OCTOBER 1982 R.A. MADDOX, D. M. RODGERS AND K. W. HOWARD 1501ANNUAL SUMMARIESMesoscale Convective Complexes Over the United States During 1981 Annual Summary R. A.' MADDOX, D. M. RODGERS AND K. W. HOWARDNOAA, Environmental Research Laboratories, Office of Weather Research and Modification, Boulder, CO 80303(Manuscript received 8 June 1982)ABSTRACT- Satellite images are used to document the life cycles of Mesoscale Convective Complexes (MCCs) whichoccurred over the United States during the warm season months of 1981. These systems were found toexhibit characteristics similar to aspects of MCCs discussed recently in the literature; however, the behaviorof several of the convective systems poses questions that can only be answered through detailed studies. Thesystems did produce a variety of significant weather events ranging from severe thunderstorms to locallyheavy rains and flooding. Information is also provided for a number of other significant mesoscale convectivesystems that, although they did not meet the stringent MCC definition criteria, caught the investigators'attention. This documentation should provide a useful starting point for scientists who might wish to pursuestudies of mesoscale convective weather systems.1. Introduction Recent studies indicate that mesoscale convectiveweather systems, particularly those of meso-a horizontal dimensions on the order of I-2 x 103 km, arefrequently responsible for widespread, significantweather events such as heavy rain and fiording. Suchevents are in addition to the severe thunderstormsthat frequently occur early in the life of these Mesoscale Convective Complexes (MCCs; see Maddox,1980a). The definition used to identify MCCs is reproduced in Table 1. This definition is based uponcharacteristics observable in satellite images becauseof the wide range of atmospheric scales that may bemonitored from satellites. It is most important to realize that this definition was designed so that verylarge and long-lived convective mesosytems could' beeasily identified and then studied utilizing synopticupper-air and surface observations. The definitionwas not designed to uniquely describe or denote allexamples of this particular type of convective system;indeed, many "MCC type" convective complexes donot meet these stringent size and duration requirements. Nevertheless, Fritsch et al. (1981) showed thatMCCs produce widespread regions of measurablerainfall and that they likely account for a significantportion of growing-season rainfall over much of theUnited States corn and wheat belts. Maddox (1981)found that, in addition to these widespread beneficialrains, many MCCs produced locally intense rainfallsand flash flooding (see also Bosart and Sanders, 1981).Maddox (1980b) noted that MCCs appeared tostrongly affect upper-tropospheric wind fields andFritsch and Maddox (1981) showed that operationalnumerical models do not forecast this type of important mesoscale/large-scale interaction. The frequent occurrence of MCC systems over thecentral United States, where there are relatively densemeteorological observation nets, provides a uniqueopportunity for studying these systems, their internalstructures and circulations and interactions betweenthe background environment and large regions ofconvective storms. With this promising research poTABLE 1. Mesoscale Convective Complex (MCC) definition,based upon analyses of enhanced IR satellite imagery.Physical characteristicsSize:Initiate:Duration:Maximum extent:Shape:Terminate:A: Cloud shield with IR temperature ~< -32-C must have an area >~ 100 000 km2B: Interior cold cloud region with temperature ~< -52-C must have an area >~ 50 000 km2Size definitions A and B are first satisfiedSize definitions A and B must be met for a period >~6hContiguous cold cloud shield (IR temperature ~< -32-C) reaches maximum sizeEccentricity (minor axis/major axis) >~ 0.7 at time of maximum extentSize definitions A and B no longer satisfied1502 MONTHLY WE/~THER REVIEW VOLUME ll0ab FIG. 1 a, b. Tracks of 1981 Mesoscale Convective Complexes. Dashed lines indicate regions and movements of initial thunderstorm developments. Triangles show location of the system at "Initiate," circlesat" Maximum Extent," and crosses at "Terminate" (refer to Table 1). Numbers in circles correspondto system numbers shown in Table 2.OCTOBER 1982 R.A. MADDOX, D. M. RODGERS AND K. W. HOWARD 1503bFIG. 2. Enhanced infrared satellite images for (a) 1130 GMT 13 May, (b) 1030 GMT 27 May, (c)0115 GMT 29 May, (d) 0600 GMT 24 June, (e) 0330 GMT 3 July, and (f) 1330 GMT 29 Sept.tential in mind, 30-minute interval satellite images(enhanced infrared-MB curve-sector KC30N95W)were surveyed for the entire 1981 warm season todocument United States MCCs. The results of thisextensive satellite survey are discussed in the following sections.2. 