THE WEATHER and CIRCULATION of SEPTEMBER 1957' Including a Discussion of Tropical Storm Activity EMANUEL M

SEPTEMBEB1967 MONTHLY WEATHER REVIEW 315

THE WEATHER AND CIRCULATION OF SEPTEMBER 1957' Including a Discussion of Tropical Storm Activity EMANUEL M. BALLENZWEIG

Extended Forecast Section U. S. Weather Bureau Washington 25, D. C.

1. COMPARISONWITH THE SUMMER MONTHS and precipitation. Only 56 of 100 stations in the United States changed temperature by one class or less compared The weather and circulation of the past summer was with a normal of 72; similarly, fewer than normal stat8ions marked byextreme heterogeneity. Themonth of June (20 instead of 35) remained in the sameprecipitation class was characterized by a mean trough through the central (last line of table 1). Part of the associated reversal in part of the United States with ridges along each coast [ti]. weather regime(figs. 2B and 2C) can be interpreted in A marked reversal in circulat,ion characteristics took place terms of the changes in 700-mb. height anomalies from between June and July [5] as a pattern more typical of August to September (fig. 2A). summer emerged with a trough off each coast and a ridge over theGreat Plains. The sharp circulationreversal 2. MONTHLY MEAN CIRCULATION AND WEATHER from June to Julywas followed by strongpersistence from July to August at middle Mitudes of the Western Rem- Thefairly well-defined temperatureand precipitation isphere [4]. patterns for t,he month of September (Charts I-B, 11, and The circulation experienced yet another oscillation in its 111-B) can be related quite well to the monthly 700-mb. transitionfrom August to September. Theplanetary height and height anomaly patterns (fig. l), despite the wave train became relnarkably akin to that of June 1957 changes in these patterns whichtook place duringthe [6], with a trough again in themiddle of the United States month (section 3).Temperatures as much as 5' F. and the coastal troughs of July andAugust supplanted by above normal occurred in thePacific Northwest and along ridges(fig. 1). The reappearance of the mid-continental the California coast under anticyclonic conditions, above trough was suggestedduring thelatter half of August normalheights, andabundant sunshine (ChartVII). (15-day mean,not shown), when the eastcoast trough The warm temperatures in the Pacific Northwest during fractured, with the southern portion retrograding toward September represented an extremely sharp reversal from the lower Mississippi Valley. the cool weather experienced there duringAugust (fig. 2B).

Despitethe large changesin hemispheric circulation Thiswarmth wasa consequence ' of upper-level ridge observed fromAugust to September (fig.ZA), blocking development along the coast (fig. 2A) and frequeslt anti- was still animportant feature of the circulation. The cyclonic passages in the vicinityof the Ilockies (Chart 1X) block that had beendominant in the westernCanadn- which kept the Northwest mainly in dry southerly flow Alaska area for the previous threemonths became a at sealevel. Warming was also experienced alorlg the strong ridge along the west c,oast of Canada, appearing as east coast, largelyas a result of anticy(4onic condit>ionsoff a closed High on many of the 5-day mean maps (fig. 3). the coast (figs. 1 and 2A). Blocking re-emerged in the Greenlandarea andin the The areafrom the Rockies to the Appalachians was cool Bering Sea, where700-mb. heights were 320 feet above (Chart I-B) as a consequence of frequentintrusions of normal and sea level pressures (Chart XI, inset) 7 mb. polarCanadian anticyclones (Chart IX). These anti- above normal. The high latitude block over Siberia in cyclones were steered rather well by the mean700-mb. August [4] movedsouthwestward into western Siberia, flow from thenorthwest. This area wasone of below and the deep polar vortex of August filled greatly as the normal 700-mb. heightsand cyclonic flow at 700 mb. polar region became a seat of above normal heights. Northerly anomalous flow from the Rockies to the hfissiu- Persistence from August to September, while clima- tologically not quite as marked asfrom July toAugust [7], TABLEI.-Persistence measures of monthly mean anomalies in the United States from dugust to September is usually much stronger than it was this year (table 1). - The -0.29 lag correlation of 700-mb. height anomalies 1057 Normal' Random over the United States is an indication of the reversal -_--__ 700-nlb. lielght (hg correlatlon) ".. .. ~-..-0. ?Y +O. 33 0. M fhattook place. The lack of persistencein the circula- Temperaturc (0 or 1 cl:rss change, permnt) __..._._~_._~_ 56 72 60 tion was reflected in large-scale changes in temperature Precipitation (0 class change, percent) ____..._._._._.___ 20 35 33 -. " 'Normal persistence measures of 7Wmb. height based on 1033-1950 data, temperature 1 See Charts I-XVII following p. 328 for analyzed climatological data for the month. on 1942-1954 data, and precipltation on 1942-1950 data.

