i ·

U. S. DEPARTMENT OF COMMERCE Luther H. Hodges, Secretary

WEATHER BUREAU F. W. Reichelderfer, Chief.

NATIONAL SEVERE STORMS PROJECT

REPORT No.7

The Vertical Structure of Three Dry Lines . ..

as Revealed by Aircraft Trave~ses

by

E. L. McGuire

Nap.onal Severe Storms Project, Kansas City, Mo.

W .. hingtOI1,D. C. April 1962 '

1. INTROOOCfION

CONTENTS

.......................................... " ........... ' ........................ .. ................ .. Page

1

2. INSTRUMENTATION OF AIRCRAFT ........ .. ............................. 2

3. METIIOD OF INVESTIGATION.. .. . .. . .. ... . . . . . . . . . . . . . . ... . ......• . .. . • 2

4. DESCRIPTION OF DATA ......... ...•.....•...........•................. 3

S. TIIERMAL AND DENSI1Y CDNSIDERATIONS ..•.........................•... 5

6 • OTIIER EXArdPLES ••••••.•••.•••.••••••••.•.•••••••.•.••••••••••••••••• 7

7. CDNCLUSIONS •................................ ' .' • . . . . . . . . . . . . . . . . . . . 8

REFERENCES •••••••••..••....•...••••.•.•.•••..•.•••••.•••.••.•.....•.•• 10

THE VERTICAL STRUCTURE OF THREE DRY LINES

AS REVEALED BY AIRCRAFT TRAVERSES 1

E. L. McGuire National Severe Storms Project

Kansas Ci ty.Missouri

ABSTRACT

Aircraft data. in conjunction with synoptic data. are analyzed for three cases of sharp moisture discontinuity near the ground. Aircraft traverses of the "dry line" yielded horizontal gradients of mixing ratio. of .the order of several grams per kilogram per kilometer. through the discontinuity surface. The discontinui ty surface was found to be nearly vertical through the lower

· 3000 to 4000 feet. No densi ty contrast across the dr~ line was apparent.

I . I NTRODUCTI ON

The terms "dry line" and "dew point front" have been used to describe a line. other than a warm or cold front . across which a sharp moisture discon­timrity .exists at the earth's surface. Discontinuity lines of this type are found most often in western Texas. where they form a boundary between the con­tinental tropical air mass of the southwestern United States and the maritime . tropical air flowing northward from the Gul f of Mexico. They are also some­times observed farther north through western Oklahoma. Kansas. and Nebraska. These lines are sometimes drawn eastward by migratory depressions and are of­ten observed in proximity to outbreaks of severe convective weather.

Fawbush. Miller. and Starrett, [2]. in their pioneer work on the fore­casting 6f tornadoes, considered the 550 F. surface dew point isotherm as a western boundary for tornado development. TIley considered the transition zone between moist and dry air as a zone of potential instability.

It is not well understood just what role the dry line plays in' the gen­eSIS of severe weathti. One hypothes.is has been that ' the verydr'y air on the west side of the line is more dense than tile moist air to the east and that there is a mechanical . lifting bf the moist air', as along a ~oldfront ; Oth~r forecasters believe that the effect is more · thermOdynamic in nature. · ;Sanders

. [5] . found a preferred region for the initiation 'of precipitation along the zone of strong gradients of potential wet-bulb temperature on the wes.tenl side · of moisture infections into the central Uniteq .S,tates. , ..

Ipresented at the Conference on Sevel'e. Storms .• Amedcan Meteorological Society, · St. Louis, Mo •• ' May lOr 1960.

2

It is known that a strong dry line may remain quasi -s tationary for several days without becoming active, but under certain conditions it appears to act as a generation line for actIve squall lines. Because of this apparent asso­ciation, one of the aims of the Tornado ~esearch Airplane Project 2 has. been to investigate the nature and structure of the dry line. One flight during which a dry line was traversed several times had been described by Beebe [1] and by Fujita [3].

