UDC 551,576.34:551.507.362.2:551.513.2(215-13)
Some Characteristics of Satellite-Observed Bands
Of Persistent Cloudiness
Over the Southern Hemisphere
NEIL A. STRETEN *--Commonwealth Meteorology Research Centre
Melbourne, Victoria, Australia
~~
ABSRACT-An analysis is made of the annual and
seasonal frequency of location and movement of maximum
brightness (cloudiness) bands of substantial extent ap-
pearing on the Southern Hemisphere 5-day-averaged
satellite data; the possible relation of such bands to the
hemispheric long-wave pattern is discussed. The number
of bands over the hemisphere for a particular 5-day period
varies with latitude and season, but a high frequency of
3-4 is observed at midlatitudes in all months. Three years
of averaged data indicate that the location of high band
frequency is largely stable in the Pacific and Atlantic
Oceans but that the Indian Ocean displays higher fre-
quencies to the west in summer and to the east in winter.
Longitudinal displacement of bands between succesive
5-day periods varies to some degree with season, but only
little with latitude. It is least and most symmetric in the
South Atlantic, but elsewhere it is predominantly eastward
with a peak frequency of 5'-10' of longitude per period.
Some evidence exists for a longer term westward trend in
the location of the Pacific band from mid 1969 to mid
1971. The frequency of 5-day-averaged cloud bands is in
qualitative agreement with patterns of rainfall over
Australia in specific winter and spring seasons and points
t o qualitative assessment of broad patterns of oceanic
rainfall.
1. INTRODUCTION
Features of both daily and multiday averages of
brightness recorded by satellite cameras and radiometers
and depicted on global digital maps (Booth and Taylor
1969, Leese et al. 1970) are the prominent bright bands
extending either predominantly zonally or meridionally
in particular geographic locations. Such bands are most
clearly defined over the oceans and, in the case of nonzonal
orientations, particularly over the Southern Hemisphere,
where the bands frequently extend to high latitudes.
Recent analyses of the zonally oriented banding in
the Tropics (Gruber 1972) has revealed aspects of the
behavior of the intertropical convergence zone (ITCZ),
and Saha (1971) has drawn attention to the apparent
association between the tropical cloud bands and ridges
of sea-surf ace temperature. Further, Erickson and Win-
ston (1972) have examined the relation between cloud
bands extending from tropical storms poleward into the
Northern Hemisphere westerlies and the increase in the
strength of the planetary circulation in autumn. They
suggest that these particular bands visually depict chan-
nels for the transfer of energy in the form of heat and
moisture from the tropical storms to midlatitud5s.
2. CLOUD BANDS AND THE LONG-WAVE
PATTERN OF THE SOUTHERN HEMISPHERE
Extratropical bright bands on daily pictures represent
either frontally organized clouds usually associated with
1 Now at the Owphysical Institute, University of Alaska, Fairbanks, Alaska
486 / Val. 101, No. 6 / Monthly Weather Review
a cyclonic cloud vortex a t middle or high latitudes or,
alternatively, an elongated region of low or midtropo-
spheric convergence frequently lying between midlatitude
anticyclones. When viewed on averaged photographs
over a number of days, the banding is related to the
persistence of particular features or to repetitive events
in the same location. Examination of Southern Hemis-
phere mosaics (e.g., Streten 1968b) indicates that the
averaged bright banding at middle to high latitudes is
due to a considerable regularity in the tracks of depres-
sions, to the persistence of convergence in particular
locations, or to a combination of both factors.
If the bands are primarily related to these features,
synoptic experience suggests that they should be located
close to, though probably eastward of, the upper long-
wave troughs in the lower troposphere. Staver (1969)
found that clouds over the southern oceans in summer
formed primarily ahead of the 700-mb trough, where
warm air is being advected over relatively cool water.
After a rapid transition in autumn, however, clouds in
other seasons were found largely behind the trough,
where cold outbreaks occur over relatively warm water.
Some doubt is expressed, however, as to whether these
results are entirely representative and whether they
can Tie applied to 5-day-averaged cloud amounts.
The upper wave pattern in the middle and high southern
latitudes revealed by mean monthly or seasonal data
does not exhibit well-defined troughs or ridges (Lamb
1959, Taljaard et al. 1969). However, a recent analysis of
the zonal harmonic standing waves over the hemisphere
using surface and upper data for the IGY (van Loon and
I
FIGURE
for
1800
1 .-(A) 5-day-averaged mosaic Nov. 10-14, 1969 and (B) location of 700-mb troughs in the long-wave, scale-separated components
corresponding indicated dates. Heavy lines represent well-defined troughs, fine lines represent minor troughs, and striped lines rep-
resent troughs in the mean 5-day field.
