The state averages presented in this paper showonly the mean conditions associated with tor­nadoes during the 8 year period of study. Theseaverages cannot be used to predict tornadoes butthey do illustrate- the weather conditions thathave accompanied past tornadoes.

FebT'Uar'Y 7,978

REFERENCES

Endlich, R. M., and R. L. Mancuso, 1968: Ob­jective analysis of environmental conditionsassociated with severe thunderstorms and tor­nadoes. Mon. Wea. Rev., V 96, pp. 342-350.

Inman, R. L., 1970: Papers on operationalobjective analysis schemes at the National SevereStorms Forecast Center. NOAA Tech. Memo.ERLTM-NSSL 51, pp.

ACKNOWLEDGMENTS

The assistance of Mrs. Sandra Boyse is appreci­ated for typing the manuscript and the work ofthe NSSFC computer staff is also gratefullyacknowledged. Williams, R. J., 1976:

sodated with tornadoes.pp. 540-545.

Surface parameters as­Mon. Wea. Rev., V 104,

THE SOUTHWEST LOW AND "HENRY'S RULE"

Walter K. HenryDepartm£nt of Meteorology

Te.~(Jf; A&M llniuersllyColll!geSlotion, 'rex<l$ 77843

ABSTRACT

The two synoptic conditions which precede the eastward movement of the Southwest Low and form the basis forHenry's Rule are described, Henry's Rule is that a stationary low over the southwestern United States will accelerate tothe east wOOn its circulation weakens, and another low arrives at a position about 2200 km to the northwest.

6

1. INTRODUCTION

On many weather maps of the United States anupper-level cydonic circulation Is centered overthe Southwestern States. This paper describes theoccurrence and behavior of these lows and pro­poses a means of forecasting their movement.

The original research of this paper was started atRandolph Fieldt Texas because of the adverseweather for flying that these lows caused and ourinability to forecast their behavior. The scope ofthe work was extended and was finalized in 1949for a thesis at the University of Chicago with Dr.Walter J. Saucier as committee chairman. Thethesis was entitled "On the movement of theSouthwest Low". Being in the United States AirForce and having a demanding assignment, publi­cation of a paper at that time did not seem to bevery important. Later, Mr. Lawrence A. Hughesof the Chicago office of the United StatesWeather Bureau started using the methods de­scribed in the thesis and "Henry's Rule" came intobeing. In January of 1962 the USWB reproducedthe thesis along with a few remarks by Mr. Hughesas OFDEV Technical Note No.6. Reference tothis Itrule" has appeared in many of the variousforecast discussions such as National Meteorologi­cal Center's twice daily map summary and the

regional critiques. The p:-ocedures have been usedfor several years by many forecasters in thesouthwestern United States. Now that theNational Weather Digest is trying to presentuseable forecasting methods it is time to publish"Henry's Rule ll

•

A Jow was classified as a "Southwest Low" if itscenter was wbthin the area bounded by 40

0N to

250 N and 100 W to 1250 W. The area was furthersubdivided into three sections, the mountainousarea 1n the center (the stippled part on Figure I),and the section east or west of the mountains.The 700 mb charts were used because they werethe best upper air charts dUring the period of 1940to 1949 when the rule was formulated. However,the reasoning and rule applies to the 500 mb map.

2. CHARACTERISTICS OF LOWS

The lows in this area are often "cutoff lows"indicating that they are separated from theprevailing westerly circulation of the hemisphere.At o,ther times the long waves may be of highamph tude and the low still embedded in thewesterlies. The lows are cold core lows. All lowswhich were in the area were considered andincluded in the final results.