2. INSTRUMENTATION OF AIRCRAFT

A description of the sensing and recording equipment used on the P-38 re­search airplane has been given by Lee and David [4].' The lag' in the beginning of response to sharp changes for both the'aspirating thermometer and the infra­red hygrometer was found to be less than two ~econds. Comparison flights in which the airplane circled the radiosonde balloon ,during ascent were made in 1958 and 1959. These showed good agreement in temperature and humidity meas­urement. They indicated that the infrared hygrometer used on the airplane is much more sensitive than the radiosonde hygrometer for mixing ratio values of less. than 12 gm./kg.

A part of the data frail the sensing instruments was recorded by photogra-. Jiling a modified instrument panel (photo panel). Temperature, pressure, and humidity were recorded in analog form on Brown Recorders. A Veeder-Root coun­ter was used for time synchronization. The counter number changed each 15 seconds. triggering the photo .. panel camera and'marking the recorder roll. A tape recorder was used to record remarks.by the pilot, which were prefaced by the current Veeder-Root number. Information on the tape recorder was synchro­nized with the Brown recorder through an ,electrical contact marmally actuated by the pilot.

3. METHOD OF INVESTIGATIDN

Two flights during the 1959 season, and one dur ing the 1960 season, were diSPatched wHh the primary objective of .dimensioning the dry line . Flights

. were conducted in av'ertlcal step patfern;wi th individual legs at constant 'pressurealtitude and ina path as nearly notm~il to the dry line as could be determined. When the sharp humidi ty change was noted, a point on the ground such as a road, bridge, or stream was chosen as a reference. On each succe­eding traverse, the pilot read the Veeder-Root cOl.mter number into the ~oice recorqer and also marked the recorder chart when directly over the reference point, Thus a vertica11ine was established. In the following diagrams ,this is labeled the "marker line". The pilot alsore~orded frequent fixes on either side of the marker line which aided in careful post-navigation of the flights. The accuracy of the post-navigatiop is 'iimited by the ability of the pilot to determine when he was directly over a gr,ound point and'a1so by the necessity .o'f using a mean ground spe,ed to locate some points. It is believed, however, that the error in lateral location of points is generally less than one-half mile close in to the marker line.

2Now expanded and renamed National Severe Storms Project.

~. DESCRIPTION OF DATA

Flight 37, conducted between 1503 and 1555 CST on May 26, 1959, yielded the strongest moisture gradients of the three cases and wi 11 be described in detail. Thp. sea-level chart based on 1400 CST observations (fig. 1), shows a weak wave cyclone over the Texas Panhandle with a line of sharp moisture discontinuity extending through this' center and southward. Note the marked drop in dew point from east to west across this line and the wind shift fran southerly to westerly. The point where the traverses were made is shown just NW of Mineral Wells, Tex. Only scattered cumulus and altocumulus clouds were present In th~ are~ of investigation.

1008

I·f JCT

I + 8~ r-- ____ , / 68

{ I. •

""I /\

\ 1010

\.

1400C 26 MAY 1959

LKC '85~ .

J

)

t 69~ . _', 101y.... ---.... _ ..-> C' \.

AOE / 88'\1 / 70/ ,.

7

Figure 1.- Sea-level chart for 1400 CST. May 26. 1959 showing location of dry line (da-sh-dotted line) at time Flight 37 was dispatched.

Figure 2 shows graphs of pressure and mixirig ratio against flight time for eleven traverses at nearly constant pressure altitude past the reference point. Notice that sharp discontinllities in moisture were observed below about 750 mb., while above that level the fluctuations were unsystematic. After the last traverse at 650 mb., 8. sounding descent was begun approxi­mately over the reference point:- Following this descent, three traverses were made at about 940 mb. All of these showed very·large moisture gradi­ents, the strongest rate of change being 6 g. /kg. in 33 seconds of flight time (about 1 1~ mi. ).