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC 316 MONTHLY WEATHER REVIEW

FIQURE1.-Mean 700-mb. contours (solid) and height departures from normal (dotted), both in tens of feet, for September 1-30, 1957. Troughs for September are indicated by heavyvertical lines; position of trough in the Southeast in theperiod mid-August to mid-Sep- temher is indicated by a heavy dashed line. Dominant feature in the United States was the mean trough in the Mississippi Valley. sippi contributed tocool temperatures-averaging as much erly flowlowered temperatures by 2 classes (fig. 2B) in as 6' F. below normalin Wichita, Kans. The carving the area to the west of the Mississippi. The lowering of out of the trough in the Mississippi Valley is evident on temperaturesalong the Gulf coastand in part of the the anomalous change chart (fig. 2A), and the more north- Southeast,despite increased southerly flow,was partly the consequence of less sunshine than normal and many TABLE2.-A sample of theeffect of precipitation and cloudiness on temperature during September 1967 rainy days (table 2 and Chart VII), as record and near recordprecipitation fell inmany areas. (Charts I1 and Percent-Tempera- III-B). This precipitationwas influenced greatly by Station precipi- po%ble a&%y tropicalstorms plus abundant Gulf moisture moving (inches) sunshine* I (" F.) inland. I--"-- Birmingham, Ala ...... _.___.____.22 9.74 49 -1.3 Record rainfall in the South did not spread far enough Mobile, Ala ..._...... _.....__~~~~~.21 13.61 "_.."_... -2. 2 Pensacola, Fla...... _._..__.." -. 21 15.15 36 -1.3 northto ease thedroughty condition in parts of the New Orleans, La ...._..._....__._____26 10.90 36 "2.7 Jackson, Miss ...__...... __.______20 8.29 34 -3.4 Northeast.At Providence, R. I., which hadthe third Memphis. Tenn ___.._...______17 7.48 45 -2.3 driestSeptember on record, totalprecipitation for the 'See Chart VII-B for percent of normal sunshine. first 9 months of the year was less than 18 inches, the

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC SEPTEMBER1957 MONTHLY WEATHER REVlEW 317 least amount for that periodin 125 years. A similar record was set in Boston, Mass., which had its driest 9 monthsin 140 years. The desiccation in parts of the Pacific Northwest and through the Plateau States (Albu- querquereported only a trace of rainfor September) was the result of anticyclonic circulation and dry northerly anomalous flow aloft. Temperaturesalong the California coast were warm despit,e recordprecipitation amounts recorded in many cities. Theweather was prevailingly fairuntil the last few days of the month when a deep upper-level trough off thenorthern California coast broughtmoderate to heavy thundershowers with some hail to-this area. 3. TRANSITIONWITHIN THE MONTH Unlike the relatively stationary wave pattern of July and August, September was characterized by change during the month (fig. 3). The first 3 weeks showed a steady retrogression of themajor features of the circulation, followed by a rapidreversal in the fourth week.Con- comitantwith these permutations were changes in the temperature and precipitation regimes (figs. 4 and 5). FIRST WEEK In general, weekly temperatures averaged above normal from theDivide westward andfrom the Appalachians eastward, with below normaltemperatures reported in the rest of theNation (fig. 4A).Under a strong ridge (fig. 3A) temperaturedepartures as large as $6' to 4-7' F. occurredin Montanaand inland portions of southern California. Record maximum t'emperatureswere observed in Helena, Mont. (92' F.), Las Vagas,Nev. (108' F.),and RedBluff, Calif. (109' F. on Sept. 6 and 7). Las Vagas had 7 consecutive days with the temperature reaching or exceeding 100' F. Temperat,ures were above normal along the east coast as southerly flow prevailed east of thetrough in the Mississippi Valley (fig. 3A). Record maximum temperatures for the dat.e were established in Charlotte, N. C. (100' F.) and Columbia, S. C. (99' F.) on September 1. Heavy precipitation in the extreme Pacific Northwest was the consequence of anomalous onshore flow of mari- time Pacificair. A storm whichdeveloped inSouth - Dakota on September 1 (Chart X) deepened rapidly on &'IQURE 2.-Changes in monthly mean 700-mb. height departures from normal, August to September 1957. The isolines of anoma- the2d and 3d, and moved across the Northern Plains, lous height change (A) are drawn at 100-ft. intervals, with the accounting forthe precipitation observed in that area zero line heavier and the centers labeled in tens of feet. A large during the week and,in fact, for a large share of the reversaltook place from August to September, as reflected in monthly total there. Its passage was followed by a the number of classes the anomaly of temperature (B) and pre- series of cold outbreaksinto the United States, with cipitation (C) changed. Increased values are considered positive and decreasedvalues negative. Areas with increases of two or freezing temperaturesreported in northern and central more classes are cross-hatched;, areas with decreases of two or lowlands of Wisconsin on the 5th and subsequent dates morc classes arc stippled. and a reading as low as 23' F. in Cranmoor. New record minima were observed in such widely scattered cities as Erie, Pa.,San Antonio, Tex., Indianapolis, Ind., and on September 25(fig. 6). More will besaid aboutits Grand Junction, Colo. development and subsequent motion in sections 4 and 5. Perhaps the most important weathernews of the month Of more consequence to the weather of the United States was hurricane Carrie which formed on September 2 and was a weak tropical storm, Debbie, which formed in t,he was in the headlines until its demise near western Europe central Gulf of Mexico on September 7 and moved north-

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC 318 MOfiTHLY WEATHER REVIEW SEPTEMBEB1957

FIGURE3.-Five-day mean 700-mb. contours (in tens of feet) for selected periods in September 1957 one week apart. In A, B, C, and D, dotted lines show height departures from normal (in tens of feet). Retrogression during the first 3 weeks was followed by an abrupt reversal in the fourth week. Maps a, b, c partially overlapthe other set. The tracks of Calrie,Debbie, Esther, and Frieda (labeled C, D, E, F) at sea level are shown by arrows; open circles and date indicate 1200 GMT position. Frieda’s track from the 16th to the 20th is hypothetical......