1981 MCCs The 1981 warm season (considered here as Marchthrough September) produced 23 easily documentedMCCs over the central and eastern United States. Thetracks followed by the centroids of the ~< -32-C cloud1504 MONTHLY WEATHER REVIEW VOLUME 1i0dFIG. 2. (Continued)shields of these MCCs are plotted on Figs. la, b. [Theprocedures followed in this study adhered to the definitions and methodologies described by Maddox(1980).] These documented MCCs generally affectedthe region of the United States between the RockyMountains and Appalachian Mountains, althoughseveral systems' remnant showers and cloud debriscould be followed to the eastern edge of the satellitesector over the Atlantic Ocean. Wetzel et al. (1982)also tracked a number of MCC-type systems from theOCTOBER 1982 R.A. MADDOX, D. M. RODGERS AND K. W. HOWARD 150500,572 13791 KC37N95W-2 ~ ..... ~ ~.. ;. .' ~ ~ ,~,~.,,,~ ; ,,~,~ ?;,~, .. '~ ~ , ~ ~' ,~ , dFIG. 2. (Continued)eastern slopes of the Rocky Mountains into the northAtlantic. Bosart and Sanders (1981) tracked a heavilyprecipitating, long-lived convective system for severaldays before the event culminated in the destructiveJohnstown flash flood. About half of the 1981 MCCsgrew from initial thunderstorms which developedover the eastern slopes of the Rocky Mountains, withthe genesis region for the remainder of the systemsgenerally lying to the west of the Mississippi River. The satellite images shown in Fig. 2 capture some1506 MONTHLY WEATHER REVIEWTABLE 2. 1981 Mesoscale Convective Complexes.VOLUME 110Time (GMT)/DateMax. cloud toparea x 103 km: First MaximumNo. Date storms Initiate extent Terminate ~< -32-C ~< -52-CSignificant weather1 31Ma~l Apr 2100/31 0001/1 0400/1 1130/1 412 2192 10/11Apr 2015/10 2315/10 0345/11 0531/11 159 1033 13/14 Apr 1730/13 0145/14 0445/14 1000/14 639 3734 27/28 Apr 2030/27 0245/28 0600/28 ,1000/28 282 1415 8/9May 2015/08 0115/09 0445/09 1015/09 267 1606 13May 0530/13 0700/13 1030/13 1615/13 310 1887 15/16May 2000/15 0445/16 0815/16 1315/16 276 1698 26/27May 2130/26 0515/27 1130/27 1400/27 272 1509 28/29May 0100/28 1330/28 1730/28 0145/29 318 19010 28/29May 1800/28 21'30/28 0145/29 0800/29 419 27611 5 June 0300/05 1000/05 1545/05 1715/05 280 14012 7/8 June 1930/07 2315/07 0445/08 0745/08 239 18813 10 Jun 0445/10 0615/10 1130/10 190~/10 207 16014 11Jun 0530/11 1030/11 1400/11 1630/11 328 21115 20/21Jun 1700/20 1245/21 1530/21 1900/21 245 18816 22/23 Jun 0545/22 1100/22 1545/22 0415/23 445 31817 22/23 Jun 2315/22 0445/23 0700/23 1000/23 304 21018 23/24 Jun 1230/23 0145/24 0515/24 1000/24 367 27019 2/3 Jul 1930/02 0000/03 0400/03 0630/03 338 26520 12 Jul 0200/12 0800/12 1030/12 ~530/12 236 12321 23 Jul 0600/23 0930/23 1200/23 1530/23 282 9222 4/5 Aug 1600/04 0400/05 0645/05 0945/05 309 18623 29 Sep 0600/29 1100/29 1330/29 1700/29 187 93 Mean 2240 0515 0905 1345 . 310 192Wind, hail, tornado, 2 deathsHailWind, hail, tornadoesNone reportedWind, hail, tornadoesWind, hailHail, wind, tornadoWind, hailHailHail, heavy rains, windHeavy rainTornadoes, wind, hail, heavy rain, I injuredNone reportedHeavy rain, 3 deathsHailHail, 'heavy rain, floodingNone reportedHail, heavy rain, tornadoesTornado, flash floodingHail, wind, heavy rainWind, hail, flooding, heavy rainWind, hailNone reportedof the 23 systems that occurred during 1981. TheMCC in Fig. 2a is the eastern convective region centered over Missouri; the strong jet-stream which intensified to the north and east of this MCC had significant impacts on cross-country commercial aviation flights (for details refer to Maddox and Fritsch,1982). Fig. 2c shows double MCC systems over thesouth-central United States. The eastern complex.over Arkansas is in a weakening phase, while theyounger and very large MCC over northwest Texasand western Oklahoma has just recently