4

IZ - - - r - - - - - - - f - -- -- -, - - -- - T - - - - - - - - r - - - - - - - r - - - - -l- - - - - r - - - - - - - - - - , - - - - - - - - - - - - lZ

I 1 1 I I Martr~.~_ • .1 I I: I 10 - - - - - - - - - r ------J - - - -- - -I- - - - - - - - _C_ -----t - - - - - t- - - - -- ~ - - - - - - - -- - - - - - - - - - - - 10

8---~ ______ )� ____ J ---L------J------L----J-----L---J --- 1--- ------ 8 ~

:-:/, -~~~---- -~~~-~ ~~~j ~-------~~~--~-!---------~~-----L--- ~~~~ I ~::~ ::~~~~:: , +-------l- ------i-----r ------ -i ------~ --j/ --~~ -i -----f---- i--- ~ --, • ~ --t---- ---f -----t ---7Di~:f~\;a;~t-~:t --- -- -7----+----+----f- ----+- ---T------ 0

I 1 1 1 I I 1 ,I 600

700 1 : I 1 I 1 ~ --- I I ! :0 700 ! ; I I '---.;-- r- 1 1 1 I

800 I • ~ - '-----I ~ 1 I .1 ~ mi: I I I 1 ~ 800 _ I I I ·I- Iso.mi. I I I ::l

I I I ' , I . I I&J

900'r-~--------~--__ --~----L--------+-------+------+-----+---~~~--~----~----~~900 I I I II I I I '- - -

~Ir_~I----~I----~I----LI _____ L-I ____ L-I __ ~: __ ~I __ ~I~ __ ~I __ ~I~ __ ~ 1000 DRY- LINE TRAVERSES

FLIGHT #37 5/26/59

Figure 2.- Graphs of mixing ratio against flight time for eleven traverses through dry line. Flight 37. May 26. 1959. Time -and space sca1.e indicated in inset.

::0 E

M~~r 600.---------~~------------------_,

70'0/------

5 N. "'I. West 0 N. Mi. East 5 10

FLIGHT #37 1503-1555 CST 26 MAY 1959

Figure 3.- Cross_ection analysis of moisture distribution. Flight 37. DOts and short arrows .rep­r.esent path of airplane and small circles locate ' points where data were read. Isopleths :repre-. sen t mixing ra tio in g. /kg.

5 In figure 3, across-section analysis of the moisture pattern is given.

The path of the aircraft is indicated by the short arrows, and the dots locate points where data were read. The interval between 6 and 8 g./kg. seems to be the interval of most rapid change and was chosen as representing the dry-line boundary zone. This boundary can be recognizeq up to near 750 mb. and is sharpest below .850 mb. We observe a nearly vertical boundary for the 'lower 100mb. and a slope to the east at higher pressure altitude. Notice that near the ground the boundary appears to be displaced somewhat to the east and is more diffuse.

5. THERMAL AND DENSITY CONSIDERATIONS

The Fort Worth and Midland adiabatic diagrams for 1800 CST (fig. 4) il­lustrate the thermal and moisture characteristics of the air masses well east and west of the dry line near the time of Flight 37. The Fort Worth sounding is typical for northward-moving mad time tropical air in spring and summer. Note the very high moisture content at'lower levels, capped by a temperature inversion near 800 :mb. The Midland sounding shows a nearly dry-adiabatic lapse rate to 700 mb. with potential temperature of 3100 K. It is seen that below 800 mb., the air was dry and warm to the west and moist and cooler to the east, while little difference was observed at lower pressures.

400r---~-----r~--'-----r----.-----r----' 400r---~-----r----,-----r-~-,~---r----.

500 500

600

iii 600

li1 ~ ~

III III

~700 ~700 fI) fI) III a::

fI) U) III a:: IL IL

·800 800

900

Figure 4. -Rawinson<;le ascents at Fort Worth (solid lines) and Midland ( d.ashed lines) for t800 CST. May' 26 .• 1959. Right hand curve in each case is temperature and le.ft hand curve is dew point.

900

Figure 5.- Virtual temperature versus pre.ssure fo~ Fort Worth (ACF) and Midland (MAF) for 1800 CST. May 26, 1959. ' .

In .figure 5, Virtual temperature is plotted against pressure for Fort Worth and Midland. This shows that even with moisture considered, the air at Fort Worth below 800 mb. has lower virtual temperatu're and hence greater den.,. sity than the air over Midland. To examine the density variations in the im­mediate vicini tyof the dry line, data were selected from traverses on either side of the sharpest moisture gradient in figure 3, the points involved being

6

MIXING RATIO gm/kg.