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC " . SEPTEMBEB 1957 MONTHLYREVIEW WEATHER 319

FIGURE.4.-Departure of average surface temperature from normal(OF.) for weeks in September1957, centered on the 5-day mean periods shown in the left column of figure 3, and ending (A) September 8, (B) September 15, (C) September 22, and (D) September 29. Re- trogression of the pattern is discernible through the first three weeks, followed by a reversal. (From Weekly Weather and Crop Bulletin, National Summary, vol. XLIV, Nos. 36-39, Sept. 9, 16,23, and 30, 1957.) eastwardthrough northwestern Florida, southeastern norma.1 in Baltimore, Md.; and other cities along the mid- Alabama, and Georgia.No casualties or severe damage Atlanticseaboard reported temperature anomalies in were reported as the highest winds were about 40 m.p.h. excess of 4- 5'. This area was under the influence of a with gusts up to 50 m.p.h. along the northwestern coast of strongridge (with heights 200 feetabove normal) and Florida. Some flooding was reported due to heavy rain abundant sunshine.Record high temperatures were set and high tides. Tides were as high as 4 feet above normal inBoston, Mass., Providence, R. I., and New Haven; atthe Panacea Coast Guard Station in northwestern Conn. Florida, and stations in that area had rain in excess of 9 The polar outbreaks continued over the central interior inches. In general the precipitationassociated with United States as the boundary of below normal tempera- Debbie was about 24 inches,extending over anarea tures retrograded to the west of the Rockies. Tempera- from Florida to North Carolina (fig. 5A). The isohyetal turesfor the week averaged 10' F. below normal at pattern was typical of tropical storms, with the heaviest Cheyenne, Wyo., Grand Island, and North Platte, Nebr. rainfall to the right of the storm's path. As the ridge in the west retrograded, the area of above normaltemperatures diminished there. The largest SECOND WEEK temperaturedeparture reported was +6" F. at Seattle, The retrogression that took place from the first to the which had its highesttemperature of theyear on Sep- second week is manifest in the 700-mb. map centered on tember 13. the middle of the second week (fig. 3B) and the associated West of the Continental Divide another rainkss week weather (figs. 4B and5B), compared with that of the continued the dry weather regime (fig. 5B). In Washing- preceding week (figs. 3A, 4A, and SA). ton and Oregon dry weather combiried with above normal Above normalt,emperatures spread westward to the temperatures and low relative humidity to intensify the Ohio Valley, with the warmesttemperatures along the extreme fire hazard,and logging operations were sus- east coast (fig. 4B). Temperatures were 7' E'. above pendedin western Washington, Very littlerain fell

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC 320 REVIEWWEATHER MONTHLY SEPTEMBER1967

FIGUREB.-Total precipitation (inches) for the sameweeks and from the same source shown in figure 4. Heavy rain fell along the Gulf coast and in parts of the Southeast each week. anywhere to the west of the United States t'rough, except records set in Schenectady, N. Y., on the 21st and 22d. for light precipitation in the Rockies, including snow in Temperature anomalies were most extreme in the North- southern Wyoming and a 5-inch fall in Colorado. hloder- east, where a departure of +IO" F. was observed in ate to heavy rains fell over the central Great Plains in an Caribou, Maine, +9" F. in Concord, N. H., and +So F. area of cydoniccirculation, with amounts exceeding 3 in Ba.ltimore, Md. Precipitation was generally abundant inches inNebraska and Kansas. Heavy rain continued in the eastern half of the Nation in broad southerly flow along the Gulf coast as deep tropical moisture was ad- ahead of the mean trough (fig. 5C). vected northward. A tongue of heavy precipitat'ion Tropical storm Esther ent,ered the mainland over south- extended northward on the windward side of the Ap- eastern Louisianrt (fig. 6)) with winds of 50 m. p. h. and palachians into the Ohio Valley. More than 7 inches fell gusts up to 75 m. p.h. It was the third tropical storm in Chattanooga, Tenn., partly in response to the remnant,s this season to ma,ke landfall on the same narrow strip of of Debbie. Thundershowers produced some small pockets coastline. Esther was responsible for torrential rains of heavy ra,in along the mid-Atlantic coast. But a la.rge whichfell in squalls northeast of its center.Rain in portion of this arca, especially New England, cxpericnccd excess of 10 inches was not uncommon, and as much as 18 light preripit'ation under anticyclonic circulation. inches was recorded at Quarantine,La. Highest tides ranged from 2 feet above normal on the western Louisiana THIRD WEEK coast and near Pensacola,'Fla., up to 6 feet above normal Further retrogression of the circulation features, and along the southeastern Louisiana and Mississippi coasts. theattendent weather anomalies, took place from the Flooding from heavy rains occurred through Southeastern second to the third week of the month (figs. 3C, 4C, 5C). Louisiana and near the Mississippi and Alabama coasts. The mid-continental t,rough sheared and joined the Cali- Heavy rains with totals upto 11 inches fell on Oklahoma fornia trough as the latter intensified. The block in the and east,ern Texas,resulting in localflooding as a cold Gulf of Alaska weakened as it moved wcsttward. The front moved throughthe area. Cyclonic conditions are east coast ridge also weakened. apparent in that region on the mean map (fig. 3C). The eastern third of the, United States had continued A small tropical depression formed south of Louisiana hot weather (fig. 4C), with newmaximum temperature on the 16t'h andentered nort,hwcst,ern Florida, causing

