James I.N. Introduction to Circulating Atmospheres

Подождите немного. Документ загружается.

7.2 Monsoon

circulations

219

(a)

equator

Fig. 7.6. Schematic illustration of idealized monsoonal circulations: (a)

flow

induced

hot continent north of the equator and cold ocean south of the equator; (b)

effect of orography across the equator; (c) effect of mountains to the north of the

heating region.

just given are mainly

terms

the feedbacks between the large scale

circulation and the heat source over south-east Asia. However, comparison

of Fig. 3.8(b) with Fig. 7.2 shows that the region of strong heating is quite

localized compared to the extent of the upper level monsoonal anticyclone.

The question arises: what determines the spatial scale

the monsoonal

circulation? The linear arguments of the preceding section provide the most

accessible simple theoretical prediction. The monsoonal circulation

not

unlike the sort

solution shown

Fig. 7.5 which shows the response of

the tropical atmosphere to

heating maximum located to the north of the

equator. The meridional scale

essentially the equatorial Rossby radius,

(2c

/j8)

1//2

The zonal scale according to this theory is related to the friction

timescale and the zonal components

the group velocity

waves, i.e.,

These ideas certainly suggest the observed zonal elongation of

220

Three-dimensional aspects of the global circulation

25.0

Fig. 7.7. Ertel's potential vorticity on the

360

K surface in the monsoon region

during July 1990, computed from ECMWF data by M. Masutani. Shading indicates

values between 0.75 and 1.5

PVU.

The region covered by the plot extends from 0°E

140 °E

and from

20 °S

°N.

the upper monsoonal circulation, but are probably too crude to give very

useful numerical estimates of the scales. It must be recognized that the

monsoonal circulation is highly nonlinear, and so the applicability of linear

theory is dubious. This is illustrated by Fig. 7.7, which shows the Ertel

potential vorticity associated with the upper monsoonal circulation. The

nonlinearity is expressed by the large distortion of the potential vorticity

contours, and the significant angle between the contours and the velocity

vectors.

7.3 Midlatitude storm zones and jets

In the midlatitudes, the distribution of transient eddy kinetic energy, and

other eddy quantities, is not uniform. Rather, the transient activity at

higher frequencies is concentrated in localized regions. Partitioning the data

according to the frequencies of the transients is simply accomplished by

the application of a numerical filter, such as that described in Section 5.1.

We shall separate high frequency transients, whose periods are less than

6-10

days,

from low frequency transients with longer periods. Figure 7.8(a)

shows the high frequency eddy kinetic energy for DJF. The

field

is dominated

by two elongated maxima at midlatitudes, one located over the Atlantic and

one over the Pacific. Much the same pattern is found when other measures of

73 Midlatitude storm zones and jets 221

eddy activity are considered. Some examples of these are shown in the other

parts of

Fig.

7.8. A similar pair of maxima are seen in the geopotential height

variance. The temperature fluxes suggest that the maxima in kinetic energy

are closely associated with baroclinic instability. For example, Fig. 7.8(c)

shows the vertical temperature flux at 70kPa. This is upwards nearly

everywhere, but has maximum values near the western end of the kinetic

energy maxima. As we saw in Section 5.3, such upward temperature fluxes

are a necessary condition for the conversions of available potential energy

into kinetic energy. Finally, Fig. 7.8(d) shows the poleward momentum flux,

at 25kPa. The largest values are again found close to the maxima in

eddy kinetic energy. This time, however, the pattern is rather more complex.

Large poleward momentum flux is found to the south and near the eastern

end of the kinetic energy maxima. Poleward of this location, there is a

weaker maximum of equatorward momentum flux.

The association between baroclinic energy conversions and the maxima of

transient eddy activity suggests that there may be a link between synoptic

weather systems and the distribution of eddy activity. Indeed, this turns out

to be the case. Figure 7.9 shows the tracks of the surface centres of major

depression systems observed during a particular winter over the Atlantic.

Superimposed is the

90 m

variance contour of the

kPa geopotential height.

The cyclogenesis region lies towards the westwards extremity of the high

pass eddy activity maximum, close to the eastern coast of North America.

Thereafter, the tracks of the depressions tend to run close to the major axis

of the geopotential height variance. The cyclones are observed to decay and

fill mainly towards the eastern side of the Atlantic, where the levels of eddy

activity are observed to decline. The close association between depression

tracks and the high pass filtered eddy variances have led to the latter being

called 'storm tracks'. A similar association is found for the Pacific storm

track.

The occurrence of discrete storm tracks is presumably related in some

way to the strong zonal variations of the surface properties in the northern

hemisphere. We will consider exactly how this might operate in the next

section. The southern hemisphere is more uniform in its surface properties,

especially at latitudes south of

°S,

where most of the baroclinic energy

conversion takes place. Yet a structure analogous to the northern hemisphere

storm tracks is still observed. Figure 7.10 is similar to Fig. 7.8, except that

it shows the southern hemisphere during the winter (JJA) season. There is a

distinct maximum in the high pass filtered eddy kinetic energy in the southern

Atlantic and Indian Oceans, and a relative minimum in the southern Pacific.

