312 MONTHLY WEATHER REVIEW AUGUST 1958
THE WEATHER AND CIRCULATION OF AUGUST 1958'
A Month with an Unusual Temperature Reversal
CHARLES M. WOFFINDEN
Extended Forecast Section, U. S. Weather Bureau, Washington, D. C.
1. INTRODUCTION
Over North America the mean flow pattern during
August 1958(fig. 1) was characterized by pronounced
blocking in the extreme north and by a planetary wave
in relatively fast westerlies at middle latitudes consisting
of a ridge in the western United States and a trough in the
East. In contrast to July 1958[3], these features of the
westerly circulations were near their climatologically pre-
ferred locations for the month and hence closely resembled
the normal map but with slightly greater amplitude. They
wereresponsible for an unusually warm regime in the
West, which completely reversed the pattern of the pre-
vious month. A July-August reversal of this magnitude
has occurred only once before in the past 17 years.
2. GENERAL CIRCULATION
The blocking High over northern Canada and its deep
companiondepression across the pole accounted for the
two largest departures from normal (+410 f t . and -500
ft., respectively) which appear on the mean chart for the
month (fig. 1). These two features became established
early in the month, moved but little throughout the period,
and completely dominated the weather of the polar basin
and adjacent land areas. The polar Low was particularly
persistent and vigorous, with a departure from normal
at one stage in excess of 700 ft. (5-day mean map for Aug.
16-20,fig. 5). However, although a pronounced depres-
sionexisted at high latitudes, the circulation over the
remainder of Europe and Asia was weak and flat.
The trough normally found in the western Pacific was
weak and ill-defined, with the principal Pacific trough
lying well to the east between 150' and 160' W. This
corresponded roughly to its position during July [3] and
illustrates the tendency of this regime to persist, even
though the long-period climatological statistics indicate
that troughs are not at all favored in this longitudinal
band in midsummer [7]. Though the amplitude of this
feature remained small, its center averaged 240 ft. below
normal, and it maintained sufficient vigor to sustain the
downstream ridge in western North America and effec-
tively fU the northern portion of the west coast trough,
leaving the latter effective only at lower latitudes and well
offshore. This constitutes a significant change from July
[3] when this depression, though similarly confined to
southerly latitudes, was deeper and lay inshore along the
*&?e oherts I-Xm ib~owlug P. a28 for sndyzed climatological &ta for the m'mth.
California coast. As a result of this filling and retro-
gression, the cyclonically curved flow, which characterized
the July pattern in the West, was replaced by anticyclonic
streamlines in August which resulted in profound changes
in temperature and precipitation anomalies.
As mentioned, eastern North America and the Atlantic
were strongly influenced by blocking. Twin warm Highs
appear on the monthly mean pattern (fig. l), one over
northern Canada and the other over southern Greenland.
These are associated with depressed polar vortices in
eastern Canada and Iceland and the maintenance of a
broad band of strong westerliesacross the Atlantic a.t
temperate latitudes. The jet stream, as illustrated by
figure 2, reached a maximum speed of 14 m. p. s. at
45' N., 50' W., or 4 m. p. s. higher than the normal at
that point. This broad westerly wind stream and the
deeper than normal trough along the east coast, combined
to favor recurvature of tropical storms at sea followed by
rapid motion across the Atlantic.
3. HARMONIC ANALYSIS OF THE MEAN MONTHLY
CHART
In recent years increasing study has been given to the
spectrum of amplitudes and energy of meridional motions
on the scale of the general circulation. Graham [5],for
example, studied the latitudinal variation of the 17,800-ft.
contour of the 500-mb. surface and found that the sum of
the first three harmonics adequately describes the normal
map for January. He also found that the total of these
components remained nearly stationary on daily 500-mb.
charts during the month of January 1952. More recently,
others [4, 6,111 have dealt with energy spectra, of the
meridional component of the circulation. In general it
has been found that two distinct wave number bands exist
in the atmosphere, one incorporating waves one through
four, and the other wave number five and above. The
first class of waves is semi-permanent and may beasso-
ciated with the fixed topographic features of the earth's
surface, while the second is connected with the smaller
migratory systems.