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC SEPTEMBEK 1957 MONTHLY WEATHER REVIEW 321 heavyprecipitation in parts of theSoutheast. It is lies at Caribou, Maine, sliding from + 10" F. for the third possible that thisdepression was the beginning of hurricane week to "8" F. Temperaturedepartures from normal Frieda, whichdeveloped off Bermuda on September 20 were asgreat as -11" F. at such diverse stationsas (fig. 6 and Chart X). Esther and Frieda g-ill be discussed Youngstown, Ohio, and Greenville, S. C. Record minima in greater detail in the following sections. for the date, t.he month, andfor so early in the season were Warm and dry weather still prevailed over anarrow set atmany scatteredstations throughout the East, strip along the Pacific coast (fig. 4C and 5C), partly due to Midwest, and South. a dim.inished sea-breeze circulationas cold inlandtcm- With clear skies and abundant sunshine in association pcraturespredominated. The remainder of t,he Nation wit,h the High in the Southwest, the western half of the was unseasonably cool(fig. 4C) as polns anticyclones Nation experimced a heat wave of record-breaking pro- cont,inued t80taraverse the Great Plains (Chart IX). Dc- portions. SaltLake City, Utah, broke maximum tem- parturesfrom normal of the averagetemperature, were perature records on the last 3 days of t'he month. New - 11" F. in Montana, -9" F. in Minnesota, North Da- daily maxima were also set on those days and on the 27th kota, South Dakota, and Wyoming, and "8" F. in Ne- at stationsin Colorado, Montana,and Wyoming. The vada. Recordminimum temperaturesand killing frost most extreme weekly temperature departures were + 12" were experienced at stationsin Oregon,Washington, F. in Great Palls, Mont., and +11" F. in Salt Lake City. Montana, and Idaho. The record minimum temperature Temperatures for theweek were 5" F. below normal in a in Helena,Mont.., contmsted with record maximathere small pocketin theupper Sacramento Valley (fig. 4D). in both the first and fourth weeks of the month. Early in the week, it was warm, but, from the 26th through As a cold airmass moved slowly from the northwest and the 30th) thick clouds blanketed the area and kept maxi- an upper-level Low from the southwest, significant snow- mumtemperatures suppressed an average of 16" F., falls were recorded in the Rocky Mountain region with while the minima remained near normal. On 3 days the Helena, Mont., observing arecord-breaking September percentage of possible sunshine ranged from0 to 7 percent. accumulation of 13.4 inches and a 24-hour total of 9 inches. Drifts over 3 feet deepwere reported in Logan Pass in 4. TROPICALSTORMS RELATED TO THE MONTHLY Glacier National Park. CIRCULATION FOURTH WEEK ATLANTIC STORMS A major circulation change took place from the third t'o Of the four tropical stormswhich formed in the Atlantic the fourth week astroughs and ridges movedra,pidly and Gulf of Mexico during September two developed into eastward (fig. 3D).The meantrough in theUnited hurricanes, compared to mean values of three storms and Statessheared once more,with thesouthern portmion two hurricanes during the past 70 Septembers. Although becoming established along the coast of Lower California fouror more tropical cyclones have occurred slightly and the remainder marching eastward across the country more than one-fourth of the time in September, since 1953 tto join the low-latitude trough of hurricane Frieda. this frequency has been observed five times in succession. A troughremained in the Gulf of Mexic,o, extending This can be explained in part perhaps by increased data. northeastward into Tennessee, and was associated with Though conditionsfor tropical cyclogenesis were not heavy precipitationalong the Gulf coast. Therains the most ideal according to composite circulation types extended northward parallel to the trough in the sout'herlyoutlined by the author [l,21 the circulation in September flow to its east. At the surface a weak wave developed in was morefavorable than in the previousmonth. In the northwestern Gulf on the 23d and stagnated until the August, westerly anomalous flow prevailed over a large 26th, when it bcgan to drift slowly eastward (Chart X). part of thesubtropical Atlantic, and, despite favorable A smallarea of heavyprecipitation south of Hatteras sea-surface temperatures,only one storm formed. This during thefourth week was associated withawave storm developed in the Gulf of Mexico, which has been cyclone that moved northeastward parallelto that portion the preferred areathis hurricane season. Todate, five of the coast on the 29th-30th. The onlyother notable of the seven tropical storms (including two in September) precipitation was innorthern California andsouthern have developed in the Gulf. 1936 was the only previous Oregon,where record 24-hour falls occurred as a conse- year on record in which five or more storms developed in quence of the onshore flow and cyclonic conditions dis- the Gulf of Mexico. In that year a total of 17 tropical cussed in section 2. The rest of the country was dry as a storms developed inthe Atlantic region, so that even large High became established in the Southwest (fig. 3D), though 6 formed in the Gulf the concentration was not and a long fetch of northwesterly flow invaded the north- as great as this year. Conditions were favorable in the ern two-thirds of the Nation. Gulf inSeptember as easterly wave activity with con- The sharp reversal in thc temperature field (fig. 4D) is comitant cyclonic vorticity prevailed, and cooler than easily understood against the background of the circulation normalmid-tropospheric temperatures overlay warm changes. Cold weather was the rule in the eastern half of sea-surface temperatures [SI. the United States. Temperatures dropped sharply as two It is believed that the favorableclimate for tropical thrusts of cold air were, injected int'othe East,with anoma- storm formation includes warm sea-surface temperatures;