Natioru:zZ Weather Digest

Figure 1. Typical 700 mb Chart with a South­west Low. Heights are in dekameters. Shadedpart represents the "over mountains area".

a. Climatology of the Lows

The distribution of the lows by month and bylocation is shown in Table 1. It should be pointedout that the area of the section west of themountains is about twice the area of each of theother two sections. Even with that considerationthe tendency for the low to remain west of themountains was noticeable. The position of eachlow center was transferred from each 700 mbchart to a topographic map. This showed that thelow areas tended to favor the regions of lowerterrain and avoid the elevated regions. Therewere very few cases with the low located overareas where the height of the terrain was greaterthan 1800 m. The locations were concentratedalong the coastal lowlands, and when the lowsmoved eastward their paths were between themajor mountain ranges. This indicated that themountains were an important factor in controHingthe location and paths of the Southwest Low. Theindividual lows often moved 950 km in twelvehours as the low crossed the mountains, a move­ment far in excess of that before or after itmoved over the mountains. Wintertime lows weremore frequent than summertime lows. Alsoduring the winter they normally moved eastwardand could be identified over the mountains, andeast of the mountains. They are an importantfactor in the circulation of the southwestern partof the United States since they are present abouttwenty-five per cent of the time during the winterseason. CyclonicaHy curved flow over this areawithout a closed circulation occurs a muchgreater portion of the time.

Summertime lows (June, July, August andSeptember) appeared as' closed circulations only tothe west of the mountains. In general, theymoved northward or southward along the coast,without any inclination to move eastward and

across the moun tains. Closed lows occurredinfrequently in the summer, but those that didoccur persisted for some time. It is noted thattwo lows in June lasted 3.5 days (d) while a low inthe winter lasted only about 2.5 d on the average.July seemed to be free of closed lows; however, ifa longer period had been considered they mightwell have appeared.

b. Paths of Lows------Approximately one-third of the lows started in theAleutian Low, moved southeastward along thecoast and down into the designated area. Thisprocess occurred when there was a trough of largeamplitude over the west coast. These lowsbecame quasi-stationary over Southern Californiafor two or three days, then moved eastwardrapidly. In this case the polar jet streamcontained troughs and ridges of large amplitude.A ridge usuaHy extended into the Gulf of Alaskaand the ridge line had a north-south orientation.The trough line was oriented north-south, ornorth-northwest to south-southeast. A ridge waspresent over the Mississippi VaHey but of lessintensity than the one in the Eastern Pacific.Figure I shows a schematic map of a low in theSouthwestern United States and the generalpattern.

The other lows could be classified into fourgroups, each containing about one-sixth of thetotal. One group of the lows retrogressed into theregion; that is, they moved southward fromMontana or Wyoming to a position over lowerCalifornia where they remained for an indefinitetime, then moved northeastward across the UnitedStates. The general circulation pattern for thisgroup was about the same as described in thepreceding paragraph. The only noticeable differ­ence was that the orientation of the trough linewas northeast to southwest. After the lowreached the southern-part of its movement thetrough line often became north to south in itsorientation.

TABLE 1

MONTHLY DISTRIBUTION OF SOUTHWEST L(l;'S 8'l SEcrION AT 700·mb LEVEL(Closed Centers only) (1940 to 1948)

Average Average ~~er~g:,,~~m~~r of D~~S L~ P" CentMonth Number No. of • of

P" Days p" West Om East HonthMonth L~ of Htns. Htns. of Mtns, LOW" Ed..lJU

Jan. 3.' 2.7S 1.57 0.72 0.46 29.6Feb. 2.S 2.20 1.10 .90 .20 19.6"'::If. 2.3 2.21 1. 29 .79 .l4 12 ,5Apr. l.O 2.84 1.83 .83 .17 9.4

H'Y 1.3 2.50 1.63 .'0 .37 10.8June .7 3.50 2.84 .67 .00 11.7July .0 .00 .00 .00 .00 .0Aug. .7 2.25 2.05 .20 .00 4.8Sept. .3 1. SO 1.50 .00 .00 .2Oct. 2.3 2.57 1.86 .S7 .l4 20.0Nov. 2.0 2.33 1.25 .92 .l7 15.6Dec. 2. , 2.50 1.45 0.80 0.25 20.8

7

8

Another group of lows had a path that was almostwest-east. They carne into t~e area from thewest, continued through the area, and moved outeastward. Seldom was their path south of 350 Nand these had a constant rate of move:nent. Theydid not have the pause over Southern Californiawhich characterized the two types discussedprev.!ously. The general circulation in these caseswas more zonal, and the polar jet was south of itsnormal position.