In order to study the amplitude spectrum of meridional
flow for August 1958, a harmonic analysis of the 700-mb.
flow pattern was prepared, following the procedure
described in [lo]. The results are shown in figure 3 in
which amplitudes are plotted against wave number for
the latitudes 76' N., 60" N., 46' N., and 30' N. The pre-
AUQUST 1958 MONTHLY WEATHER REVIEW 313
FIGURE 1.-Mean 700-mb. height contours (solid) and departures from normal (dotted) both in tens of feet with troughs indicated by
heavy lines, for August 1958. The circulation pattern over North America included a ridge in the West and trough in the East.
The trough along the west coast was weak and effective only a t low latitudes.
eminence of the first harmonic at high latitudes is strik-
ingly illustrated by the curve at 75' N. (and to a lesser
extent by that for 60' N.). This reflects the existence of
the two large anomaly centers on either side of the pole
already described and indicates the degree of eccentricity
introduced by such a circulation. The normal circulation
for August similarly shows a maximum amplitude in wave
one at 75' N. but only of about one-third the magnitude.
At 60' N. the maximum amplitude was still associated
with the first harmonic, with a weak secondary peak
beginning to appear in the fourth harmonic. This second
maximum was also in evidence at 45' N. and 30' N. but
displaced toward the higher wave numbers (dashed line).
This characteristic dual maximum a t temperate latitudes
agrees roughly with the data of Eliassen [4] who found
maximum amplitude at wave number one and again at
wave number five, although his data were for a different
level (500 mb.) and period (October 21 to November 30,
1950). Conspicuous in the graph for 30' N. was the very
large amplitude accompanying the lowwave numbers.
However, interestingly enough, the normal also exhibits a
similar distribution, except that the amplitude of the first
harmonic is even larger, a surprising 167 ft. The phase
angle is such in each instance that it is largely attributable
314 MONTHLY WEATHER REVIEW AUGUST 1958
FIGURE 2.-Mean 700-mb. isotachs (in meters per second) for
August 1958. Solid arrows indicate the position of the primary
jet axis which roughly paralleled the northern border but dipped
southward over the Atlantic because of blocking to the north.
to the Asiatic monsoon and illustrates the marked eccen-
tricity which this phenomenon impresses on the low-
latitude circulation in midsummer.
The first four harmonics accounted for 91 percent of the
variability of the pattern for August 1958, and the addi-
tion of the next four increased this figure to over 99 per-
cent. For the normal chart for August, the first four
harmonics similarly contain most of the variance (97
percent), in agreement with the work of Graham [5], and
only an additional three are needed to increase this to 99
percent. Thus, waves one through four strongly dominate
both the pattern for the month and its normal. This
highlights the semi-permanence of these long planetary
waves as opposea to the random nature of the shorter
transient waves. However, th.e former are just those
components which are inadequately handled by simple
numerical models, which confirms the necessity of a differ-
ent approach to this problem, as recently emphasized by
Burger [l]. A step in this direction was taken by the
Joint Numerical Weather Prediction Unit at Suitland in
an attempt to take into consideration the special behavior
of these ultra-long waves [2, 121. Also it was this con-
sideration which led Namias [9] to incorporate the extrac-
tion of “latitudinal anomalies” as an integral part of his
“basic current” model.
4. RETROGRESSION O F 5-DAY MEAN FEATURES
As is often the case, considerable variation about the
monthly mean occurred, and one of the more interesting
trends discernible on 5-day mean charts was a long-period
retrogression. To illustrate this tendency, three 5-day
mean maps centered one week apart were selected (fig. 4).
- In the initial state, ,represented by the chart for ,August
25 e
20
15
I O ‘
1 5 ’
0 .
W
3
a
-I
z
IO
n
t 5
Q O
IO
5
0
IO
5
n - 0 I 2 3 4 5 6 7 0
WAVE NUMBER
FIGURE 3.-Amplitude in tens of feet of the zonal harmonic wave
components a t 75’ N., 60’ N., 45’ N., and 30’ N. for the 700-mb.
mean chart for August 1958 plotted against wave number. For
wave number one the amplitude was a maximum at both high
and low latitudes, the former reflecting the eccentricity of the
polar vortex and the latter the eccentricity imposed by the
Indian monsoon. A secondary peak appeared a t 60’ N. in the
fourth harmonic which was displaced toward higher wave
numbers.