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC 322 MONTHLYREVIEW WEATHER SEPTEMBER1957

. /I. -?,,>.. \ J SEPTEMBER 1957 ~ ?“ 2’ \ ,,A, I FIGURE7.-Mean sea-surface temperatures (OF.) by 5’ squares for 6.-Preliminary tropicalstorm tracks for September in FIGURE themonth of September 1957. Numbers in lower right hand both the Atlantic and Pacific Oceans. The track of Frieda (F) corner of each box are the departure from normalin OF.; those in from the Gulf of Mexico to its “official” origin near Bermuda is upperleft give thenumber of ship observations (from twice- hypothetical. daily synoptic maps) on which the data are based. Water temp eratures were especially warm throughout the area.

Palmh [ll]set a criterion of 78O-8lo F. Since June 1957 allship observations of watertemperatures plotted on twice-daily synoptic charts have been tabulated for the of easterlyanomalous flow at lower latitudes (fig.8B). 5O squaresdelineated in figure7 (and for the squares This resulting increase inlatitudinal wind shear encourages extendingeast to Africa) and computed by 10-day and the fracture of extended troughs and the possible retro- monthly means. The mean temperatures for the month gression of theirsouthern portions. Observation has of September appear in the center of each square (fig. 7). shown the importance of a band of easterlies in the zone A comparison with the long-period (1887-1936) averages of formation [15], a conditionrealized this month as B given by Riehl [12] is in the lower right corner, and the continuous band of easterlies girdled lower latitudes from number of observations is given in the upper left corner Africa to the western Pacific. of each box. The behavior of the tropicalstorms (fig.6) can be It is apparentthat throughout September the sea interpreted in terms of the monthly circulation (fig. 1). surface was warmer than both the critical temperature Both Debbie and Esther moved roughly parallel to the of 81’ F. and the long-period average in the ocean area isopleths of 700-mb. height departure from normal. shown. Temperatures were extremely high from June to Namias [8] pointed out several cases of storm trajectories Septemberin the Gulf of Mexico,reaching a peakin behavingin similar fashion. In addition, Esther moved August; and they exceeded or equalled the threshold value northward just east of the mean trough for September, throughout the hurricane breeding ground of the Atlantic and Debbie moved northwardjust east of the mean during both July and August, in addition to September. trough for mid-August to mid-September (dashed trough Despite this, tropicalstorm activitywas suppressed during in fig. 1). Suchbehavior has been notedin the case of Julyand August. It isprobable that a diminutionin Audrey byElein [6],who listsmany other MonthZy sea-surface temperature below acritical value would Weather Review articleson theweather and circulation inhibit the formation of tropical storms,but that tempera- illustratingsimilar hurricane motion in relation to the tures equal to, or in excess of, the threshold value do not mean circulation. guaranteeformation unless synoptic conditions are also Therecurvature of Carrie also paralleled a mean favorable. trough. Thistrough in the western Atlantic was not This September the peak speed of the westerlies (jet), merely a reflection of the passage of the storm on 700-mb. computed geostrophically fromthe monthly mean700-mb. heights in the region traversed, since removal of the days chart (fig. l), was farther south than normal throughout when Carrie was in thatregion from the 30-day mean does the Atlantic (fig. 8A). Although it is popularly believed not completely e,liminate the trough (not reproduced). that when the jetis north of normal, tropicalstorm activity Carrie had an extremelylong life, traveling approxi- isgreat, and when south of normaldiminished, this mately 6,000 miles in 24 days,and was classified asa genera,lization is perhaps secondary in importance to the hurricane for 20 days from the Cape Verde Islands to a wind distribution tothe north of thenormal breeding point east of the Azores. The maintenance of the hurri- grounds. Note the belt of stronger than normalwesterlies cane circulation for so long a period is associated with at mid-latitudes of the Atlantic, surmounting anotherone its oceanic trajectory and points up many problems involv-