Some lows appeared to form in the area. Thesewere usually the true "cutoff lows". Once formedthey remained almost stationary for two or threedays before making any noticeable move",enteastward. The general flow pattern for the"cutoff" was a trough of large ampli tude and alarge pressure ridge to the west in the EasternPacific. After the "cutoff" was completed themain westerly current had less amplitude and~~~~~t1~~8~~d north of the "cutoff low", usually

The last group consisted of the remaining miscel­laneous cases. It included the few summertimelows which developed through the various methodsdiscussed above; after remaining a few days offthe coast they either filled or shifted northwardbefore moving eastward. South of 400 N there wasno tendency for these summer cyclones to movetoward the east. Also in this last group were afew lows that appeared to drift in from the south.After some time they drifted southward again anddisappeared. Because of lack of data the exactnature of these cyclones was not known, but itwas suspected that each was connected with adisturbance in the tropical easterlies. When thelow from the south moved into the area the mainwesterly current was usually zonal and located atabout 500 N. The large scale circulation patternwas different for the indi vidual cyclone cases.

A closed low pressure area in the southwesternUnited States usually had little movement for aperiod of 36 to 72 h. Then suddenly it movedeastward or northeastward rather rapidly. Attimes it may have appeared to "split" with onepart remaining stationary or retrogressing and theother part moving eastwardly rapidly; this process'lias associated with a quasi-stationary Ilattern ofthe westerly current. The sequence of eventsunder these circumstances revealed that theactual splitting was very rare. Usually a weakpressure trough moved in fro'" the northwest andformed a second low to the west of the originallow which was displaced eastward. Also theoriginal low moved eastward as a weak wave whilea new low formed to the west of the position ofthe former. This gave the appearance of awestward movement of the cyclonic center. In afew instances the low filled so rapidly that aftertwelve hours there was little trace remaining of aclosed cyclonic flow.

Feb!'UaPy 197B

TABU 2

!'!EAt; v~s '" THr. C%liTKAt. HEIGHT "I\'D INtni!llTYOF THE SOlInlIo'2St LoWS AT Tl-X 100-mb 1ZY!l

-Icentral H~41ht. II'l Illt ..ns ity II .v880 K.

Month W"51 of avo< l!:.~t O~i W.. st n£ av", Ene of!".oulltalru Hounealn. ""uncain HouDe.&1na HOIIn[.lnt Hoontalns

Jan. 2919 2973 2961 98.8 18.0f ..b. 2'l~" 29107 29':1')66. I

Ha~. 299",87 .~ as.o 84.1

2939 29'1') 86. J 71.0 ....A"r. '''''' :l0'.1 Jl,•• 79.2 37.2 42.7HOy 3019 30CG 3043 77 .1 71.6 6J.4June 3028 3075 , ... * "'.8 53. ~July ...Au,;;. 3072

.... .... .... ....317C .... 83.5 l2.2Sl'pt. 3042 .. ,. 81.4

, ...OCI, 102~ J026 JUb) 87.1

.... , ...110\1. 2~51

73.2 63,42959 2995 103.3 90.0Do<. 29')5 107.1t2930 2972 93.3 77.1 89.9 I

c. Mean Circulation of Lows

The average gradient around each closed low wastaken to be a measure of intensity. The radius of880 km was selected arbitrarily and the heightdifference from the center in four directions wasdetermined, averaged and called II H. The valueof II H 'vas 'cquivalent to the average gradientaround the low. (With additional computation therelative vorticity could be determined from theaverage gradient.) The intensity of each low wasdetermined every 12 hours (h) by this method. Theresults for each month are compiled in Table 2.