Jenne 1972) has revealed that waves 1 and 3 have signifi-
cant standing components, and that wave number 3 is
particularly prominent between latitudes 40' and 60's
with ridges near the three low-latitude continents. Further
inferences to frequently occurring maxima and minima
of trough and ridge frequency may be obtained by study
of 5-day means of pressure and geopotential derived from
synoptic chart series (e.g., Noar 1973) by a statistical
procedure using multiple regression with specific cloudiness
indices to estimate 700-mb heights around particular
latitude circles (Staver 1969), and by inference from the
pattern of formation and movement of satellite-observed
cloud vortexes associated with depressions (Streten 1968a,
Streten and Troup 1973).
Techniques for the separation of so-called inherent
scales (Holl 1963a, 1963b, Riegel 1965) enable realistic
approximations to the long-wave field to be obtained
from individual synoptic charts for cases where high-
quality basic analyses are available. The smoothing is
defined by
ZLw=Zo +c V"=Zo-Z,,
where ZLw is the long-wave field, 2, is the original field,
Zsw is the short-wave field, C is a constant, and CY is a
parameter representing the degree of smoothing.
An example of this technique is shown in figure 1 for a
5-day period (Nov. 10-14, 1969) when an optimum ob-
servational distribution for the first Global Atmospheric
Research Program (GARP) project was available and
when a reasonably well-defined cloud band pattern was
evident on the 5-day-averaged hemispheric mosaic. A
third-order smoothing program modified from that of
Nagle and Hayden (1971) was employed using a=3 for
an N23 grid (Le,, 23 points between the Equator and the
1
pole). This results in a 90-percent short-wave amplitude
reduction for zonal wave numbers greater than 7 at 60's.
Five individual, and an average of the five long-wave
scale-separated 700-mb, patterns were obtained from the
0000 QMT daily analyses. In this test case, the prominent
cloud bands are generally located close to or slightly
eastward of well-defined groups of 700-mb long-wave
troughs occurring in the daily scale separated fields.
To obtain a seasonal statistical relationship between
the position of the scale-separated long-wave troughs
and the band axes, one would need a substantial series
of high-quality gridded charts for the hemisphere coinci-
dent with suitable 5-day-averaged pictures-a situation
not yet realized in practice. However, synoptic experience
and the evidence present131 available suggest that there is
a close, though not necessarily unique, association be-
tween the cloud bands and the configuration of the upper
long-wave pattern.
The high frequency of the extratrcpical banding over
the Southern Hemisphere and the frequent apparent
spatial continuity from period to period in the 5-day-
averaged mosaic sequences is notable. This suggests that
examination of the band sequences (which are themselves
of practical interest) may reveal features of the hemi-
spheric circulation and its temporal variations. The bands
may thus act as markers of the broad-scale circulation
features over the data-void southern oceans.
3. HEMISPHERIC CLOUD BAND DATA
The data employed in the investigation of the ba,nds
were the 200 5-day-averaged brightness mosaics covering
the Southern Hemisphere for 1,000 days of a 3-yr period
(November 1968-October 1971). The source satellites for
June 1973 1 Streten 1 487
the period were ESSA 7, ESSA 9, ITOS 1, and NOAA 1.
The mosaics varied in quality and degree of continuity
throughout the period and were only of limited value in
winter a t latitudes higher than 404s. The pictorial data
are not strictly comparable from one period to uncther.
On individual mosaics, however, the regions of maximum
and minimum brightness can usually be reliably distin-
guished, and a geometric axis can be delineated for each
organized band that is bright in relation to the surrounding
areas of the mosaic.
To be included in the statistics, a cloud band must
intersect at least 20 consecutive parallels of south latitude
and must average a t least 5 O latitude in width. I n this
way: only the most marked bands were recorded. All
quasi-zonal bands, in particular those associated with the
sectors of the intertropical convergence zone (ITCZ)
sometimes located south of the Equator (Kornfield et al.
1987, Gruber 1972), that do not extend over a wide
latitude range are excluded, as are those zonal bands
sometimes found close to Antarctica. Figure 2 shows an
example of a 5-day-averaged mosaic and the band axes
ascribed to it.
A complication exists in a zone of varying width lying
between 70° and 90°E associated with the “daybreak” in
the hemispheric data. The mosaics in this region are often
difficult to assess, but the continuity and orientation of
bands across the area frequently enable a cloud band to
be delineated with some degree of confidence. Unfortun-
ately, as will be seen later, this particular region is one of
considerable interest in the seasonal variation of highest
frequency of Indian Ocean cloud bands.
