312 MONTHLY WEATHER REVIEW AUQUST1959

THE WEATHER AND CIRCULATION OF AUGUST 1959 A Hot Month from the Central Plains to the Atlantic Coast

L. P. STARK Extended Forecast Section,U.S. Weather Bureau, Washington, D.C.

1. THEHEAT WAVE days}.Washington, D.C., and Los Angeles, Calif.air- INRELATION TO THECIRCULATION ports recorded their warmest Augusts of record. At Los Angeles airport monthly mean temperatures have been August 1959 was an exceptionally hot month for most warmer than the 1921-50 normal since April1956, a note- of the United States. Figure 1shows that inabout three- worthyclimatic fluctuation. Several cities had their fourths of the country, monthly temperatures averaged warmest August since the heat wave of 1947, and othen warmer than the seasonal normal. From the Dakotas to since 1937. Atthe other extreme,Miami International the Lower Great Lakes and from the Ohio Valley to the Airport had its coolest August of record. Middle Atlantic States, temperatures averaged from 4’ The upper-level circulation at 700 mb. for the month to more than 6’ above normal for the month. Cooler than is shown in figure 2. Consideringonly North Amer.ica normal weather was confined principally to a patch in and hhe adjacent marine areas, the immediate impression the Southern Plains States and to a belt extending from is one of a moderate zonal flow with the westerlies dis- the Pacific Northwest to partsof the southwestern deserts. placed well to the north. That condition is notuncommon Extremelyhigh temperatures did not typify this in the summer months when the subtropical ridge pene- August’s heat wave, but its severity was manifested in trates well .into the United States as an extension of the other ways. It is best described as one with successive Bermuda High. days of discomfort in which maximum temperatureshov- The August pattern was generally less complex than ered around 90°F., while humidities and nighttime mini- that of June [l] or July [2]. The dotted lines in figure mum temperatures were both high. 2 represent departures ofobserved height from normal All-time recordsfor number of days witha maximum of and thereby show the relative strength of major features. 90°F. or morewere reported by Cincinnati, Ohio (22 Of considerable importance to theweather in the United days), Newark,N.J. (15 days),and Boston, Mass. (11 States was the lack of amplitude of the wave pattern there.However, the trough in western North America was sufficiently intense to keep the West cooler than nor- mal as the eastern Pacific ridge propelled cool, maritime air into it. Confluence of cool and warm streamsof air in central Canada was associated with fast westerlies across Canada, thus precluding strong ridgedevelopment north of the United States border. Height departures in the downstream trough off the Atlantic Coast were negative only in its southern portion. Much of the United States mighthave been cool instead of hot had there been a mechanism to transport the cool Pacific air deeply into the country. But the combination of the trough inwestern North America, the ridge ineast- ern United States, and fast westerlies in Canada was not conducive to theinvasion of cool air masses. Even though FIGURE1.-Departure of average surface temperature from normal the ridge was only some 80 feet above normal, its per- (“F.), August 1959. Hatching indicat,es areas of normalor sistencepermitted a continuation of warm conditions. above normaltemperatures. Departures of 4” F. ormore were cotnmon from the Northern Plains to t8he Atlantic Coast. (From The broad southerly flow at sea level (see Chart XI of [4] Sept. 7.) [3] ) supplemented that at the upper level. Its presence

Unauthenticated | Downloaded 09/25/21 05:29 AM UTC MONTHLY WEATHER REVIEW 313

FIGURE2.-Mean 700-mb. contours (solid) and height departures from normal (dotted) (both in t,ens of feet) for August 1959. Over the United States and adjacent oceans the westerlies were weak or absent south of 45' K., and the warm, subtropical ridge had a great influence on the weather. also suggests warming and recurrent thrusts of tropical perature anomaly. In many respects the observed upper- air east of the Rockies. level mean for August 1959 closely approximated thecom- Figure 3 is a composite chart prepared from ten cases posite(compare figs. 2 and 3). of heat waves in the northeastern quadrant of the country The composite represents the type of flow in which the in June, July, or August selected from the years 1933 to polar front may be expected to lie along the northern 1958. For thisstudy a heat wave was defined asthe border of the United States east of the Continental Di- predominant occurrence of the much above normal tem- vide. The absence of a ridge in Canadaprecludes massive cold advection into the eastern half of the country. The 1 Temperatureanomalies are divided intothe following classes : much extensive tropical ridge in the southern and central States aboveand much below (12% percentoccurrence each)and above,near normal,and below (25 percentoccurrence each). and the intensified southerly flow insures repeated advec-

Unauthenticated | Downloaded 09/25/21 05:29 AM UTC 314 MONTHLYREVIEW WEA .THER AUGUST1959

