of a gaseous constituent of air, such as ozone, is

the volume mixing ratio, that is, the number density of the constituent divided by the total

number density of the ‘air’. A useful property of the volume mixing ratio is that, unlike

the number density, it is constant for a moving parcel of air in the absence of chemical

production and loss processes. Figure 1.9 gives a latitude–height cross-section of the zonal-

mean volume mixing ratio of ozone for January. This shows that the maximum values are

in the low-latitude stratosphere (where the photochemical production of ozone is greatest)

but also that there are signiﬁcant values over the winter pole, where no production takes

place since the Sun is below the horizon all day. This suggests that ozone is transported

into the ‘polar night’ region by wind motions.

Further information on the global ozone distribution is given in Figure 1.10,which

shows the ‘column ozone’, a measure of the total number of ozone molecules in a vertical

column of atmosphere, as a function of latitude and season. Low values of column ozone

are found all year round at low latitudes, despite the fact that this is where most production

of ozone takes place. Maximum amounts are found in spring: in the Northern Hemisphere

16 Introduction

Fig. 1.10

The observed annual cycle in column ozone, based on the 5-year climatology of Li and Shine

(1995). The units are Dobson Units: see Section 6.7.DataprovidedbyDrD.Li.

this maximum occurs at high latitudes, whereas in the Southern Hemisphere it occurs at

middle latitudes; a relative minimum is found in the Antarctic spring.

The Antarctic spring minimum of column ozone in Figure 1.10 is a manifestation of

the Antarctic ozone hole, a dramatic reduction of ozone in the Antarctic stratosphere

that has been observed in spring since the mid-1970s. Intensive international efforts, both

observational and theoretical, established that this loss of ozone is due to complex chemical

and physical processes, involving chlorine resulting from man-made compounds such as

chloroﬂuorocarbons (CFCs), taking place on the ice and water particles that can form in the

extremely low Antarctic winter temperatures. The ozone lost during spring over Antarctica

is mostly replenished there later in the year, but this effect may be contributing to a slow

global loss of ozone. This research led to the Montreal Protocol, in which a phasing out

of CFCs was agreed by the international community. Depletion of ozone is also occurring

at other latitudes and is currently the subject of global monitoring and modelling. Further

details are given in Chapter 6.

17 Weather and climate

1.5 Weather and climate

The word weather, while having a clear enough meaning to the layperson, does not have

a precise deﬁnition in atmospheric physics. However, it tends to encompass tropospheric

events associated with atmospheric ﬂows with length scales of hundreds of metres and

more, and time scales of a few days or less, although weather patterns may occasionally

persist for a week or two. Weather phenomena are notoriously irregular: indeed, Lorenz’s

famous pioneering work on chaotic systems originated in a study of a simple nonlinear

atmospheric model and was motivated by considerations of weather prediction. However,

atmospheric data averaged over a month or so usually behave in a more regular manner.

Most localities have regular longer-term variations in the average weather conditions,

with a roughly annual cycle. This annual cycle does not repeat precisely from one year

to the next: there may be dramatic interannual variations in the average weather at a

given place.

Much of the thrust of atmospheric research has been directed towards the improvement of

weather forecasting, for periods of up to two weeks ahead. Although present-day forecasting

is partly limited by the insufﬁciency of computer power and incompleteness of atmospheric

data, it is also limited to some extent by gaps in our understanding of the basic physics and

dynamics of the atmosphere.

The word climate refers to the state of the atmosphere on longer time scales, typically

averaged over several years or more. The understanding of climate and climate change does

not necessarily require a complete understanding of every weather event; conversely there

are physical processes, operating on long time scales, that are unimportant for weather

prediction but crucial for climate prediction. (An example is heat transport from the deep

ocean, which may vary on decadal or longer time scales.)

There is much current concern about whether human activity may be changing the

climate signiﬁcantly, principally through an enhancement of the greenhouse effect due

to increasing levels of carbon dioxide resulting from the burning of fossil fuels and the

destruction of the tropical rain forests. Given the long time scales involved, the detection

of changes in climate is difﬁcult; however, the latest estimates suggest, for example, that

the global mean surface air temperature has increased by about 0.7 ± 0.2

◦

C over the last

century. It is even more difﬁcult to distinguish between man-made and natural climate

change, given the complexity of the climate system (including the atmosphere, oceans, ice

cover, biosphere and solid Earth). Since controlled experiments cannot be performed on the

climate system, we must use models to identify cause-and-effect relationships, as noted in

Section 1.2. Even the most sophisticated current models are still over-simpliﬁed. However,

the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (Solomon

et al. (2007)) states with very high conﬁdence that the global average net effect of human

activities since 1750 has been one of warming, amounting to an increase of between 0.6

and 2.4 W m

−2

in the power per unit area entering the climate system. See Chapter 8 for

more details.

18 Introduction