4. GEOGRAPHICAL FREQUENCY
OF 5-DAY-AVERAGED BANDS
The frequency of occurrence of the bands based on 3-yr
data is given in figure 3 for four periods-summer
(December-March), “intermediate season” (April, May,
October, November), Winter (June-September), and
annual. The data are expressed as a percentage of mosaics
having axes of brightness maxima located within a 5 O
latitude by 10’ longitude sector of the polar stereographic
projection map. Several main features are apparent.
1. The distinct triple maxima of summer compared with the
more complicated pattern in the intermediate and winter months
over the Indian Ocean and Auatralasian redon.
2. The high frequencies in all seasons over the South Pacific and
South Atlantic. These are consistent with the depression tracks
and frequency of cyclogenesis in summer (Streten and Troup 1973)
and with the typical cyclone tracks of the International Geophysical
Year (IGY) (Taljaard 1967). It is intereating to note that as early
as 1930, on the basis of a wind analysis by Koppen, Bergeron (1930)
located a winter frontal zone in the South Pacific closely approximat-
ing the axis of the high cloudiness region now observed by satellite.
3. The apparent more westward location of the South Atlantic
maximum in winter and the higher frequencies in the eastern as
compared with the western part of the Indian Ocean in the same
season.
4. The apparent secondary maxima appearing a t lower latitudes
off the weat coasts of South America and South Africa in winter.
These are clearly associated with the stratus and fog forming in the
region of the Peru and Benguela Currents and are more prominent
in the colder months.
488 / Vol. 101, NO. 6 / Monthly Weather Review
FIGURE 2.-Five-day-averaged brightness mosaic with axes of major
cloud bands delineated by superimposed lines.
Individual seasonal maps (not reproduced) show great
similarity to the 3-yr averages of figure 3. If the axes of
maximum and minimum band frequency are drawn for
the 9 individual seasons of the sample period for the
region south of 2OoS, a pattern is established that enables
us to draw envelopes of these axes for particular geo-
graphical areas (fig. 4).
Prominent high-frequency axes lie from northwest to
southeast across the South Pacific and South Atlantic,
falling within narrow bounds. Equally prominent regions
of very low band frequency are enclosed by relatively
narrow envelopes over the southeast Pacific and South-
West Africa. Over the Indian Ocean and Australasia, the
band frequency is more complex. The Indian Ocean en-
velope is much broader in extent than that of the Pacific
and Atlantic with a much weaker detached secondary
maximum evident over south and east Australia. Such a
pattern is, in general, consistent with the location of the
ridges of standing waves 1 and 3 found by van Loon and
Jenne (1972) in the annual 500-mb map. The pattern over
Australasfa points to this region as being the major
location of change and readjustment in the band pattern.
5. SPATIAL AND TEMPORAL VARIATION
OF BAND FREQUENCY
Before discussing the individual characteristics of the
bands over particular ocean areas, some statistical features
of the band patterns as a whole will be presented.
Band Number
The number of bands (“band number”) on a particular
5day-averaged mosaic ranges from 0 to 7. As shown
in figure 5, however, there is a predominance of band
FIGURE 3.-Percentage frequency of 5-day-averaged mosaics having axes of major cloud bands within a 5O-latitude by lo0-longitude square
for (A) summer (December-March), (B) intermediate season (April, May, October, November), (C) winter (June-September), and
(D) annual. Data based on period November 1968-October 1971.
numbers 3 and 4 averaged over the latitude band from
20' to 5OoS, with band number 4 predominating from
30' to 40"s. This pattern may be compared with the
statistical analysis of 500-mb data for 1960-61 (Noar
1973), which indicated that wave number 4 was a princi-
pal mode of the broad-scale Southern Hemisphere cir-
culation, and with the four-wave pattern in the 700-mb
mean monthly trough and ridge locations at 40's for the
period January to May of 1967 (Staver 1969).
The-number of bands at particular latitudes on individ-
ual 5-day mosaics also displays some monthly variation.
Figure 6 indicates the percentage of the total number of
mosaics in each month having a band number 1 4 aver-
aged over latitudes 20°, 30°, and 40's. The period with
the smaller percentages is clearly in midsummer to early
autumn and the highest values occur in late winter and
spring. An apparent increase occurs in May with low
frequencies in June and July. However, the midwinter
period data must be regarded with some caution as the
observations are then of lower quality.
Longitude of Maximum Band Frequency
More detailed information on the location of the cloud
bands may be obtained by plotting in time section the
June 1973 Streten 1 489
FIQURE 4.-Envelopes of axes of maximum (solid line) and minimum
(broken line) frequency of cloud bands for the 9 seasons (3-yr)
between 20' and 60's. Figures along 40's are percentages of
seasons having maxima (minima) within each envelope.