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC SEPTEMBER1957 MONTHLY WEATHER REVIEW 323 ing atmosphericvortices. Carrie reached itsgreatest intensityabout 1,000 miles east of Puerto Rico in the area where themonthly mean 700-mb. vorticity (computed from fig. 1, but not reprodumd) was much more cyclonic thannormal. (An indication of thisvorticity is the - 10 center in the height anomaly field of fig. 1.) At that time its winds were 160 m. p. h.; thoughhurricane- force windsdid not affect any land areas, Carrie pre- senteda hazard to NorthAtlantic shippingduring its long existence. On September 21, Carriewm associated with the sinking of the Germm sailing mssel Pamir with a loss of 80 lives. Hurricane Frieda reached its greatest intensity in an area of cyclonic vorticitynear the apex of the rnmn troughtraversed by Carrie. Its motionsubsequent to its intensification does not fit the 30-day mean pattern as well as it does the 5-day patterns (section 5). How- ever, features of themonthly mean that do appearto have influenced its path, and also that of Carrie, are the prevailing ridge along the east coast of the United States (fig. 1) and, particularly, the stronger thannormal westerlies over theNortheast (fig. 8) which deterredtropical storm penetration into that area [9, 11. EASTERN PACIFIC Of interest this month were the three tropical storms which formed in the eastern Pacific near Mexico (fig. 6), one of which was a hurricane. They formed in an mea of predominantly below normalheights on the mean 700-mb. chart, and two of them moved toward the north in thesoutherly flow ahead of themean trough. The third stormhad a forward speed of about 10 m.p. h. FIGURE K-Mean 700-mb. isotachs of the zonal wind speed com- A finalbulletin issued by SanFrancisco on the27th ponent (A) and its departure from normal (B) (both in meters stated that it was impossible to follow that storm beyond per second). Solid arrows in (A) indicate positions of the mean 15' N., 104' W., due to lack of reports in the area. Sub- axes of 700-mb. zonal wind maxima, and dashedarrows give their sequently a circulation was detected at 0000 on normalSeptember position. Centers of easterlyand westerly GMT flow are labeled E and W, respectively. Easterly flow is consid- October .1 at 19' N., 119O W. (not shown on fig. 6). ered negative and shaded. The hatched shading in (B) delineates DuringOctober thestorm developedinto a hurricane areas where the departure from the normal zonal flow equalled thatrecurved northeastward across Lower California or exceeded 2 meters per second from the east. Note the effect into Mexico. If stormnumber 3 (fig. 6), which could of blocking in middle and high latitudes, producing anomalous not betracked beyond the position given above, had flow components from the east. continued todrift west-northwestwardwith a speed of 11 m. p. h., it would have arrived at the point where the circulation was detected on October 1. It istherefore into the Aleutian Low south of a strong blocking High conceivable that these two storms were the same one. over Kamchatka. This Low was regenerated from time to time and acted as a sink for the cyclonic vorticity of WESTERN PACIFIC these typhoons and a source for the cyclone tracks leav- When this month began, Typhoon Bess (B' on fig. 6) ing that area (Chart X) and traveling across the top of had already formed. Five other typhoons were observed the blocking ridge in the eastern Pacific. in the western Pacific during September, although only four typhoonsformed. This paradox is theresult of semantics, since tropical storms with winds in excess of 5. TROPICAL STORMS RELATED TO THE 5-DAY MEAN 75 m. p. h. are called hurricanes east of 180' and typhoons CIRCULATIONS west of 180°, and a hurricane observed first on Septem- ber 4 at 163' W. was renamed typhoon Della (D') after The relation of the tropical storms of t8heatlantic to it crossed the 180th meridian on the 7th. the 5-daymean circulations deserves special attention. Four of the six typhoons observed this montb recurved Features that were not too clear in the previous section in the southerly flow between the Asiatic trough and the will bebrought into sharper focus. Each 5-day mean subtropicalHigh. Theso fourstorms recurved sharply map (fig. 3) will be discussed in chronological order with.