developed.A late season MCC which occurred at the end ofSeptember is illustrated in Fig. 2f. Details concerning the life cycles of each of the1981 MCCs are listed in Table 2. The average timethat first thunderstorms could be identified (definedas the first appearance of enhancement on the satelliteimage) was 2240 GMT, mid-afternoon local time.The eventual mesosystem reached the 105 km2 areacriterion (refer to Table 1) by 0515 GMT and grewto maximum extent by 0905 GMT (i.e., during thevery early morning). The average time of termination(see Table 1) was 1345 GMT. Thus, even though thetimes did vary widely, most of these MCCs were distinctly nocturnal precipitation systems. These averagetimes are slightly ( 1-3 hours) later than those reportedby Maddox (1981) for 1978 and 1979 MCCs; however, the physical reasons for these differences are notobvious. The ave(age 1981 system had a duration(initial thunderstorm to MCC termination) of 15hours. The shortest-lived system had a duration ofonly slightly more than nine hours as it developedexplosively but did not persist long. [It is importantto remember that the Table I definition creates asomewhat artificial (physically based only upon spacing of U.S. upber-air soundings) lower size bound tothe systems that are termed MCCs.] Only four of the23 MCCs did not result in Storm Data (NOAA, EDISpublication) reports of significant weather events. Although only 23 MCCs fully satisfied the rigidcriteria prescribed by Maddox (1980), many "MCCtype" convective complexes were noted during theimage overview. Fig. 3 presents satellite perspectivesof several such systems which did not meet eithershape or duration criteria but nevertheless appear tobe important mesoscale convective weather events.Note in Fig. 3c that the system over Missouri didmeet MCC definitions while the extensive thunderstorm complex over the Ohio Valley was too elongated-to meet the shape criterion.3. Case example--22/23 June 1981' The period from local midnight on 22 Junethrough midnight the next day produced several inOCTOBER 1982 R.A. MADDOX, D. M. RODGERS AND K. W. HOWARD 1507a bFIG. 3. Enhanced infrared satellite images for (a) 0730 GMT 25 May, (b) 0100 GMT 9 July, and 0715 GMT 10 June.tense (as viewed from the satellite perspective) convective systems that are briefly described in this section. At 0545 GMT on the 22rid, a pronounced squallline was depicted in the satellite image (Fig. 4a) extending from northwestern Ohio to southwesternMissouri. However, by 0900 GMT (Fig. 4b) the linehad weakened markedly while nocturnal convectionhad developed and intensified over western Kansas.This system grew very rapidly (see Figs. 4c, d) so thatby mid-morning a large MCC extended from centralKansas across western Missouri. During the day (seeFigs. 4e, f) this system split into two distinct convective complexes. The leading system moved rapidlyeastward and reached the Appalachians by evening(Fig. 4g) while the trailing system moved more slowlyand dissipated during the afternoon (note that thetrailing system did not satisfy the MCC definition'used in this study). By midnight another MCC haddeveloped far to the west over New Mexico. Thissystem (number 17, see Fig. lb) moved in an unusualsouthwestward direction quite contrary to its predecessors over the Plains. Important weather eventsassociated with the central United States convectivesystems are plotted on Fig. 5. Although these systemsproduced several reports of high winds and large hail,the principal "weather" was an extended swath ofmoderate to heavy rainfall. For example, the areaaround Hoisington, Kansas, received more than 185mm of rain with significant flooding reported. Radardepiction charts (Fig. 6) during the early morning also1508 MONTHLY WEATHER REVIEW VOLUME II0 FIG. 4. Enhanced infrared satellite images for (a) 0545 GMT, (b) 0900 GMT, (c) 1130 GMT, (d) 1445GMT, (e) 1715 GMT, (f) 2015 GMT, all on 22 June, and (g) 0115 GMT and (h) 0630 GMT, both on.23 June,indicate the widespread na.ture of the systems andespecially the embedded intense elements (note thevery high echo tops over Kansas). This particular episode poses many intriguing questions which can onlybe answered through more extensive research. Forexample, why did the system over Kansas split intoOCTOBER 1982 R.A. MADDOX, D. M. RODGERS