FIGURE4.-Mean relative humidity for August 1959 baaed on the FIGURE3.-Composite mean 700-mb.contours (solid) andheight average of 0000 and 1200 GMT values.Very high relative departuresfrom normal (dotted) (both in tens of feet) for ten humiditiesdominated the country from the Mississippi Valley selectedcases of hotweather in thenortheastern quadrant of to the Atlantic Coast. the United States for summer months (June, July, August) from 1933 to 1959. The obviousfeatures common to allhot spells in theNortheast are: a strongridee in the Culf of Alaska,a trough along the west coast, an extension of the Bermuda ridge was thelower-latitude component of blocking. The into the Midwest, and below normal heights in Canada. higher-latitude component was evidenced by the + 220-ft. height departure from normal off Labrador. Thus, e'ast- erly onshore flow prevded along the Atlantic Seaboard tion of tropical air from the Gulf of Mexico and thus from New England to Georgia. perpetuates warm temperatures. Perha7ps as a consequence of the generally weak flow, Accompanying the heatwave was a regime of high rela- both observed andanomalous, temperature departures tive humidity. This too was largely a result of low-level from normal for the first half-month were rather small transport from tropical maritime source regions. Figure in most of the United States (fig. 6A). Exceptions in- 4 shows the monthly mean relative humidities of the 0000 clude the warm pockets in S0ut.h Dakota,New Mexico, and and 1200 GMT observationsmade duringAugust 1959. the coast of California. In the last instance mean tem- Note that va,lues greaterthan 70 percent (shaded) en- peratures were more than 6" above normal. Portions of veloped the country from the Mississippi Valley e.astward.the Pacific Northwest and Arizona were nearly 3" below Also of interest are the low relative humidities in Mon- normal.Cooler thannormal temperatures in the East tana and Wyoming, where the ground hasbeen especially were confined to coastal areas. dry all summer. -30, 1959 2. EFFECT OF CHANGESIN MEAN CIRCULATION Broadscale readjustmentof the circulation characterized ON TEMPERATURES INTHE UNITED STATES the la.& half of August in and near North America. Fig- ure 5B shows well-organized centers of a,ction and sub- AUGUST 1-15, 1959 stantial centers of height departure from normal. Anti- The mean circulation at 700 mb. over the United Statm cyclonic conditions in the Gulf of Alaska replaced the during the first half of August was generally quite weak, generally cyclonic conditions observed there a half month as shown in figure 5A. Except along the northernborder, earlier,with heights rising as much as 480 feet. Asso- where moderate westerlies prevailed, the gradient of '700- ciated with this strengthening ridge, a strong troughap- mb. height was seldom as much as 100 feet. Another fea- peared downstream over the western United States. ture of the chart, one which sometimes signifies an im- Marked cooling became associated with the trough in pending change, is the long wave spacing at 50" N. The the West, and widespread heating accompanied intensifi- next trough upstream from the one over Maine was at cation of the ridge in the East. The extent of the heat- ISSOW., a distance in excess of 90" of longitude. ing and cooling are portrayed in figure 6B. The trough along the eastcoast was of considerable Cold air penetrated into the interior valleys of Cali- importance to the weather in spite of its small amplitude. fornia where temperatures averaged more than 4" below Although not too apparent from thecontours, this trough normal, contrasted with 4" above normal during the first

Unauthenticated | Downloaded 09/25/21 05:29 AM UTC \

AUGUBT 1959 MONTHLYWEATHERREVIEW 315

FIGTJRE6.-Departure of average surface temperature from normal (OF.) for (A) August 1-15, 1959, and (B) August 16-31,1959. Hatching indicates areas of normal and above norm91 tempera- FIGURE5.-Mean 700-mb.contours (solid) and height departures ture anomalies.Warming inthe latter half of August was from normal (dotted) (both in tens of feet) for (A) August 1-15, greatestin the eastern Great Lakes, and widespread in other 1959, and (B) August 16-30, 1959. Notethe reversal of the areas east of the Rockies. circulation in the eastern Pacific and western North America as amplification occurred in the last half of the month.