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC 324 MONTHLY WENTHER REVIEW SEPTEMBEB1957 Subsequent to its development on September 2, Carrie moved eastward southof a zonally oriented ridge (fig. 3A). The storm continued its westward march until the 10th when it encountered a low-latitude trough (fig. 3a) that had moved about 5O t’o the east from its former position (fig.3A) and influenced Carrie’s abortive recurvature. This low-latitude trough joined with a high-latitudetrough that deepencd asthe upstream wave train amplified. Carriedrifted north in t,he mean trough at about 5-10 m. p. h. during the period September 10-14 (fig. 3B). The strong ridge along t,he east coast of North America (fig. 3B) sheared during the next period (fig. 3b), and the westerlies over eastern Canada increased greatly in speed. As a result, the short wavelength between the mid-continental trough and the western Atlantic trough (fig. 3B) could no longer be supported, and the upper portion of theAtlantic trough sheared andamalgamated with a retrogressivetrough in the eastern Atlantic. In association with these fundamental changes in the circulation, a ridge crested across the Atlantic to the north of Carrie, FIGURE9.-Five-day mean 200-mb. contours (in hundreds of feet) for September 19-23, 1957, the period in which Frieda was impeding itsnorthward motion. Carrie turned to the detected.Although the position of Frieda on September 21 northwest on the 15th and on the 17th recurved sharply (designated by thetropical storm symbol) appearedfavorable as it came under the influence of the westerlies north of for deepening, it did not deepen until the 25th when a westerly the ridge. trough moved into the area. A mid-tropospheric cyclonic circulation over Nicaragua on September 12 drifted northwestward to the southwest- ern Gulf of Mexicoon the 15th. Tropical storm Esther respect to its relationto the tropical storm(s) of that formed underthis circulation at the base of the mid- period. continentaltrough. Relatively few of the Gulf storms Namias and Dunn [lo]hypothesized that the frequency deepen into hurricanes (less than one-third of the Septem- of tropical storms of the Cape Verde type depends on the ber storms), and Esther was part of the majority. As it degree of development of the Azores ridge to the north. moved northward just to the east of the trough, t’hebase They stated thatwhen the ridge of the Azores upper-level of the westerlies was apparently too low for major deep- anticyclone is thrust strongly northeastward into Europe, ening [14]. Esther moved inland on the southeastern cyclonic vorticity is injected into a trough off the African Louisiana coast about daybreak of the 1Sth and continued coast. They cited theantecedent conditions assoc.iatcd northward up the h,fississippi Valley, losing its identity with hurricane Connie in August 1955 as an example of on the 19th (fig. 3C). this type of development(their fig. 4). The author [16] On the 16th a weak tropical depression was located in hasnoted that conditionsfavorable for tropical storm thenortheastern Gulf of Mcxico, and it moved inland generationin the Cape Verde areaare such that large that morning (fig. 3b).This Low movednortheastward positive height anomalies appear near wcstern Europe- and eastward to the North Carolina-South Carolina bor- possibly a manifestation of a stronger than normalAzores der early on the morning of the18th. At t,hat timeit High extending farnorthward. The 700-mb. conditions was a frontalwave which progressed eastward by a method prior to the formation of Carrie on September 2 (not re- of discolltinuous cyclogenesis. It is thoughtthat this produced) were very reminiscent of those preceding Con- wave cyclone eventually deepened into hurricane Frieda. nie, and provide further evidence for thehypothcsis Dunn [3] hasindicated that trailing, stationary, or expounded by Namias and Dunn [lo]. fractured portions of old polar troughs may provide the On the morning of September 7, a weak cyclonic circu- initialconcentration of cyclonic vorticity necessary for lation, Debbie, developed in the central Gulf of Mexico hurricaneformation, a common occurrence inthe area as the result of juxtaposition of an easterly wave and a between theBahamas and Bermuda [16]. It wasmen- stagnanttrough aloft. Debbie intensified verylittle as tioned in section 4 that Frieda formed at the apex of the it moved northeastward in the southerly flow ahead of monthlytrough associated withCarrie; but these two the trough.Conditions were notfavorable for intense stormscan be even more closely linked. As Carrie deepening assouthwesterlies were abovethe storm in moved rapidly eastward (Sept. 17 to 25) it left a trough boththe middle and uppertroposphere [13]. OnSep- behind. The extratropical Low over the South Atlantic tember 8 (fig. 3a)Debbie went inland at Fort Walton, States deepened as it approachedthis concentration of Fla., with winds of 40 m.p.h. (associated wea.ther described cyclonic vorticitynear Bermuda, and the “official” in section 3). preliminary track places the origin of Friedain that