7 to 11 (fig. 4A), a fast westerly flow characterized the
circulation with two pronounced troughs in the Pacific
and a third off the United States east coast.. One week
later (fig. 4B) the pattern in the Pacific underwent a major
transformation. The deep trough off Japan filledab-
ruptly, resulting in an extended wavelength between the
eastern Pacific trough and its new upstream counterpart
over eastern Asia. It adjusted to the new situation by
retrograding to the position shown in figure 4B, leaving in
its wake a cut-off and greatly weakened trough near the
California coast,. This was associated with a westward
displacement of the ridge-trough structure over North
AUUUST 1958 MONTHLY WEATHER REVIEW 31 5
America. This retrogression of the east coast trough
coincided with the periodwhen tropical storm Becky
was approaching recurvature and permitted that storm to
recurve off the coast in approximate conformity with the
normal track for August.
The Pacific trough continued its retrogression during
the third week (fig. 4C), and once again the downstream
features over North Americafollowed suit. By this time
the Canadian ridge was sufficiently far west to become
realignedwith the intensified maritime ridge in the
eastern Pacific rather than with the High in the Plateau
States as previously. As a result, the United States trough
shifted all the way back to the Mississippi Valley, which
led to the only outbreak of really cold air into the central
portion of the country. At this time, the Atlantic circula-
tion also reacted by reversing the tendency for eastward
motion of the Atlantic ridge. In fact, comparison of the
change in position a.nd strength of the Atlantic anti-
cyclone from figure 4B to figure 4C indicates a net west-
ward displacement of some 10". This sharp retrogression
of both the United States t,rough and the Atlantic ridge
was instrumental in influencing hurricane Daisy to recurve
farther west than either of its predecessors.
I t is noteworthy that, despite the overall retrogressive
tendency described above, the t'rough in the eastern
Atlantic succeeded in making eastward progress as far as
the European coast (fig. 4C). However, the resultant
overextended wavelength was subsequently adjusted by
rapid discontinuous retrogression of this feature as well.
5. TROPICAL STORMS IN RELATION T O
THE MEAN CIRCULATION
During August, three of the four storms which developed
over tbe tropical Atlantic became hurricanes. Except
for Ella, these fortunately remained at sea and recurved
offshore, in rough accord with the mean circulation for the
month(fig. 1). The f i s t tropical storm, Becky, was
detected near the Cape Verde Islands on August 10,
though its exact position could not be established until a
reconnaissance flight fixed the center about 20" east of
San Juan on the 12th. From that point the storm con-
tinued moving westward at a rapid rate, averaging over 20
knots over its entire course prior to recurvature and at one
time reaching a forward motion of 25 knots. As it came
under the influence of the mean trough along the east coast
it turned northward and recurved well offshore (Chart X).
Becky never attained f u l l maturity as a hurricane but
remained fairly weak over most of its long trajectory.
On the other hand, Cleo, the second storm, which
followed Becky by only two days, was already a severe
hurricane when first explored by reconnaissance on August
14 when about 750 miles east of the LesserAntilles.
Although reports were insufficient to establish a definite
track, it is probable that this storm had its origin in a
tropical depression to the southeast of the Cape Verde
Islands on the 11th. If &is is the case, it progressed
rather rapidly (about'il m. p. h.) over the first portion'of
PIGURE 4.-Five-day mean 700-mb. charts (in tens of feet) for (A)
August 7-11, (B) August 14-18, and (C) August 21-25,1958.
The filling of the deep mean trough in the western Pacific resulted
in retrogression of the west coast trough the following week (B)
and of the United States trough and Atlantic ridge the third week
(C). After the first period the west coast trough was weak and
confined mainly to low latitudes.
its journey, but deceleration must have begun near 45" W.
prior to the reconnaissance penetration of the storm since
thereafter it moved more slowly and shifted to a more
north-northwest.erly course (fig. 5).