In considering these results the fact that onlyclosed lows were measured masks the results.Any low that weakened and formed a trough as itmoved eastward was evaluated only as long as ithad a closed circulation at the 700 mb level. Thevalues given for the sections over the mountainsand east of the mountains were somewhat higherthan they would be if a cyclone could be iollowedall the way across. A closed circulation weakenedas it crossed the mountains, as will be discussedlater, and usually became a trough. Only thestronger lows retained their identity whilecrossing the mountains; therefore, the averagevalues of II H fort: l·~ two eastern sections in Table2 were exaggerated relative to the average valuesfor the western section.

This table indicates that the lows had the lowestcentral height during the winter months. Inwinter the lowest central height was inland, whileduring the warm months the lowest central heightwas over the coast. This follows the seasonaltemperature distribution. In winter the surfacewas colder over the mountains, and in summer thecoast and ocean had the cool surface.

The intensity of the circulation also wasdependent upon the season, with the most intensecirculation being in winter. A value of lIH of 100m is equivalent to a geostrophic wind of 25 kt.Because of the cyclonic curvature the actual wind

NationaZ Weather Digestwas weaker. This was a mean value around thecyclone, and since the wind on the north side ofthe low was usually quite weak, this representedstronger winds on the other three sides. In somecases there appeared to be a sym metrical circu­lation around the low, but most of the time thestrongest flow was on the southwestern side of thelow.

Another point of interest was that the circulationintensity decreased over the mountains comparedto the strong closed cyclonic flow west of themountains. This was more noticeable in individuallows and will be discussed later in connection withthe forecasting of the movement of the low. Thecontinuance of the cyclonic circulation of the lowwhen it arrived east of the mountains wasdependent on the frontal features of the surfacemap. The conditions needed by the low to persistwill be discussed in conjunction with methods offorecasting the weather associated with the stormthat the low may cause.

3. WEATHER WITH LOW

With all weather events there may be greatvariations. The weather patterns associated withthe Southwest Low are variable also. West of themountains the low causes much cloudiness. Theheavy rainfall in Southern California in winter hasbeen caused by this low. With some lows the rainextended into Mexico. When 'the low moved overthe mountains it was usually weaker and movedrapidly so the weather was characterized byclouds and some light rain.

When the low encountered a strong ridge over theMississippi Valley it usually continued to weakenand moved around the periphery of the ridge.Little precipitation and some cloudiness has beenassociated with this condition.

Often however, the ridge was weak or a weaktrough extended southwestward from Hudson Bay.With this favorable condition a Southwest Lowcould intensify over the Great Plains. During thewinter months a polar front generally was alongthe east slopes of the mountains even if not shownon the analysis of the National MeteorologicalService. The Southwest Low generated waves onthe fronts. Then snow, rain or freezing rain fellover the Great Plains and the upper Mississippiand Ohio Valleys. The storm often developed intoblizzard conditions. In Texas, Oklahoma,Arkansas, and Louisiana thunderstorms occurred,followed by rapid clearing as the trough aloftpassed. Often in springtime tornadoes and duststorms occurred whenever the Southwest Lowentered the Plains States. In summary, manytimes severe weather of some type has formed asthe Southwest Low moved to the east side of themountains. Therefore timing of the movement ofthe low became important.

9

Figure 2. Path of low (solid line) and isallobaricminimum to the northwest (dashed line) for I I to15 January 1946, inclusive.

Code: I - II January:1500 GMT; 2 - 12:03; 3 ­12:15; 4 - 13:03; 5 - 13:15; 6 - 14:03; 7 - 14:15; 8­15:03; 9 - 15 January:1500 GMT.

4. FORECASTING THE MOVEMENT OF THELOW

The forecast problem is to decide when the lowwill start to move eastward across the moutains.Two synoptic conditions have a bearing on themovement, a) the intensity of the low and b) atrough approaching from the west.

a. The Intensity of the Low

It was found that the strong lows with large valuesof Li H did not move far, at least not eastward. Alow that was weakening tended to move eastwardrapidly. An example is shown in Table 3 andFigure 2. Because of the sparse data bothpositions of the lows could not be determinedexactly and some indicated movement may be dueto analytical techniques, but the jump eastwardfollowing point 7 was much more than variation ofanalysis. Also the decrease of 1I H by a factor oftwo was more than the whim of an analyst.