AND K. W. HOWARD 150900012 130!8100471 13841 KC37N95W-2 d FIo. 4. (Continued)two distinct complexes; why did the eastern system circular complexes; why did the MCC over Newthen move faster than the western complex; why did Mexico move in such an unusual southwesterly dithe apparent mode of convection change from that rection? Studies of this entire period are underwayof a squall line on the night of the 22nd t~ several in an attempt to answer some of these questions.1510 MONTHLY WEATHER REVIEW VOLUME 110FIG. 4. (Continued)4. Other significant convective episodes During the perusal of 1981 warm season satelliteimages, a number of convective episodes caught theinvestigators' attention as being particularly significant, at least from the satellite perspective. Table 3presents a listing of these events; brief descriptions oftheir apparent structures; the dates and times of theirOCTOBER 1982 R.A. MADDOX, D. M. RODGERS AND K. W. HOWARD 151113391 KC37N95W-2 t ~ ~ ~ .~ ~ , ~.~m~,~~~12E-4MB 00102 12781 KC30t-195bJ-2 ( ~ ) ) m . . - . ~; ~q~,.hFIG. 4. (Continued)occurrence; and the parts of the country affected. Thistable was constructed to provide a general startingpoint for scientists who might be interested in studying 1981 convective mesosystems. Indeed, the largenumber of events listed in Table 3 indicates that theimpact (on precipitation, severe weather, etc.) of con1512 MONTHLY WEATHER REVIEW VOLUME II0 , '152 '~-i 187~_~!laflin -~--~.o,s~ngTon 'B~arton I ~ *~ County[25~ ~ ~Gr~en~'--~ ~~ ~WindHailTornadoesFlooding m .,.J[' ..... I \ ~ I I ~ ~ I ~ / \ I ~ I ~ ~ t / ~ k ~ ~1 .x[. x-x I ' i \FIG. 5. Severe storms and precipitation for 24rhour period ending at 12 GMT on 23 June. Precipitation is in mm with first contourdrawn for measurable rainfall (i.e., ~0.3 mm). Inset shows some of the very heavy rain amounts reported in central Kansas.vective mesosystems is considerably greater thanwould be implied merely by counting MCC events.5. Summary Satellite imagery was used to document the occurrence of 23 Mesoscale Convective Complexes during the warm season of 1981. These systems displayedcharacteristics during their life cycles which were ba- sically similar to those already documented in theliterature for other MCC weather systems. The caseexample of 22/23 June illustrates some of'the typesof MCC behavior which require further and very detailed study to explain. The MCC tracks and tabulardocumentation, along with brief descriptions of other1981 convective systems, are presented as a satellitebased overview of' significant thunderstorm systemswhich occurred during the year. Much work is ob NE ..... ~ TRW m~ 8 ~ ~ NEa 1035z June 22, 1981 \ x NE NA \ ...... "1, GRI TRW~ III __,'.. t I ::;.; ....... ;;::!;~:.::: :??' ::;.'::::::::::--i~i---::.~=*' ? ":::/..-' [ TRW ---- -'~"~ 18.9' 20b 1435Z June 22 FIG. 6. National Weather Service radar summary charts for (a) 1035 GMT and (b) 1435 GMT on 22 June. Echo topheights are in km; cell movement given at end of arrows in m s-t; system movement given by arrows with barbs (full barb= 5 ms-').OCTOBER 1982 R.A. MADDOX, D. M. RODGERS AND K. W. HOWARD 1513 TABLE 3. 1981 non-MCC significant convective events. Begin EndDate (GMT/Date) (GMT/Date) Character of event Areas affected08 April 0330 1500 Small convective Complex South Dakota into Iowa12 0001 2330 Convective complex Illinois into Indiana12-13 2200/12 0530/13 Small convective complex Oklahoma13 0630 1200 Small convective complex Illinois14 0045 0930 Convective line Texas to Lake Michigan19-20 1830/19 1100/20 Convective complex Kansas into Missouri21-22 1100/21 0900/22 Convective complex Texas to Oklahoma23 0000 2200 Small convective complex Texas25-26 0700/25 0600/26 Convective complex Texas01 May 0000 0600 Convective complex Texas02 0000 2300/03 Convective complex Texas03 0200 0800 Convective line South Dakota into 'Minnesota03-04 2130/03 0530/04 Convective line Kansas-Nebraska-Minnesota09-10 ' 1500/09 2000/10 Convective line Texas13-14 1830/13 1430/14 Convective line Texas-Oklahoma-Missouri17-18 2000/17 2330/18 Convective complex Texas-Oklahoma-Kansas-Missouri18-19 1000/18 2330/19 Convective line Texas-Arkansas-Louisiana23-24 2045/23 1300/24 Large convective line Texas-Oklahoma-Kansas-Missouri-Iowa 01-02 June 1830/01 0730/02 Convective line