The key transformation, accompanying this shift in
trajectory from that of Becky appears to have been the
eastward motion of the Atlantrc ridge and the splitting of
316 MONTHLY W E A T H E R R E V I E W AUQVBT 1958
the subtropical high pressure cell(fig. 5). The retrogres-
sion,described earlier as affecting the Pacific and the
United States, had not yet begun to influence the Atlantic.
As a result the ridge succeeded in making eastward prog-
ress, though in the face of a lengthening wavelength.
Subsequently (fig. 4C) this ridge did retrograde sharply,
but it remained far enough east just long enough for Cleo
to penetrate the subtropical high pressure belt. The
evolution of this development is of some interest. On the
5-daymean map for August 7-11(fig.4A) the western-
mostcell of the subtropical high pressure belt was the
stronger. Oneweek later, however(fig. 4B), the eastern
center had grown at the expense of the western one, and
two days later (fig. 5) it was definitely predominant. It is
quite possible that the storm itself contributed to the
anticyclogenesis and thus influenced its own steering
pattern.
6. THE TEMPERATURE REVERSAL
The outstanding aspect of the temperature pattern for
this month was the pronounced reversal from July which
took place over most of the western portion of the country.
Ordinarily, the reverse is true and the transition from July
to August has usually been characterized by marked sta-
bility of the temperature pattern. In fact, in a study based
on the period 1942 through 1950 and later extended to
1957, Namias [8] found persistence to be higher between
this pair of months than for any other pair during the year.
In his paper, persistence was expressed in terms of the per-
centage of stations at which the temperature did not vary
by more than one class (out of five). For the change from
July to August this statistic turned out to be 80 percent,
the maximum for the year. This August, however, only 59
percent of the stations fell within this interval. Since the
chance expectance of this event is also59 percent, no
significant persistence occurred. During the past 17
years, only August 1947 had a lower persistence tendency,
with 51 percent.
In August 1958, warming of two classes or more (cross
hatched in fig. 6 ) encompassed an extensive area from
the interior of California to Illinois, with maximum
change (four classes) over Nevada and western Montana.
The departure from normal of the monthly mean tem-
perature rose 9.5' and 9.2' at Billings and Havre, Mont.,
respectively, and 9.1' at Winnemucca, Nev., to cite the
three highest. These are unusually high figures and high-
light tbe unusual nature of the change whichoccurred
this month.
The key changes in circulation accompanying this re-
versal in pattern have already been described in section 2,
namely, the strong ridging in the West which changed the
Bow from cyclonic in July to anticyclonic in August and
relegated the west coast trough to a minor role offshore.
Temperatures along the Pacific coast continued well
above normal for the fourth successive month. Record
warmth for August was reported from several locations in
Oregon and from Seattle, Wash. Tatoosh Island experi-
enced the warmest August in at least 56 years, and record
or near record t'emperatures were the rule at most Cali-
fornia coastal locations. This prolonged period of warmth
along the Pacific coast has constituted one of the most
distinctive characteristics of the temperature regime so
far this year and has been linked to abnormally warm
ocean temperatures in previous articles of this series [3].
While a similar effect in all likelihood continued this
month, the heat wave in the West can be adequately ex-
plained in terms of the large-scale ridging in the West,
above normal sunshine, and southerly components of flow.
In general, temperatures in the rest of the country
remained nearer normal, ranging slightly above normal
over the Appalachians and eastward and slightly below
normal over the States bordering the western Lakes and in
AUGUST 1958 MONTHLY WEATHER REVIEW
"
FIQURE 7.-Departure of average surface temperature from normal
(" F.) for the weeks ending at midnight (A) August 10 and (B)
August 24. The abnormal warmth which prevailed during
the first period, especially over the Northern Plains, was followed
two weeks later by an intense cold snap affecting central portions
of the oountry.
the Ohio and Lower Mississippi Valleys. Departures from
normalwere small because a warm regime early in the
month was largely cancelled in the mean by a marked
outbreak of polar air in the eastern two-thirds of the
country toward month end. Figure 7A was selected to
represent the warm regime. During this week tempera-
tures were unseasonably high nearly everywhere but,
relative to normal, were warmest, in the northern Great
Plains and Southwest where temperatures ranged up to
6' above normal. Maximum temperatures in the northern
Great Plains rose into the 100's on the 8th and 9th) and
Bismarck, N. Dak. recorded 107' F., the highest there
since1949. For these States and for those bordering the
Lakes, where June and July had been unusually cool, this
was the f b t real taste of summer weather. The circulation
for this period (fig. 4A) was characterized by broad west-
erlyflow along the northern border with mainly anti-
cyclonic curvature. However, it will be recalled from
section 4 that the stage was about to be set for a large-scale
retrogression of planetary waves on 5day mean maps.