Other high pass eddy statistics show much the same variation along this

222 Three-dimensional aspects of the global circulation

(a)

Fig. 7.8. Various high pass filtered transient eddy statistics for the northern hemi-

sphere, DJF period. Based on six years of ECMWF data, (a) Eddy kinetic energy,

+ i/

)/2. Contour interval 25m

, values in excess of 100m

s~~

shaded, (b)

Geopotential height variance, Z'

. Contour interval

m, values in excess of 90 m

shaded.

73 Midlatitude storm zones and jets

223

(d)

Fig. 7.8 (cont). (c) Vertical eddy temperature flux,

Contour interval

0.05KPas"

, values less than —0.2KPas"

shaded, (d) Northward momentum

flux, wV. Contour interval

. Dashed contours indicate equatorward mo-

mentum flux.

224

Three-dimensional aspects of the global circulation

Fig. 7.9. The tracks of low pressure centres over the North Atlantic for the period

December 1985 to February 1986. The shading indicates the region where the high

frequency Z

exceeded

90 m

in the ECMWF analyses for the same period.

'storm track' as observed in the northern hemisphere, namely, the elongated

maximum in geopotential height variance, the large vertical temperature

flux at low levels, and the dipolar structure of the poleward momentum flux

towards the downstream end of the storm track. Attempts to correlate the

tracks of synoptic systems with the variance maximum are less successful

than in the northern hemisphere. There is a tendency for the cyclonic systems

to spiral polewards from cyclogenesis regions on the equatorward flank of

the 'storm track' to decay regions in the 'circumpolar trough', the region of

low pressure around the Antarctic coast. Partly at least, this is the result

of attempting to identify the centres of synoptic weather systems by means

of extrema in the surface pressure field. Because the surface wind field is

strong around the southern hemisphere baroclinic zone, there is a natural

tendency for centres of low pressure to be displaced poleward of the vortex

centre, and for centres of high pressure to be displaced equatorward. But,

partly, it seems that this spiral trajectory of weather systems is real. Perhaps

it would be better to describe the regions of large high frequency variance

as 'storm zones' rather than 'storm tracks'. However, the latter nomenclature

is in general use despite being rather misleading.

Each of the three major storm zones has a distinctive seasonal behaviour.

Midlatitude storm zones and jets

225

The Atlantic storm track is most pronounced during DJF and least pro-

nounced in JJA. Its actual location varies very little. The Pacific storm track

is at its most intense during the transition seasons, MAM and SON, is

rather weaker during DJF and is much weaker during JJA. Thus, there is

a clear semi-annual as well as annual cycle in its behaviour. The southern

hemisphere storm track shows rather little seasonal variation, though it is

at somewhat lower latitudes during DJF than the remaining seasons, and is

marginally more intense during MAM. Once again, the seasonal cycle has a

semi-annual as well as an annual component. The most significant feature

is that, for all three storm zones, the location of the beginning of the zone

seems to be independent of season, even though the intensity and length of

the storm zone varies.

The storm zones occur in association with major jet streams in the tropo-

sphere. Particularly in the northern hemisphere winter, the zonal flow varies

considerably with longitude. Fig. 7.11 shows the mean wind speed, averaged

with respect to pressure, in both hemispheres for DJF and JJA. In the

northern hemisphere winter, the most prominent jet is located in the western

Pacific, with a shorter and weaker jet over the east coast of North America

and the west Atlantic. A third jet is located over Arabia. In the summer,

the jets are much weaker, but occur in much the same places. The southern

hemisphere jet has a much smaller seasonal cycle. It is strongest over the

southern Atlantic and Indian Oceans. Particularly in winter (JJA), there is

a minimum of wind speed in the vicinity of New Zealand, with the depth

averaged flow splitting to north and south. Comparison between Fig. 7.11

and the various transient eddy quantities shown in Figs 7.8 and 7.9 reveals

that the northern hemisphere storm zones are most intense near the longitude

of the jet exits, but rather on their poleward flanks. The Atlantic storm zone

conforms most closely to this pattern. In the southern hemisphere, the storm

zone is much longer, and it begins roughly in phase with the acceleration of

the depth average wind. If anything, the strongest eddy activity is slightly

towards the equatorward side of the jet. Because the Atlantic storm zone is

rather better observed, and certainly more fully documented than the other

major storm zones, writers have a tendency to regard it as the 'normal' storm

zone from which the Pacific and southern hemisphere storm zones deviate

to some extent. But it could equally be that the Atlantic storm zone is the

anomalous one.

In Section 5.5, the lifecycle of a single idealized baroclinically unstable

disturbance was described. The growing phase of the wave was charac-

terized by large poleward and upward temperature fluxes, associated with

the conversion of available potential energy to eddy kinetic energy. As the

226

Three-dimensional aspects of the global circulation

(a)

(b)

Fig. 7.10. As Fig. 7.8, but showing the southern hemisphere during the JJA season:

(a) Eddy kinetic energy,

+1/

)/2; (b) Geopotential height variance, Z

7.3 Midlatitude storm zones and jets

227

(d)

Fig. 7.10

(cont.).

flux, u'v'. Plotting conventions as for

Fig.

7.8 except for

(c),

where the contour interval

is 0.02KPas"

and values less than

0.1

KPas"

are shaded.

228

Three-dimensional aspects of the global circulation

(b)

Fig.

7.11.

The zonal wind speed, averaged with respect to time and pressure between

100

kPa and 15kPa. Contour interval

ms"

Shading indicates values in excess of

ms"

(a) northern hemisphere, DJF; (b) northern hemisphere, JJA.