To consider all cases, note Figure 3. It iUustratesthe relationship of the 1I H value and the distancemoved during the subsequent twelve hours. Ascan be seen the strong lows move slowly. AlSoimportant was the decrease of 1I H as well as theactual value of 1I H.

b. A Trough to the West

It also was noted that whenever a low approachedfrom the northwest that the Southwest Lowmoved eastward. This series of events is shown inFigure 4 where the twelve-hour position of bothlows are shown. On the 700 mb map of 27October 1948, at 1500 GMT a low was centered

9

TABLE 3

RELATION OF INTENSITY AT 700~mb LEVEL AND MO\'EMJo:NT OFTHE SOUTHWEST L(}; OF 11 to 15 JANUARY 1946, INCLUSIVE

Februa:ty Z978

further study. In this case the katallobaric centerbecame the next Southwest Low and movedeastward on the 18th.

... ...

. .

-. -. -: ~~:- r-;;;---.",. . --- ....... "' - -.- -. I····· 4/'-"• a.:i..~:.:- • ••••~ I •••• • '" 41' •/••• -r _.,. .,,;••••••• :, .../ t!-. ...• "',nl - ••I iii· • .. ••• -. •I •••• -.-. •-, :.~ .... ,..... .

•"•• •.Ji'.. ,.~

.1 ••••:,. _.... , ..I • •

-\ • • J

'..:,. ", /

~

"glD

iI

.., 100>-t:VJzw....z

2000 600 1200

DiSTANCE KM

Figure 3. The relationship of the 12 h eastwardmovement to the intensity of the Southwest Low.

For each cyclone the location of the low centerand of the effective katallobaric center weredetermined from the last chart before the cyclonebegan to move eastward. The distance anddirection from the low to the isallobaric minimumwere obtained from the map. When the lows weregiven a common location it was found that thepositions of the isallobaric minima showed pro­nounced clustering. This distribution is shown inFigure 5. The location of the points on the mapare not the exact geographical positions of thekatallobaric centers; they were the positions ofthe katallobaric centers with respect to theSouthwest Low. It was the relationship of theisallobaric minimum to the low that was im­portant. The actual strength of the katallobariccenter did not seem to have much effect. Thirty­four lows (all that occurred in 19lf6 and 19lf7except during the summer) were checked. Allthese lows had a katallobaric center located tothe northwest when they began to move. Onlyonce did a katallobaric center appear in theproper relationship to the low, and the low did notmove. That time the t:, H value of the low wasgreater than 160 m/880 km, a very high value (seeFigure 3).

0.------------....,

SubsequentTWel'le-L1uur Movemenr

Code Time ;H

Fil;;ure D', "''' 11./880 KillDist"nce

K" Direction

\ 1\ 1'i00 74.7 685 SW, 12 0300 128.0 75 5

1 12 1JOO 152.4 HiS SW

4 11 0300 147,9 75 S

5 13 1500 lL.0.2 165 '"E6 14 0]00 115.9 18'i E, 14 1500 51.8 556 '"EB 15 IUOO 54.9 "" "

just west of Salt Lake City (position I). This 101/1moved southwestward and in twelve hours wasover eastern Nevada (position 2). On the samemap a closed low had moved around the top of theridge in the Pacific and was just west of theCanadian coast (point 2). This Jow moved south­eastward (points 3, If and 5) and the low in Nevada(positions 2 and 3) subsequently moved rapidly tothe northeast (positions 3 and If).

In most cases the disturbance approaching fromthe Gulf of Alaska was not strong enough to havea closed center. Also since the troughs weresteered around the northern periphery of thePacific anticyclone and because of the dominantanticyclone curvature they usually were of smallamplitude while in the Gulf of Alaska. Inaddition, the scarcity of data in that region madeit difficult to follow these troughs precisely.Therefore, an objective alternative was taken tolocate the center of acti vity of those small waves.Since in the belt of the westerlies the position ofa twenty-four hour isallobaric minimum is a goodapproximation to the position of the associatedpressure trough, it was decided that the locationof the katallobaric centers be investigated. Theisallobaric fields were determined from the ana­lyzed 700 mb charts by subtracting the contourheights on one chart from the heights of a chart2lf h later. This was done at 12 h intervals for theeastern north Pacific area whenever there was aSouthwest Low. The centers of pressure fall(katallobaric centers) and the centers of pressurerises moved in the general direction of thesteering current.