This was accomplished by the week of August 18-24, and
during this period (fig. 4C) the principal mean trough
became entrenched over the Mississippi Valley, bringing
483864--68-3
FIGURE 8.-Number of days in August 1958 with surface fronts
of any type (within squares with sides approximately 500 miles).
The band of maximum frontal activity extended from the Central
Atlantic States west-northwestward.
an abrupt termination to the hot weather in the Midwest.
The strongest and most widespread outbreak of cold air
of the month was initiated a t this time and penetrated
to the Gulf of Mexico. Figure 7B depicts the thermal
distribution which ensued. Temperatures dropped rap-
idly, and Madison, Wis. recorded 37' F. on the 25th, the
lowest temperature reached there in August during
the 80-year period of record. Near-freezing temperature
occurred in the region of the western Lakes, and early
season frost was reported in the higher elevations in the
Black Hills of South Dakota. Thus, after only a brief
period of summerlike weather in the Midwest, conditions
more typical of fall were suddenly reintroduced.
7. PRECIPITATION
In general the precipitation for August, while not nearly
as heavy as for June and July, was nevertheless adequate
over most areas. However, thc July pattern of heavy
rains and flooding through the Ohio Valley and Central
Plains continued to dominate early August. Flood stages
maintained in a number of streams in Illinois and Indiana;
some of these streams had been in floodsince about
midJune. A record 30.0-ft. stage was reached a t New
London, Mo. as a result of rains in excess of 6 in. in the
upper reaches of the Salt River. This exceeded the pre-
vious high reached in June 1928 by 1.2 ft. Thereafter,
however, these excessive rains terminated as the ridge
moved into the Midwest and brought warm drying winds
to the area.
During this period, when the subtropical ridge was at ita
strongest across the Atlantic and the Southern States
(fig. 4A), onshore winds and easterly wave activity brought
tropical rains to the Gulf States. In fact, this type of
activity was quite promineGt during August, and a number
of 5day mean maps (figs.4, 5) exhibit pieces of weak
318 MONTHLY WEATHER REVIEW AUGUET 1958
low-latitude troughs in the Gulf or along the Mexican
border. These also served to furnish a supply ofGulf
moisture so that frontal activity, which was fairly frequent
in the East-Central States (fig. 8), succeeded in releasing
substantial rainfall. In this manner a band of above
normal precipitation for the month was produced, extend-
ing from Texas northeastward to the Central Atlantic
States (Chart 111-B). One of the heaviest of these frontal
rains transpired in central and eastern Oklahoma on the
20th when 5 to 9 inches fell locally and caused considerable
flood damage.
Frequent shower activity was widespread in the South-
west and, as a consequence, rainfall exceededtwice the
normal over an extensive belt extending from California
to southern Oregon (Chart 111-B). This precipitation
resulted from a moist tongue which curved anticyclonically
around the western High and was sustained by Gulf
moisture which was transported across Mexico by a broad
current from the east-southeast.
During this period Blue Canyon, Calif. reported the
greatest, number of thunderstorms ever to occur in August;
Winnemucca, Nev. experienced the wettest August of
record; and heavy showers and even a fewflashfloods
occurredover the southern Sierras and desert regions
of California. On the other hand, Grand Junction, Colo.,
which lay in the dry tongue on the opposite side of the
mean ridge, (fig. l), experienced the driest August in
the station recordwhich dates back t,o 1892.
In the Northwest, which remained hot and tinder-dry
for the entire month, a serious range and forest fire hazard
developed. Scores of fires were reported in the mountains
of Oregon with several reaching significant proportions
beforsbeing brought under control. Also, lightning
started many fires in the Sequoia National Forest and
REFERENCES
1. A. P. Burger, “Scale Consideration of Planetary Motions of
the Atmosphere,” Tellus, vol. 10, No. 2, May 1958, pp.