Kansas-Oklahoma 10-11 0045/10 0400/11 Large convective line Illinois-Indiana-Ohio 11-12 2100/11 0600/12 Large convective line Texas Panhandle-Colorado-Oklahoma 13-14 0900/13 0245/14 Convective complex Indiana-Ohio 14-15 0415/14 0145/15 Convective complex Coast of Virginia 14 2300/13 1445/14 Small convective complex South Dakota-Minnesota 16 0030/16 2330/16 Large convective complex Texas 28-29 2030/28 1200/29 Convective line Colorado-Nebraska-Minnesota 29-30 1800/29 1630/30 Convective complex New Mexico-Colorado-Nebraska-Wisconsin04-05 July 2315/04 0830/03 ' Large convective complex Texas09-10 1800/09 1630/10 Convective line Gulf Coast11 0500/11 1730/11 Large convective complex Minnesota-Wisconsin13 unknown 1400/t 3 Large convective complex Central United States14-15 2230/14 1130/15 Convective complex Central United States16 0000/16 1700/16 Convective complex Coast of Carolinas and Georgia17-18 1930/17 1230/18 Convective line Texas-Oklahoma-Kansas23 0200 1830 Large convective complex South Dakota-Nebraska25-26 2000/25. 0230/26 Convective complex Southeast United States26 0130/26 1430/26 Convective line Texas2:7 0000/27 1000/27 Convective line Nebraska-Mexico29 0000/29 0500/29 Large convective complex Mid-Atlantic States 04 Aug. 0530/04 1730/04 Convective complex Northern Mexico04-05 2000/04 2200/05 Convective complex South Dakota-Nebraska06-07 2100/06 0630/07 Convective complex New Mexico09-10 2200/09 0830/10 Large convective complex Northwest Utah-Arizona11 0100/11 0730/11 Convective complex Nevada14-15 2000/14 1600/15 Convective line Central United States15-16 2100/15 1400/16 Convective line Oklahoma-Kansas-Missouri-Arkansas16 0300/16 1030/16 Convective complex Oklahoma16-17 2030/16 0200/17 Convective line Texas29 0030/29 2000/29 Large convective complex Texas30-31 1600/30 0130/31 Convective complex Illinois 01 Sept. 2200/31 0930/01 Convective line Kansas-Missouri01 1300/01 2300/01 Convective line Texas07-08 2200/07 0300/08 Convective complex Texas23 0415/23 1000/23 Convective complex South Dakota26 0045/23 1500/23 Convective complex South Dakota-North Dakota26-27 2200/26 0700/27 Convective line Illinois30 0445/30 0930/30 Convective line Great Lakes area 17-18 Oct. 2015/17 0630/18 Convective line Texas-Tennessee1514MON'~HLY' WEATHER REVIEWVOLUME 110viously required before we fully understand the physical intricacies of convective mesosystems. This overview and subsequent annual summaries is presentedto provide a starting point for those interested in pursuing such studies. Acknowledgments. The authors thank Linda Atrteridge for her careful preparation of the manuscript.The cooperation of Brian Heckman and the staff ofthe NOAA/NESS Kansas City Field Service Stationmade the survey of satellite imagery possible and thispaper could not have been completed without theirkind assistance. Dr. Dave Olson of NOAA/NWS/NMC provided the authors with copies of the HeavyPrecipitation Branch detailed rainfall analyses whichwere used in the construction of Fig. 5. REFERENCESBosan, L. F., and F. Sanders, .1981: The Johnstown flood of July 1977: A long-lived convective storm. J. Atmos. Sci., 38, 1616 1642.Fritsch, J. M., and R. A. Maddox, 1981: Convectively driven me soscale weather systems aloft. Part I: Observations. J. Appl. Meteor., 20, 9-19. , -- and A. G. Barnston, 1981: The character of mesoscale convective complex precipitation and its contribution to warm season rainfall in the U.S. Preprints, Fourth Conf. on Hydro meteorology, Reno, Amer. Meteor. Soc., 94-99.Maddox, R. A., 1980a: Mesoseale convective complexes. Bull. Amer. Meteor. Soc., 61, 1374-1387.--, 1980b: An objective technique for separating macroscale and mesoscale features in meteorological data. Mort. Wea. Rev., 108, 1108-1121.--, 1981: The structure and life cycle of midlatitude mesoseale convective complexes. Atmos. Sci. Pap. No. 336, Dept. At mos. Sci., Colorado State University, Fort Collins, 311 pp.--, and J. M. Fritsch, 1982: Mesoscale convective weather sys tems and aviation operations. AIAA-82-0015, American In stitute of Aeronautics and Astronautics, New York, NY, 8 pp.Wetzei, P. J., W. R. Cotton and R. L. McAnelly, 1982: The dy namic structure of the mesoscale convective complex--some case studies. Preprints, 12th Conf. on Severe Local Storms, San Antonio, Amer. Meteor. Sot., 265-268.