No storm centers penetrated the Eastern States from August 16 to 31, thus permitting an uninterrupted heat half of August.During the week endingAugust 24, wave. The principal storm track was near Hudson Bay temperatures of 10"-20" F. below normal were responsible and was reflected aloft as the channel of negative height for new low daily maximum records [4]. departures in that area (fig. 5B). The positive height anomaly centernear thelower Great The temperate westerly index * is frequently a satis- Lakes (fig. 5B) was also the seat of the maximum surface factory reflection of the circulation. In August the be- temperaturedepartures from normal (fig. 6s). Lower havior of theindex wasreasonably descriptive of the Michigan and northern portions of Illinois and Indiana long-waveevolution nearNorth America.The index recorded mean temperatures for the last half of August reached a maximum of 9.1 m.p.s. (the highest since early greater than 9O F. above normal. It was during this pe- June) for the 5-day period ending . That was riod that Chicago equalled an existing record of 11 con- the termination of a generally rising index thefirst fifteen secutive days with 90" F. or above. Cincinnati had only days of the month. Coincident with amplification of the three days out of the last 16 days of August in which the trough in western North America the index started fall- maximum temperaturefell below 90" F. The Dakotas had ing, and by the end of the month reached4.8 m.p.s., more numerous readings above 100" F.-to as high as 112O F. than 2 m.p.s. below normal. at Vivian, S. Dak.-during the week ending August24 [a]. 2Computed at 35"-55' N. from 5' W.-17J0 E.

Unauthenticated | Downloaded 09/25/21 05:29 AM UTC 316 REVIEWWEATHER MONTHLY AUQUST1959 3. NON-PERSISTENCE OF TEMPERATURE ANOMALIES > A 16-year record shows that July and August aremore 4 90 persistent in of temperature anomaly than any other z terms a Y pair of months. Namias [5] foundthat in the period Lf :80 from 1942 to 1957 the change in temperature anomaly a w from July to August did not exceed one class at 82 per- a I cent of100 stations in the United States. The other 11 70

LL months averaged a change of one class or none in 67 percent 0 w of the cases. U 60 Temperature anomaly changes by class from July to + E August 1959 are shown in table 1. Persistence is defined Lo w a 50 as the total number of cities with no change in tempera- 19421943 1944 1945 1946 1947 1948 1949 1950 1951 19521953 1954 1955 1956 1957 1958 1959 ture class from the preceding month or with a change of AUGUST only one class (either warmeror colder). From the table PIGITRE7.--l'ersist,ence of t,emperatureanomalies (percent) from it can be seen that, with this defin.ition, August 1959 had Julyto August 1942-1959. Augustnormally has the highest index of persistence of only 54 percent,. This excep- pereistence of any month averaging between 70 and 90 percent. an In 1959 August was unusuallynon-persistent. Only August tionally low figure can be compared with all the Augusts 1947 R~Fless persistent. since1942 on the accompanying graph (fig. 7). From 1942 through 1958 only August 1947 (hot in the East) and August 1958 (hot in the West)had persistence values mean 700-mb. height was appreciably below normal. below the random curve. Even the long-wave troughs were relatively weak,and The geographical distribution of temperature anomaly some were actually above normal in height. The trough changes from July to August is shown in figure 8. ,4reas in western North America was the strongest in the West- of persistence are those without shadingor hatching. The ern Hemisphere, and the one east of the Urals was the wide hatching depicts changesof two or more classes and, most vigorous on themap. In general,polar latitudes to some extent, is indicative of the heat wave. The four- were under the influence of blocking, as attested by the class change in the Central Plains States is an area in widespreadpositive height anomalies inthat area. which much below temperature anomalies in July were Middle latitudeswere moderately zonal. For example, in replaced by much above anomalies in August. the western sector of the Northern Hemisphere, 700-mb.the Persistence in temperatures does not necessarily imply zonal index of temperate westerlies was 1-2 m.p.s. below a persistence of upper-level pattern. However, a month normal most of the month and increased to normal near in which the temperature anomaly is pal.ticularly non- the end of the month. persistent is almost always associated wit.h a major circu- although the Tropics were predominantry below nor- lation change from the preceding month. In the present mal in 700-mb. height, tropical storms were not especially instance a reversal in circulation in the lasthalf of August active. In theNorth Atlantic (including the Gulf of (discussed above) wasresponsible for a change of the Mexico andthe Caribbean) tropical developmentwas whole monthly pattern from July to August. slight in August,as only one tropical depression managed to develop into a tropicalstorm. An easterly wave 4. OTHERASPECTS strengthened in theLesser Antilles was designated Edith, OF THE CIRCULATIONAND WEATHER then weakened as it moved westward. A week later the easterly wave brought heavy rains to theTexas coast, Figure 2 shows that the Low in the western Siberian where t,he depression first deepened sl.ightly then weak- lowland was the only vortex on the map whose monthly ened again as it moved rapidly inland. The probability of tropical storm development in the North Atlantic in TABLE1.-Class changes of surfacetemperature anomalies in the Augustis 1.6 stormsper year according toDunn [6]. United States from July to August 1959 (A later computation based on nine more years of data Class change Frequency (percent) increased the average to 1.8 per year [7] ). ~- On the other hand, in the western North Pacific there +4 _...... "~~ "." ~~." 3 was greater than normal tropical storm activity. There +3 """.~~. ..~. ." ..~ ." 7