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC SEPTEMBEB1957 MONTHLY WEATHER REVlEW 325 vicinity (fig. 3c). It continued to movesouthward, 4. R. A. Green, “The Weather andCirculation of August steered by the large-scale flow which appears on the 5-day 1957,” Monthly Weather Review, vol.85, No. 8, mean (also paralleling the isopleths of heightanomaly) August 1957, pp. 282-287. (fig. 3c). Such a path, while uncommon, is not unprece- 5. H. F. Hawkins,Jr., “The Weather and Circulation dented; e. g., thefamous “Yankee”hurricane of 1935 of July1957”Drought in the East,” Monthly and a tropical storm in October 1938. Weather Review, vol. 85, No. 7, July 1957, pp. As Frieda continued to move southward and then loop 2 54-2 58. northward (fig. 3, C and c), it proved to be an exception 6. W. H. Klein, “The Weather and Circulation of June to therule [13] that anortherly current at 200 mb. is 1957-Including an Analysis of Hurricane Audrey favorable to deepening (fig. 9).This exception was inRelation to the Mean Circulation,” Monthly expecially noteworthy in view of the fact that the anti- Weather Review, vol. 85, No. 6, June 1957, pp. cyclone above Friedawas part of a ridge of the long-wave 208-220. pattern in the westerlies during a period of considerable 7. J. Namias, “The Annual Course of Month-to-Month amplitude. Persistence in Climatic Anomalies,” Bulletin of the During the period September 21-25 (fig. 3c) large-scale American Meteorological Society, vol.33, No. 7, retrogression ceased and eastward motion again became September 1952, pp. 279-285. established. Thiseastward motion continued atan 8. J. Namias,“Long-Range Factors Affecting the accelerated pace, and a deepened trough off the east Genesis and Paths of Tropical Cyclones,” Proceed- coast of North America swept Frieda rapidly northward. ings of the UNESCO Symposium on Typhoons 9-12 It was not until this trough at both 700 mb. and 200 nlb. November 1954, Tokyo, 1955, pp. 213-219. came over the surface cyclone that the storm deepened 9. J. Namias, “Secular Fluctuations in Vulnerability to to hurricaneintensity (fig. 3D). It appearsas if the Tropical Cyclones Inand Off New England,” intensification of Friedacannot be divorced from the Monthly Weather Review, vol. 83, No. 8, August baroclinic conditions that accompany the transforma t’Ion 1955, pp. 155-162. of tropical cyclones intloextratropical ones. Trough 10. J. Namias and C. R. Dunn, “The Weather andCircu- intensification notonly affects themotions of these lation of August 1955-Including the Climatological storms but also providesa mechanism for extratropical Backgroundfor Hurricanes Connie and Diane,” development. Monthly Weather Rewiew, vol. 83, No. 8, August Meanwhile (fig. 3c) Carriestreaked eastward under 1955, pp. 163-170. westerly mean 700-mb. flow. On the 21st the Pamir sank 11. E. Palm@“On the Formation and Structure of in the vicinity of thestorm, and on the 25th Carrie TropicalHurricanes,” Geophysical vol. 3, 1948, passed overIreland as strong gales lashed the coast. pp. 26-38. It helped to deepen the Low over northernEurope on 12. H. Riehl, Sea Surface Temperatures of the North the last map of this series (fig. 3D). Atlantic, 188?”1936, University of Chicago, Dept. of Meteorology, January 1956, 9 pp. plus 124 pp. REFERENCES of figs. (Multigraphed) 1. E. M. Ballenzweig, “Seasonal Variations in the 13. H. Riehl,“Formation of Hurricanes,”WMO Sym- Frequency of NorthAtlantic Tropical Cyclones posiumon Hurricanes,Ciudad Trujillo, February Relatedtothe General Circulation,” hrational 1956,14 pp.plus figs. (Multigraphed) Hurricane ResearchProject Report No. 9, U. S. 14. H. Riehl and N. M. Burgner, “Further Studies of the Weather Bureau, July 1957,32 pp. Movement and Formation of Hurricanes and Their 2. E. M. Ballenzweig, Relation of Long Period Circula- Forecasting,” Bulletin of the American Meteoro- tion Anomalies to Tropical Storm Formation and logical Society, vol. 31, No. 7, September 1950, Motion,unpublished report of U. S. Weather pp. 244-253. Bureau,Extended Forecast Section,September 15. H. Riehl and R. J. Shafer, “The Recurvatureof Tropi- 1957. calStorms,” Journal of Meteorology, vol. 1, Nos. 3. G. E. Dun, “Tropical Cyclones,” Compendium of 1-2, September 1944, pp. 42-54. Meteorology, American Meteorological Society, LG. U. S. WeatherBureau, “Hurricane Forecasting.” Boston, 1951, pp. 887-901. Forecasting Guide No. 3, (in preparation).

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC SEPTEMBER 1957 M. W. R. LXXXV-129

Chart I. A. Average Temperature ([IF.)at Surface, September 1957.

B. Departure of Average Temperature from Normal (OF.), September 1957.

average maximum and monthly averageminimum, which are the averageof 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 25 years or more (mostly 1931-55) for cooperative stations.

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC L2XXV-130 SEPTEMBER 1957 M. W. R. p b 1

C I- k

7 +n: C €-

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC SEPTEMBER 1957 M. W. R. LXXXV-13 1

B. Percentage of Normal Precipitation, September 1957.

Normal monthly precipitation amounts are computed from the records for 1921-50 for Weather hureau stations and from records of 25 years or more (mostly 1931-55) for cooperative stations.

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC SEPTEMBER 1957 M. W. R. LXXXV-132

Chart VI. A. Percentage of Sky Cover Between Sunrise and Sunset, September 1957

B. Percentage of Normal Sky Cover Between Sunrise and Sunset, September 1957.

A. In addition to cloudiness, sky cover includes obscuration of the sky by fog, smoke, snow, etc. Chart based on visual observations made hourly at Weather Bureau stations and averaged over the month. B. Computations of normal amount of sky cover are made for stations having at least 10 years of record.

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC SEPTEMBER 1957 M. W. R. LXXXV-133

Chart VII. A. Percentage of Possible Sunshine, September 1957.

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC LXXXV-134 SEPTEMBER 1957 M. W. R. 1

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC w 0 E

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC LXXXV-136 SEPTEMBER 1957 M. w. R.

(u 0

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC SEPTEMBER 1957 M. w. t(. LXXXV-137

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC Unauthenticated | Downloaded 10/11/21 09:01 AM UTC LXXXV-139

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC wUrv"lp0 SEPTEMBER 1957 M. W.R.

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC SEPTEMBER 1957 M. W. R. LXXXV-141

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC LXXXV-142 SEPTEMBER 1957 M. w. R.

Unauthenticated | Downloaded 10/11/21 09:01 AM UTC Unauthenticated | Downloaded 10/11/21 09:01 AM UTC