30-
20-
%
10-
0-
monthly median longitude at which band axes intersect
particular parallels of latitude. Figure 7 shows such data
averaged between 30' and 40's intersections for four
separate zones; that is, the South Pacific (170'-70°W),
the South Atlantic (7O0W-2O0E), the Indian Ocean (20'-
120'E), and Australasia (120'E-170°W). The boundary
longitudes are those at which the band frequency ap-
proaches a minimum at midlatitudes. The longitudes are
sh0.m as an averaged annual cycle based on the 3-yr
data and as median monthly positions in time section for
the whole observation period.
lnterperiod Displacement of Bands
Examining the mosaics period by period, we observed
that the interperiod movement of the most prominent
cloud bands is usually fairly regular and a sequence can
often be followed as with daily cloud mosaics or synoptic
charts. Where at least three successive 5-day-averaged
mosaics were available and displayed well-defined bands,
the interperiod longitudinal displacement of the bands at
a number of latitude intersections was calculated. There
are, of course, many situations when no such clear pattern
occurs. To some extent, therefore, the calculated motions
are biased toward periods where distinct patterns are
displayed. Despite this subjective element, we believe
that the results represent a reasonable approximation to
the motion pattern. The distributions of such movement
data are shown for the previously defined zones at lati-
tude 40's for the whole period (fig. 8), for particular
months s t latitude 40's based on all zones (fig. 9), and
as 8. function of latitude (fig. lo).
In general, eastward interperiod movement tends to
predominate at all latitudes (fig. 10); westward displace-
490 / Vol. 101, No. 6 / Monthly Weather Review
LAT
(O S )
10
20
30
4 0
50
1 2 3 4 5 6 7
N
(A )
1 2 3 4 9 6 7
(B ) (C )
FIQURE 5.-Percentage frequency of (A) the number of cloud bands
on individual mosaics (band number N) a t particular latitudes,
(B) the particular band numbers averaged at 10'-latitude in-
tervals over the range 10'-60°S, and (C) the particular band
numbers averaged a t 10'-latitude intervals over the range 20"-
50's. Data for November 1968-October 1971.
70
2 0 7 J F M A M J J A S O N Q J
MONTH
FIQURE 6.-Monthly percentage frequency of mosaics (3-yr data)
having band number 2 4 (solid line). The broken line is the annual
average.
ment is more frequent in early autumn and winter (fig. 9).
It should be noted that the displacements of figure 10
are shown in terms of degrees of longitude measured on the
polar stereographic projection of the mosaic. Further,
the angle of intersection of the bands with the latitude
parallel, a!, is generally smaller at low latitudes than that
at higher latitudes, where the bands are more meridionally
alined. Thus, small changes in band orientation at 20'5,
for instance, may indicate a large, apparent displacement
along the latitude circle. Typical values of a! are shown
in figure 10.
6. BAND CHARACTERISTICS IN RELATION
CIRCULATION FEATURES
TO LONG- AND SHORT-TERM HEMISPHERIC
Since the behsvior of the bands within particular zones
with regard to seasonal and longer term location is
different, each will be examined separately.
SOUTH P A C I F I C SOUTH ATLANTIC INDIAN OCEAN AUSTRALASIA
J
F
M
A
M
J
J
A
S
0
N
0
I ..,..l .I
150 180
-
,..I
FIGURE 7.-(A) 3-yr average and (B) time section of monthly median longitude of 5-day-averaged bands for four geographical sectors aver-
aged between 30' and 40's. The shaded squares in (A) indicate months and regions with band frequency 2 1 per mo a t either 30'
or 40'5.
The South Pacific Zone
The median location of the Pacific band axes at mid-
latitudes moves little from month to month (figs. 3, 7A),
and the permanence of this broad-scale feature is reflected
in the limited longer term climatological data that are
available (Streten ..970). However, figure 7B indicate?
a slow trend in the band axis toward the west from the
spring of 1969 to that of 1971, the movement being about
1Oo-2O0 of longitude during this period. There is some
confirmation of this movement in rainfall data for two
island stations, Rapa (27OS, 144OW) and Pitcairn (25OS,
13OoW), which are given in table 1. The rainfall at Rapa
is slightly above the long-term average during the period,
but, as the band position tends to lie increasingly west of
Pitcairn Island, this station records many very dry
months with a large overall rainfall deficit. This situation
is probably associated at least in part with the westward
displacement of highest band frequency.
Figure 8 indicates largely eastward interperiod dis-
placements at 4OoS, the westward movement being of
lower frequency and smaller magnitude.