When the data from two years (l9lf6 and 19lf7)were tabulated it was found that the results wereabout the same as if a closed cyclone coming outof the Aleutian Low could have been tracked.When the katallobaric center approached from thenorthwest, the low moved eastward. An exampleof this is given in Figure 2.

This low was one of a series that formed duringJanuary 19lf6. Cyclones moved eastward from theSouthwest on the lfth, 8th, 10th, [lfth and 18th ofJanuary 19lf6. The low shown in Figure 2 "backedin" from Colorado and its path is indicated by thesolid line. The location of the low is shown each12 h. After it arrived in the vicinity of the Gulfof California, its movements were insignificantuntil after Ilf January at 1500 GMT. Thekatallobaric center in the meantime had beenmoving eastward across the North Pacific. Its 12h positions are indicated in Figure 2 by a dashedline for the period of 13 January at 0300 GMT toIlf January at 1500 GMT, inclusive. When thecenter of the low and katallobaric center had therelati ve positions shown on the map of Ilf Januaryat 1500 GMT (point 7 on Figure 2) the low beganto move eastward. Such a relationship betweenthe Southwest Low and the upstream katallobariccenter was evident in other cases, and worthy of

10

National Weather Digest •

5. SUMMARY

In conclusion the following outline for forecastingthe movement of the lows that occur over thesouthwestern part of the United States and theadjoining PacificOcean is presented,

Figure 5. Location of isallobaric minima to theSouthwest Low when the low started to moverapidly. The + marks the position of eachisallobaric minima.

cyclone as the original low moved into the areaeast of the mountains. If the former low lost itscirculation and became a weak wave, then the lowappeared to "jump" westward, especially if thetime interval between charts was greater thantwelve hours. The low that formed was a new lowand usually not a retrogression of the former. Ifthe new low formed while the former low was stillover the mountains their outer circulations mighthave combined, and as the low to the east movedacross the mountains it often appeared that thelow split.

In a few cases when the katallobaric centers weremoving extremely fast they caught the SouthwestLow, and then a double low did form. This was anunstable condition and the "split" took place withthe low to the east moving very rapidly, forming atrough of small amplitude. The few lows thatoccurred during the summer did not move east­ward while south of 40oN. For this reason theypose problems only to forecasters on the westcoast.

Figure 4. Path of lows of 27 October to 2November 1948 inclusive at the 700 mb level.

Code: 1 - October 27:1500 GMT; 2 - 28:03; 3 ­28:15; 4 - 29:03; 5 - 29:15; 6 - 30:03; 7 - 30:15; 8­1:03; 9 - 1:15; 10 - 2 November:0300 GMT.

The positions shown in Figure 5 are well clustered,especially if the accuracy of the analysis over thispart of the ocean is considered. Also a tIme lagcould enter into the relationship of the katal­lobaric center and the low. The low could startmoving, even before the time of the map and notmove enough to give a positive indication ofmovement, or it could wait several hours after themap time to start moving. Therefore there couldbe up to eight or nine hours time differencebetween the different locations of the isallobaricminima. Even without these substantiating con­siderations the locations of the points clusterwell, and if more information were available theycould cluster even better. Therefore, wheneverthere is a low in the Southwestern United States aforecaster may expect it to move whenever akatallobaric center is located about 2200 km tothe northwest of the low.

On today's maps the positive vorticity maximummay be used in lieu of the closed low or thekatallobaric center. All three are indications ofcyclonic activity. The positive vorticity maxi­mum, the low, and the katallobaric center at the500 mb and 700 mb levels are not at the samegeographic ·location on all maps, if on any.However, given the accuracy of the analysis in theGulf of Alaska and the diversity of points onFigure 5, any of the indications will suffice.