+2 "...... ~ "."" ~. 18 +1"...".... ~.~~..~."" 23 has been an average of 4.6 storms per year in August[6], o""...~~~.....~~.~.....~ 22 - l".....~~~~....~~~....~9 but, t,llis year 6 were found. Of those, there werefive -~~~~.~.....~~~....~~~~.~ 16 -3"..~...... ~~.~....~..~ 2 typhoons and one tropical storm. One hurricane in the

- ~ ~~~ 4"""- "" """" n easternNorth Pacific was destructive on theisland of

Unauthenticated | Downloaded 09/25/21 05:29 AM UTC AUQUST 1959 MONTHLY WEATHER REVIEW 317

Kauai, .. Rain at Lihue was 8.13 in. for the month, associated with less than normal rainfall. Perhaps more almost 6 in. in excess of normal. surprising was the liberal distribution of heavy rainfall Precipitation in the United States (fig. 9) not directly in many portions of the heat wave zone, since hot weather attributable tolocal tropical developments was rather cha- usually goes with dryconditions. otic, but many features could be partially explained by physical considerations implicit in the monthlymean cir- REFERENCES culation at 700 mb. (fig. 2).For example, rainfall 1. R. A. Green, “The Weather and Circulaltion of -A amounts in Oregon and Idahowere as much asthree times Month with an TJnusual Blocking Wave,” Monthly Weather normal under faster than normalwesterly flow in advance Review, vol. 87, No. 6, June 1959, pp. 231-238. of the mean trough lying near the coasts of Washington and Oregon. In a similar manner, upslope cyclonic flow at more northerly latitudes contributedto a total of 10.57 in. of rain at Annette, Alaska, which was more than 5 in. above normal. East of the Continental Divide the west- erlies produced foehn drying. In and Wyoming observed precipitation totals were less than 25 percent of the monthly normal (fig. 9) in a large area. Glasgow re- ported that August was its sixth consecutive month with below normal precipitation [4], and Billings recorded its driest May through August since 1894 [4]. A belt of generally heavy rain extended fromthe Dakotas to New England, and paralleled the axisof max- imum frontal activity in August, along which there were 15-25 days with fronts. Of additional interestis the mean flow (fig. 2) which shows confluence (andthus fronto- .- I genesis) in that area, along the bounda‘&, between tropi- FIGIJRE9.-Percenta-e of normalprecipitation for huqust 1959. Notethe discontinuous nature of precipitationgreater than cal and maritime polar air masses- Note that mean normalwhich occur! edsouth of the Lakesand east of the Rockies. 700-mb. height anomalies were positive, a condition usually (From [4] Sept. 7.)

Unauthenticated | Downloaded 09/25/21 05:29 AM UTC 318 MONTHLY WEATHER REVIEW AUGUBT1959

2. 0. R. Dum, 'The Weather and Circulation of "The 5. J. Namias, "The Annual Course of Month40"onth Persistence Second Oonsecutive July with an UnusualCirculation Pat- in C?lima8ticAnomalies," Bulletin of the American Meteorobgi- tern," MonthlyWeather Review, vol. 87, No. 7, July 1959, COZ Society, vol. 33, No. 7, Sept. 1952, pp. 279-285, and an nn- pp. 275-282. published extension through1954. 3. U.S. Weather Bureau, Climatological Data-National Sumary, 6. G. E. Dum, "Tropical Cyclones," Compendium of Meteorology, vol. 10,No. 8, Aug. 1959. (in press) American MeteorologicalSociety, Boston, Mass., 1951, pp. 4. U.S. Weather Bureau, Weekly Weather md Crop Bulletin, Na- 887-901. tional Xumarg, vol. XLVI, Nos. 31-36, Aug. 3, 10, 17, 24, 7. U. S. WeatherBureau, "Hurricane Forecasting," Forecasting 31, and Sept. 7,1959. Guide No. 3, Washington, D.C., ,180 pp.

CORRECTION Monthly Weather Review, vol. 87, No. 7, July 1959, p. 269: In column 2, first complete paragraph, change -99.9" C. to -90.9" C. The sentence should read, "The lowest tempera- tures registered during this ascent were : - 89.5" C. ( - 129.1" %'.) at 17,000 m. on the way up, and -90.9O C. (-131.6" F.) at 16,500 m. on the way down."

I1 6. GOVERNHEN1 PRINTING OIFICP~IOBB

Unauthenticated | Downloaded 09/25/21 05:29 AM UTC