The South Atlantic Zone
The median location of the bands in the Atlantic is
between 1 5 O and 25OW in the latitude zone from 30' to
40's (fig. 7). A band is prominent in most multiday
imagery extending southeastward across the ocean. In
the mean (fig. 7A), there is some indication of a more
WEST& EAST
D ( Degrees longitude )
FIGURE 8.-Percentage frequency distribution of longitudinal dis-
placement of cloud bands a t 40'5 between successive 5-day-
averaged mosaics for the Atlantic Ocean (A), Australasia (As),
the Indian Ocean (I), and the Pacific Ocean (P) .
TABLE 1.-Rainfall Data for South Pacific Islands from October
1969 to September 1971
Station P(P) Ni Na
~~~
Rapa +6(2893) 6 4
Pitcairn Is. - 18( 1830) 2 9
P i s the percentage departure of rainfall total from long-term mean p (mm).
NI is the number of months with rainfall 250 percent above normal.
Na is the number of months with rainfall 550 percent below normal.
June 1973 1 Streten 491
506-256 0 - 73 - 4
J
A
S
0
N
~
D
30 20 10 0 10 20 30
WESTLEAST
D (Degrees Longitude)
FIGURE 9.-Percentage frequency distribution of longitudinal
displacement of cloud bands a t 40's between successive 5-day-
averaged mosaics as a function of time (month) based on 3 yr
of data for the whole hemisphere.
westerly location in winter at midlatitudes. Such a west-
ward location in this season appears to be consistent with
the winter synoptic situations on the Brazilian east coast.
(Trewartha 1962), particularly with the secondary fronto-
genesis and the patterns of coastal winter rainfall at low
latitudes in the region described by this author. The high
winter cloudiness in the coastal region is further revealed
in the data given by Miller and Feddes (1971).
The interperiod movement of the bands at 40's (fig. 8)
is more symmetric in the South Atlantic than elsewhere
with the distribution showing peaks a t from 5' to 10' of
longitude in both east and west displacement.
The Indian Ocean Zone
This sector displays a substantial (30'-40' longitude)
seasonal variation in the median location of the band from
the western part of the ocean in January eastward to the
vicinity of 90°E in late winter (fig. 7A) with a transition
occurring in the mean, quite rapidly in spring. Despite the
previously discussed difficulty in assessing the data in
this area, the displacement was similar in all 3 yr (fig. 7B),
and the pattern appears to be real.
Van Loon (1971) indicates a midlatitude association
between the location of the speed maximum in the sea-
surface current and that of the center of the mean sub-
tropical High over the Indian Ocean, both being farther
east in summer and west in winter. The variation in the
location of the anticyclone is further confirmed by the
total cloudiness pattern revealed in the 4-yr data of
Miller and Feddes (197 1).
I
j::
6 9
20 30 30 20 10 0 10
WEST& EAST
D (Degrees Longttude 1
FIGURE 10.-Percentage frequency distribution of longitudinal
displacement of cloud bands between successive 5-day-averaged
mosaics as a function of latitude based on 3 yr of data for the
whole hemisphere. Column on the right shows typical values
of a, the angle of intersection of the cloud bands with the latitude
parallel.
The band location is in opposite phase to that of the
anticyclone. I n summer, the band most frequently extends
from tropical Africa and Madagascar southeastward into
the ocean to the west of the High. In winter, when the
anticyclone is located farther west, the median position of
the cloud bands is t o the west of Australia in longitudes
85' to 95OE. I n the latter season, another band is often
located to the west of the High (fig. 3C), but with lower
frequency than that near 90°E.
The westward location of the median position of the
band in summer is consistent with the analysis of Lamb
(1959), who concluded that the summer movement of
the broad trough in the Indian Ocean westward to about
60°E was the only recognizable seasonal shift in the trough
ridge pattern that could be inferred from conventional
observations for the entire hemisphere. Other investigators
(e.g., Noar 1973), however, have found evidence for a
westward displacement in winter. The present data appear
to imply substantial frequencies of troughs in both the
eastern and western parts of the ocean in winter, but
with troughs in the east predominating.
It may be significant that, in summer, frequent advec-
tion of moist air is occurring from tropical Africa south-
eastward along the longitudes of the prominent band. I n
winter, advection of tropical maritime air frequently
occurs to the west of large, detached high-pressure sys-
tems centered over Australia. However, at this time, drier
air from continental southern Africa may be translated
over southern waters westward of the principal oceanic
anticyclone resulting in less-prominent cloud band devel-
opment. The striking differences in the cloud cover sta-
tistics from summer to winter over southern Africa are
evident in the seasonal data of Miller and Feddes (1971).
Examination of interperiod displacement data (fig. 8)
indicates a predominance of 5'- to 10'- longitude move-
ments to the east a t 40's.
The Australasian Zone
The bands in this region are less frequent and well-
defined than those over the major ocean areas (figs. 3, 4).