195-205.
2. G. P. Cressman, “Barotropic Divergence and Very Long
Atmospheric Waves,” Monthly Weather Review, vol. 86, No,
3. C. R. Dunn, “The Weather and Circulation of July 1958-
Heavy Precipitation Associated with a Trough in Central
United States,” Monthly Weather Review, vol. 86, No. 7,
July 1958, pp. 268-276.
4. E. Eliassen, ‘‘A Study of the Long Atmospheric Waves on the
Basis of Zonal Harmonic Analysis,” Tellus, vol. 10, No. 2,
May 1958, pp. 206-215.
5. R. D. Graham, “An Empirical Study of Planetary Waves by
Means of Harmonic Analysis,” Journal of Meteorology, vol,
12, No. 4, Aug. 1955, pp. 298-307.
6. R. M. Henry and S. L. Hess, “A Study of the Large-Scale
Spectra of Atmospheric Kinetic Energy, Wave Speed,
Momentum Flux, and Heat Flux,” Technical Repolt No. 9
on Contract Nonr-l600(00), Florida State Universit,y, March
1, 1957.
7. W. H. Klein and J. S. Winston, “Geographical Frequency of
Troughs and Ridges on Mean 700-mb. Charts,” to be pub-
lished in Monthly Weather Review, vol. 86, 1958.
8. J. Namias, “Annual Course of Month-to-Month Persistence in
Climatic Anomalies,” Bulletin of the American Meteorological
Society, vol. 33, No. 7, Sept. 1952, pp. 279-285.
9. J. Namias, Application of Numerical Methods t,o Extended
Forecasting Practices in the U. S. Weather Bureau, unpub-
lished report of Extended Forecasting Section, U. S. Weather
Bureau, 1958.
10. H. A. Panofsky and G. W. Brier, Some Applications of Statistics
to Meteorology, The Pennsylvania State University, 1958,
224 pp. (See Chap. VI.)
11. R. M. White and D. S. Cooley, “Kinetic-Energy Spectrum of
Meridional Motion in the Mid-Troposphere,” Journal of
Meteorology, vol. 13, NO. 1, Feb. 1956, pp. 67-69.
12. P. M. Wolff, “The Error in Numerical Forecasts Due to Retro-
gression of Ultra-Long Waves,” Monthly Weather Review.
8, Aug. 1958, pp. 293-297.
other brush and forest coveredregions of California. Gal. 86, No. 7, July 1958, pp. 245-250.
A. Basedon reports from over 900 Weather Bureau and cooperative stations. The monthly average is half the sum of the monthly
average maximum and monthly average minimum, which are the average of the daily maxima and daily minima, respectively.
B. Departures from normal are based on the 30-yr. normals (1921-50) for Weather Bureau stations and on means of
26 years or more (mostly 1931-55) for cooperative stations.
A. Basedon reports from over 900 Weather Bureau and cooperative stations. The monthly average is half the sum of the monthly
average maximum and monthly average minimum, which are the average of the daily maxima and daily minima, respectively.
B. Departures from normal are based on the 30-yr. normals (1921-50) for Weather Bureau stations and on means of
26 years or more (mostly 1931-55) for cooperative stations.
AUGUST 1958
Chart 111, A. Departure of Precipitation from Normal (Inches), August 1958.
B. Percentage of Normal Precipitation, August 1958.
"
records of 25 years or- more (mostly 1931-55) for cooperative stations.
AUGUST 1958
-I , Chart VI. A. Percentage of Sky Cover Between Sunrise and Sunset, August 1958. ' :'
L ._
. , -B. Percentage of Normal Sky Cover BetweenSunriseandSunset,August 1958.
AUGUST 1958
:,
Chart VII. A. Percentage of Possible Sunshine, August 1958.
B; Percentage of Normal Sunshine, August 1958.
A. Computed from total number of hours of observed sunshine in relation to total number of possible hours of
sunshine during month. B. Normals are computed for stations having a t least 10 years of record.
AUGUST 1958
AUGUST 1958
AUGUST 1958
"
AUGUST 1958
"
AUGUST 1958
AUGUST 1958
J
\
\
\
\