The same technique was used for a period of 24 hbefore movement. The points were somewhatscattered and did not seem to be of much use.The isallobaric minimum had a rather erraticmovement.

It was a common occurrence for the katallobariccenter to move into the trough and form another

I. Determine the relationship of the katal­lobaric (or positive vorticity maximum)center to the northwest of\the low. If amoving trough or closed cyG!lonic circula­tion can be located, it may be used inlieu of the isallobaric minil)1um. When­ever any indicator has: the properrelationship to the low, that is, locatedapproximately 2200 km to the northwest(see Figure 5), the low should move.When the isallobaric minimum is some

11

distance away from that locatisn nomovement of the low should be forecast.This relationship is valid at 500 mb aswell as the 700 mb level. This relation­ship 1s referred to as IlHenry's Rule" andcan be determined by inspection of themaps. The next parts should be consider­ed also.

2. Determine the intensity of the circula­tion around the low. With an intensityvalue (ll H) greater than 100 m (25 kt) nomovement eastward would be expected.With values less than 60 eastward move­ment is likely. For the 500 mb levelthese values should be increased to 125and 85. A decrease of 1I H by a factor oftwo indicates that the low is ready tomove.

Febl'Ulwy Z9 78

3. A Southwest Low will intensify after itcrosses the mountains if there is afavorable cyclonic curvature of the con­tours in the area into which it moves. Ifa front exists along the east slopes of themountains expect a wave (Colorado-low)to develop. If a strong ridge is predomi­nant over the Mississippi Valley the lowwill be weakened. The extent of theprecipitati.on depends upon the intensityof the low.

ACKNOWLEDGEMENTS

The author wishes to express his thanks to Drs. A.H. Thompson, D. Djuric and Professor J. F.Griffiths for their critical reviews of the manu­script and their helpful suggestions.

/!NJijj]/

MESOSCALE LOW LEVEL VORTICITY CENTERS ASSOCIATED WITH CONVECTIONAS VIEWED BY SATELLITE

Edward C. JohnstonNuliuna/1'.'nuironmental Satellite Servu;e, SFSS

Kansas Cily, Missouri 64106

Several times during the spring and summer of1977, an interesting small scale cloud feature wasobserved in morning satellite pictures and persist­ed throughout most of the day. This phenomenon,which displayed certain similar characteristicswith each occurrence, apparently became a factorin significant weather developments later in theafternoon. Four examples will be shown here,with more detailed comments on the first two.This is by no means a complete study, but merelyan initial presentation of some cases noted duringthis particular convective season.

The most striking example occurred on September7 over Iowa and west central Illinois. A cluster ofthunderstorms over central Iowa (Figure la, takenat 1000Z), was gradually decreasing in intensity.By 1500Z, most of this convection had dissipated,leaving behind a small area of clouds oversoutheastern Iowa with tops to mid levels (FigureIb). As this cloud mass moved eastward, itgradually assumed a comma-shaped configuration,much like a well developed synoptic-scale vorti­city center (Figure Ie, taken at 1630Z). A

12

definite cyclonic twisting motion could be seen inthe movie loops at the time.

The 12Z LFM initial analysis (Figure Id) did notindicate anything of this nature over the area ofinterest. The prog for OOZ was equally unimpres­sive over l1Iinois (Figure Ie). (Barotropic and PEsimilar.) Furthermore, the morning 500 mb chart(Figure If), indicated anticyclonic directional andspeed shear, as well as marked diffluence over thearea.

Nevertheless, this cloud feature remained intactas it drifted eastward toward the l1Iinois border.By l8Z (Figure Ig), the first signs of newconvective development appeared near the Iowa­Illinois border, just downstream from the twistingcloud mass. Weather radar indicated tops to36,000 feet half an hour later. The succeedingchart depicted a small cluster with tops above40,000 feet just ahead of the apparent circulationcenter. The largest cell can be clearly seen onthe 20Z picture (Figure Ih).