492 / Vol. 101, No. 6 / Monthly Weather Review
FIGURE 11.-Relative cloud band frequency and corresponding rainfall departure from normal for specific seasons in the Australian region.
Although the variation in monthly location is not notable,
the bands appear, in median position, to be farther west
in winter at the time of the farthest eastward displacement
of the Indian Ocean band (fig. 7). Thus, the distance
between the two bands is at a minimum at that season.
There is some indication of a slight eastward trend during
the 3-yr period (fig. 7B), thus shortening the distance
between the Australasian and South Pacific bands. Inter-
period displacements showing a maximum frequency east-
ward of less than 5 O of longitude are smaller than in the
other zones.
In the colder months, the rainfall over much of Aus-
tralia, particularly the southern half, results from the
passage of active depressions over the waters south of the
continent. The associated frontal bands (of ten displaying
lower latitude secondary developments) traverse the
continent from west to east with varying rates of motion
and frequent stagnation in particular areas. Examination
of the data on the frequency of 5-day-averaged bands for
individual seasons indicates considerable qualitahive
similarity between the pattern of seasonal rainfall anomaly
and that of band frequency. Figure 11 shows t,hese
corresponding patterns for 3 seasons. The relative band
frequencies are shown in terms of the percentage of
5-day-averaged mosaics having a band axis located
within the appropriate 5O-latitude by 10°-longitude
square. Rainfall departure from normal is shown as
derived from the reporting network of the Australian
Bureau of Meteorology (1969-1970). It appears that
even such a crude measure of the frequency of organized
June 1973 j Streten 1 493
cloud as the “band frequency” employed here may be
able to provide inferences to broad-scale seasonal anom-
alies of precipitation over the middle latitudes of the
southern oceans.
Some steps toward a more quantitative assessment of
climatological rainfall patterns from satellite data have
been taken (e.g., Barrett 1970), and the utilization of
digitized brightness data in conjunction with scme
measure of cloud organization such as the frequency of
5-day-averaged bands may enable broad working relation-
ships to be derived.
7. CONCLUDING RE
In general, the bands at 3Oo-4O0S include an apparent
half yearly component in the annual march of their median
longitude (fig. 7A). This half yearly component is in the
sense that when the subantarctic trough moves poleward
(van Loon 1967) the bands, on the average, tend to move
eastward ; when the trough moves equatorward, the bands
tend to move westward. This movement is not evident
in the Pacific zone, however, possibly because the trough
is about ‘7°-100 farther poleward in this region.
In this description of the frequency of 5-day-averaged
cloud bands and the interperiod, seasonal, and longer
term characteristics of their motion, the principal results
are as follows:
1. The band pattern is found to display a high frequency of band
numbers 3 and 4. This is in general agreement with the analyses of
the modes of long-wave pattern for the hemisphere as deduced
from conventional data.
2. In contrast to the relatively stable median monthly locations
in the South Pacific and South Atlantic, bands over the Indian
Ocean display a large variation in position from summer to winter.
Such locations are in opposite phase to those of the centers of Highs
over the ocean and to the maxima of ocean surface current speed.
3. Apparent long-term westward motion of the median location
of the Pacific bands between 1969 and 1971 are reflected in the
rainfall pattern at Rapa and Pitcairn Islands.
4. Qualitative pattern agreement is found between the frequency
of bands and the rainfall anomaly over continental Australia for
three colder month seasons.
The nature of the midlatitude band structure is still
uncertain. However, the evidence suggests that their
location is, in general, related closely to that of the long-
wave hemispheric pattern. It is notable that the three
more distinct low-latitude termini of high band frequency
are located over Africa, South America, and, in particular,
the eastern tip of New Guinea-the eastern extremity of
the so called “maritime continent”. The bands may thus
visually represent the mean channels wherein energy flow
occurs into the midlatitude westerlies from these tropical-
continental areas of most active convection. It is interest-
ing that ATS 3 pictures for the South American region
give a distinct impression of the movement of cloud
features from the tropical Amazon region southeastward
into the persistent cloud band over the South Atlantic.
The mean bands may thus be climatologically analogous
to the bands connected with particular tropical cyclones,
studied by Erickson and Winston (1972), as channels of
poleward energy transfer.
494 j Vol. 101, No. 6 1 Monthly Weather Review
Xopefully, we may, in the near future, be able to test
and numerically reproduce the satellite-observed features
of the circulation described here by general circulation
models of the Southern Hemisphere currently being
developed in Australia.
ACKNOWLEDGMENTS
The author is grateful to Harry van Loon of the National Center
for Atmospheric Research, Boulder, Colo., for his helpful review
and suggestions. The assistance of G. Kelly of the Commonwealth
Meteorology Research Centre in the provision of the long-wave,
scale-separated field of figure 1 and of W. Kellas and Wendy Kenny
in the data analysis is gratefully acknowledged.