DRY MOIST SIDE SIDE 3.3 6.9 4.1 1.7

5.9

4.3 4.5 4.1

9.8

9.5 10.2 10.3

lI: I- I-<I <I c c a 0 CD CD :E ~

/, /1

-: I ,I I

~ 1 '. I

.~T*I 1

-700 -..Q E -

&&J 0:: ::>

-800 (I) (I)

&&J 0:: Q..

-900

~T

Figure 6.~' Temperature change (6T) and virtual temperature change (6T*) observed on six traverses through dry line (Flight 37.). The mean temperature varfation (moist to dry side) was + O. 7 0 C. while· the mean · variation of virtual temperature was only· -0. 1°C.

N. MI.--------------

312 0

311 .5-----311

309.2

308.8 309.1

.310°

_----3120

315.2

313

312

310.5

306.4

305.8

304.1

304

----- I 'I 1 10 ':------'--___ ----I'-____ -'--__ --'-'-....J1 -..,: -..;. ~ _

Midland 5 West 0 East 5 10 Fort Worth 18C N. M i. from Marker Line

; ~ . lac

FLIGHT #37 26 MAY 1959

MB.

600

- 700

800

- 900

~IOoo

Figure 7.'" Cross-section showing distribution of potential temperature across dry line (Flight 37). Solid portions of. isentrop'es were analyzed . from aircraft data and dashed extensions weredei'ived from Fort Worth, and Midland' soundIngs (shown on ei.ther s'ideof cross section for .comparison).

7

1/2 to 2 mi. apart. Figure 6 shows that despite the appreciable temperature difference in this small distance, there was a much smaller difference 1n V1r­tual temperature across the sharpest portion of the dry line.

Figure 7 shows the potential isentropes analyzed from aircraft data, which may be compared with the potential temperatures tabumted in the figure for Fort Worth and Midland (2 - 3 hr. after flight time). It is seen that virtually all, the lower-level tem~erature ,contrast between these stations, 270 mi. apart, was concentrated in a region about 5 mi. across. Comparison of figures 3 and 7 reveals a close resemblance between the patterns of poten­tial temperature and mixing ratio.

6. OTHER EXAMPLES

In two other cases, "stair-step" traverses of dry lines were carried out. Analyses of the mixing-ratio patte'rns are shown in figures 9 and 11. The cor­responding surf·ace charts (figs. 8 and 10) show that there were large over­all moisture contrasts between the two air masses involved. Again (figs. 9 and 11) the horizontal gradients of moisture were very large, when considered in comparison with the synoptic scale, although they were very much weaker than in the example di~cussed above.

• GI:o.:---- ------, ,r 1010 1010 10~4

781,.......... I 1012

~54 \J I 1012 \ TAD ~1 . ~fK DOC / iNU

---~~~~~+,~ L-7--/~ ,J IDHT· GAG

LVS e -7- ~ ~ /31. / I' OKC

ABO 16 ..t.- POINT 0lle5 ~/I/// ",' TRAV Rsts #'

1012 1008 3 l . ~~f ~r:} 31A LBB ~~v

$~7Y~ ~ ~, -t/jt-i:y ~

1010 '</. 1012 /

101Cl'~ 1014 1016

1011

( , {

1500C 28' MAY 1959

Figure 8.- Sea-level chart for 1500 CST, May 28,1959 riear time of Flight 39', showing location of dry line (dash~dotted ~ine) and pla'ce where trav­verses were·made.

7001-----

~ ::::I en

'" '., 0:800

900~----~--~---r--------r-~--__ -r-

5 N. Mi. West 0 N. Mi. East 5 10

FLIGHT # 3\ 1659-1730 CST 28 MAY 1959

FigUre 9.- Cross section showing p'ath of, plane and observed moisture distribut,ion (mixing ratio) (or Flight 39.

In these cases, as in the other case, there was evidently 1i ttle density contrast: ac'i'oss the dry line (over the distance covered by the aircraft trav­erses). Particularly in Flight 39 (fig. 9), the zone of sharp moisture gra­dient again had a very steep slope.

8

GUl-----'OO6 TOP' 8

~ ~ r '(004 \--57 ,004 1OO~2 ~. L t:!J GCK

\ ~ ~:~J :]1 --~I- I;;;'-;-\- OAG--t -

LVS ~---. ---;?l :lp TCC 27 AMA b OKC ~ 91 190 PITOF80

~ 1 ~ .,. T~A ERSESl~ .3 • CDS I I iF!