REFERENCES
Australian Bureau of Meteorology, Department of Interior, Sea-
sonal Weather Review-Australia, Winter 1969, Spring 1969,
Winter 1970, Melbourne, 1969-1970.
Barrett, Eric C., “The Estimation of Monthly Rainfall From
Satellite Data,” Monthly Weather Review, Vol. 98, No. 4, Apr.
Bergeron, Tor, “Richtlinien einer Dynamischen Klimatologie,”
Meteorologische Zeitschift, VO~. 47, No. 7, Braunschweig, Germany,
Booth, Arthur L., and Taylor, V. Ray, “Mesoscale Archive and
Computer Products of Digitized Video Data From ESSA Satel-
lites,” Bulletin of the American Meteorological Society, Vol. 50,
No. 6, June 1969, pp. 431-438.
Erickson, Carl O., and Winston, Jay S., “Tropical Storm, Mid-
Latitude, Cloud-Band Connections and the Autumnal Buildup
of the Planetary Circulation,” Journal of Applied Meteorology,
Vol. 11, No. 1, Feb. 1972, pp. 23-36.
Gruber, hnold, “Fluctuations in the Position of the ITCZ in the
Atlantic and Pacific Oceans,” Journal of the Atmospheric Sciences,
Vol. 29, No. 1, Jan. 1972, pp. 193-197.
Holl, Manfred M., “A Diagnostic-cycle Routine,” Final Report,
Contract Cwb 10314, Meteorology International lnc., Monterey,
Calif., Feb. 1963a, 37 pp.
Holl, Manfred M., “Scale and Pattern Spectra and Decomposi-
tions,” Technical Memorandum No. 3, Contract N228-(62271)
60550 Meteorology International Inc., Monterey, Calif., Nov.
1963b.
Kornfield, J. A., Hasler, A. F., Hanson, K. J., and Suomi, V. E.,
“Photographic Cloud Climatology from ESSA I11 and V Com-
puter Produced Mosaics,” Bulletin of the American Meteorological
Society, Vol. 48, No. 12, Dec. 1967, pp. 878-883.
Lamb, Hubert H., “The Southern Westerlies: A Preliminary Sur-
vey: Main Characteristics and Apparent Associations,” Quarterly
Journal of the Royal Meteorological Society, Vol. 85, No. 363,
London, England, Jan. 1959, pp. 1-23.
Leese, John A., Booth, Arthur L., and Godshall, Frederick A.,
‘(Archiving and Climatological Applications of Meteorological
Satellite Data,” ESSA Technical Report NESC 53, Washington,
D.C., July 1970, section 3, 23 pp.
Miller, Donald B., and Feddes, Robert G., “Global Atlas of Rela-
tive Cloud Cover 1967-70 Based on Photographic Signals From
Meteorological Satellites,” NOAA (NESS)/USAF (Air Weather
Service-MAC) Joint Production, Washington, D.C., Sept. 1971,
237 pp.
Nagle, Roland E., and Hayden, Christopher M., “The Use of
Satellite-Observed Cloud Patterns in Northern-Hemisphere 500-
mb. Numerical Analysis,” NOAA Technical Report NESS 55,
Washington, D.C., Apr. 1971, 25 pp.
Noar, Peter F., “Energy Dispersion and Other Features of the
Middle Latitude Circulation of the Australian Region,” Meleoro-
logical Study No. 24, Australian Bureau of Meteorology, Mel-
bourne, Feb. 1973, 96 pp.
Riegel, Christopher A., “An Interpretation of Holl’s Inherent
Scale Techniques,” Meteorology International Inc., Monterey,
Calif., Mar. 1965, 18 pp.
1970, pp. 322-327.
July 1930, pp. 246-262.
Saha, Kshudiram, “Mean Cloud Distributions Over Tropical
Oceans,” Tellus, Vol. 23, No. 2, Stockholm, Sweden, 1971, pp.
Staver, Allen E., “Dynamic Characteristics of the Southern Hemi-
phere’s Circumpolar Vortex and a Comparison With its Northern
Hemisphere Counterpart,” Ph. D. thesis, University of Wisconsin,
University Microfilms Inc., Ann Arbor, Mich., Jan. 1969, 113 pp.
Streten, Neil A., “Some Aspects of High Latitude Southern Hemi-
sphere Summer Circulation as Viewed by ESSA 3,” Journal of
Applied Meteorology, Vol. 7, No. 3, June 1968a, pp. 324-332.
Streten, Neil A., “A Note on Multiple Image Photo-Mosaics for
the Southern Hemisphere,” Australian Meteorological Magazine,
Vol. 16, No. 4, Melbourne, Dec. 19683, pp. 127-136.
Streten, Neil A., “A Note on the Climatology of the Satellite
Observed Zone of High Cloudiness in the Central South Pacific,”
Australian Meteorological Magazine, Vol. 18, No. 1, Melbourne,
Mar. 1970, pp. 31-38.
Streten, Neil A., and Troup, A. J., “A Synoptic Climatology of
Satellite Observed Cloud Vortices Over the Southern Hemi-
183-195.
Taljaard, J. J., “Development, Distribution and Movement of
Cyclones and Anticyclones in the Southern Hemisphere During
the IGY,” Journal of Applied Meteorology, Vol. 6, No. 6, Dec.
Taljaard, J. J., van Loon, H., Crutcher, H. L., and Jenne, R. L.,
Climate of the Upper A i r Part I-Southern Hemisphere, U.S. Navy
NAVAIR 50-1C-55, Vol. 1, Washington, D.C., Sept. 1969.
Trewartha, Glen T., The Earth’s Problem Climates, University of
Wisconsin Press, Madison, Methuen & Co., London, 1962, 334
pp. (See pp. 41-56.)
van Loon, Harry, “The Half-Yearly Oscillations in Middle and
High Southern Latitudes and the Coreless Winter,” Journal of
the Atmospheric Sciences, Vol. 24, No. 5, Sept. 1967, pp. 472486.
van Loon, Harry, “A Half Yearly Variation of the Circumpolar
Surface Drift in the Southern Hemisphere,” Tellus, Vol. 23, No.
6, Stockholm, Sweden, 1971, pp. 511-516.
van Loon, Harry, and Jenne, Roy L., “The Zonal Harmonic
1967, pp. 973-987.
sphere,” Quarterly Journal of the Royal Meteorological Society,
Vol. 99, No. 419, London, England, Jan. 1973, pp. 56-72.
Standing Waves in the Southern Hemisphere,’’ Journal of Gee-
physical Research, Vol. 77, No. 6, Feb. 20, 1972, pp. 992-1003.
[Received November 7, 1972; revised March 14, 19731
Urgent Notice to Authors and Subscribers
The National Oceanic and Atmospheric Adminis-
tration announces that the Monthly Weather Review
will become a publication of the American Meteoro-
logical Society beginning with Volume 102, No. 1,
January 1974. During 101 years of publication by
NOAA and its predecessors, the Monthly Weather
Review has served as a medium for meteorological
information originating both within and outside the
Federal Government; in recent years, as a research
journal, its contributors have been a representative
cross section of the entire meteorological community,
both national and international; now more than half
of the authors have non-NOAA affiliations. Under
these circumstances, NOAA has concluded that the
public interest would be better served by a Monthly
Weather Review under non-Government sponsorship.
The American Meteorological Society, as a major
scientific society and publisher of several important
journals serving atmospheric and hydrospheric
scientists throughout the world, is uniquely qualified
to continue publication of the Monthly Weather
Review. Under the editorship of Chester W. Newton,
the Monthly Weather Review will continue its basic
emphases on the meteorological topics of weather
observation, analysis and forecasting, and instrumen-
tation. For further information the reader is referred
to the August 1973 issue of the Bulletin of the
American Meteorological Society.
INFORMATION FOR AUTHORS
The American Meteorological Society announces
that henceforth manuscripts (Articles and Notes or
Correspondence) for the Monthly Weather Review
should be submitted to Dr. Chester W. Newton,
Editor, Monthly Weather Review, National Center
for Atmospheric Research, P.O. Box 1470, Boulder,
Colo. 80302 (phone: 303494-5151). Authors may be
members or nonmembers of the Society, and of any
nationality, but only manuscripts in the English
language can be accepted. Reference should be made
to current issues of the Journal of the Atmospheric
Sciences, Journal of Applied Meteorology, or Journal
of Physical Oceanography for guidance in the pre-
paration of papers. Authors’ institutions will be
requested to pay a publication page charge in
conformance with the current rate for all Society
journals.
INFORMATION FOR SUBSCRIBERS
The December 1973 Monthly Weather Review will
be the last issue available from the Superintendent
of Documents; he will make an adjustment on any
present subscription that extends beyond that issue.
Subscription orders beginning with the January 1974
issue should be submitted to the American Meteoro-
logical Society, 45 Beacon Street, Boston, Mass.
02108. The Society announces that the Monthly
Weather Review will be available on a calendar year
subscription-only basis at an annual rate of $60
($20 to AMS members). The single issue price will
be $6 ($3 to AMS members).
June1973 1 Streten 1 495