D=WKlO2 i ~B i L I BGS

. /MAF ;~ -- - __ 195 3-5; SJT

~ 1004 " \ 1004 .

) roQS

1500C 23 MAY 1960

'008

Figure 10.- Sea-level chart for 1500 CST, May 23, 1960, near tim~ of Flight 56. showing location of dry lin~. and place wh~re trav­~rs~s ' w~r~ made.

7.

Marker Lin.

650r-----------~------------------_.

700r----

4>

5800r----

'" '" ., ... Q..

9001----

6 .0 950~--,_------_;--------_r--------~

5 N.Mi. We.! 0 N.Mi. Eas! 5 10

FLIGHT # 56 1616-1654 CST 23 MAY 1960

Figure 11.- Cross s~ction showing path of plane and observ~d moistur~ distribution (mixing ratio) for Flight 56.

CONCLUSIONS

The limited number of cases' for which detailed f light data are avai'lable allows only quite tentative conclusions; however, some bases for hypothesiz·­ing seem to have been established. The pattern which emerges is that of a sharp boundary between the nor~hward-streaming Gulf air and the warm, dry air of semidesert origin. The close juxtapositibn of these contrasting air masses is 'apparently maintained kinematically (note that the low-level wind field is frontogenetic).The sha.rpness df the dry-line zone indicates that this fron­togenesis is sufficient to offset the effects of lateral mixing. Sincethe general configuration seems to involve rather delicately balanced static and kinematic factors, this configuration may undergo considerable fluctuations, especially of a diurnal nature. The near absence of' any slope in the dry­line surface in figure 3 is compatible with the apparent lack of any density difference, but this is nqt necessarily representative of all times of the d~. '

It is interesting to consider, for instance, the ~ffect of noctuinal ra­diation along a line separat.lng very dry from very moist air. We can visual­ize ' comparatively large nocturnal cooling west 6f the dry line, due to the .absence of a re-radiating moisture blanket aloft. This would result directly in a tenqency to decrease the temperature contrast across the dry line.

In addition, the cold air would tend to drain eastward, undercutting the moist air and perhaps leading to some diffusivity of the contrasting dry and moist air during early morning hours. Increased strength of surface winds due to daytime heating would presumably contribute to kinematic frontogene­sis by afternootl . . Al though detailed observations are not' available, synoptic

/

9

analyses suggest that the dry line is more intense in the afternoon than in early morning.

The effects of diurnal variations, particularly in the vertical flux of momentum, offer an area of interest for further investigation. Study of such variations is likely to be important in determining the role of the dry line in generating convective activity, . since diurnal variations of convergence and of frontal slope, in addition to variations in air-mass structure, are factors in the formation of convection.

Since squall lines frequently appear to form at or near dry lines, ad­ditional investigations will be carried out as a primary objective of NSSP. These will in future involve wind measurements as well as measurement of the other meteorological quantities, so that a study of kinematic properties can be made.

10

REFERENCES

1. Beebe, R. G., "An Instability Line Development as Observed by the Tornado Research Airplane," Journal of Meteorology, vol. 15, No.3, June 1958, pp~ 278-282.

2. Fawbush, E. J . , R. C. Miller, and L. G. Starrett, "An Empirical Method of Forecasting Tornado Development," Bulletin of the American Meteorolog­ical Society, vol. 32, No.1, Jan. 1951, pp. 1-9.

3. Fujita, T., "Structure and Movement of a Dry Front," Bulletin of the American Meteorological Society, vol. 39, No. 11, Nov. 1958, pp. 574-582.

4. Lee, J. T., · and C. L. David, "The Tornado Research Airplane, 1958-1959," Bulletin of the American Meteorological Society, vol. 42, No.4, Apr. 1961, pp. 231-238.

5. Sanders, R. A., "An Approach to Quanti tative Precipi tation in the Central United States Basin by Use of the Theta Prime Chart," Unpublished Man­Uscript, U. S. Weather Bureau, Kansas City, Mo., 1957.

. USCOMM-WB.DC