W$Vfim"!:o:;.. J",£“:.]"‘V..‘Q -_', ..o'~o.t|o. ...... .u-. ‘w. v' ' ..‘ .-. . . . . I ' I. . - - . f , ‘ I. ‘ '3 ‘ O :; i.’o'.’*?9i”>"' ’ GENETIC CLIMATOLOGY OF THE GREAT LAKES REGION Thesis far the Degree of M. A. MICHIGAN STATE UNIVERSW Larry :Edw'in .Hodgins' 19:60 ' '13 ail-1:1? LIBRARY Michigan State University GENETIC CLDiATulUCE OF THE; GE‘LimT LARES ELEJGIUN By Lima Eng-IN HO‘UGINS AN ABSTRACT Submitted to the College of Science and Arts Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER UF ARTS Department of Geography JWWF— Approved 00:00.00002500 Larry Edwin Hodgins ABeThACT The anomolous nature of the climate of the Great Lakes Region due to the presence of the lakes themselves and to the convergence of cyclonic tracks in that part of North America is generally recognized. Yet, in Spite of the great importance of the region, its peculiar climate has received surprisingly little attention on a comprehensive and regional scale. Furthermore, relevant studies, whether statistical or genetic in approach, have tended to explain that climate only in general and qualitative terms. The need, therefore, has arisen for a genetic climatology of the region which will draw the require quanti- tative relationships between the dynamics of the atmospheric circulation and actual regional element occurrences. Various methods of analyzing regional atmospheric behavior are examined to determine the type of classification to which elements may best be related. Of particular note are “weather type" classifications based on either air masses or recurrent synoptic patterns. A.modifi- cation of the latter method is selected over the air mass approach be- cause it is more easily adapted to accommodate frontal precipitation and more easily applied to simultaneous regional analysis in a region of frequent air mass conflict. Three criteria form the bases of differentiation of the major weather types of the Great Lakes Region: (1) surface synOptic circu- lation pattern, that is, whether they are represented by cyclonic or anticyclonic systems; (2) approximate direction of origin of these Larry Edwin Hodgins systems; and (3) local trajectory of the systems relative to the Great Lakes. As such, the various weather types are recognized large- ly by their surface reflections which, for the present, are most easily and directly associated with the elements. They are, none the less, based on upper—level flow patterns. Because the types are represented by synoptic systems following definite tracks which are related to the upper-level westerlies and the general circulation they are truly dy- namic and can be referred to as "dynamic—synOptic system weather types". In the Great Lakes Region, eleven dynamic-synoptic systems are sufficiently distinctive for isolation as weather types. For each type, day frequencies and normal element characteristics are established. Element characteristics associated with each type are found to be dis— inct and remarkably constant. On the other hand, type frequency variation from season to season is quite marked. Type frequencies and element characteristics, therefore, are clearly reflected in, and in fact explain, not only the day to day character of the climate, but also the distributional differences of average temperature and total precipitation from season to season and year to year. Each of the four mid-season months is analyzed with resnect to: (l) frequencies of dynamic-synoptic system weather types, (2) effect of the individual types on daily element values; (3) inter— pretation of the distribution of "normal" temperature and precipitation in terms of type frequencies; (A) deviations from "normal" as explained by type frequency variation. January is found to be characterized by a high frequency of the two coldest anticyclonic types and by a diversity of the cyclonic types which bring warmer temperatures to the Larry Edwin Hodgins southeast and fairly reliable and evenly distributed :recipitation to the region as a whole. rModification of temperatures by the lakes is at a maximum in this season, particularly during the coldest type, and precipitation is significantly heavier to the lee of the lakes with several of the types, both cyclonic and anticyclonic. in April, two cyclonic types bringing high warm sector temperatures to the southeast and heavy but variable precipitation to the entire region reach a maximum frequency. Anticyclonic types appear in greater variety than in winter and are less severe. In July the coldest anticyclones are at a minimum. Milder anticyclonic types, however, are at a maximum and along with a fair number of cyclonic types with warm sectors reach— ing quite far north maintain high average temperatures with low lat- itudinal gradient. Precipitation is primarily associated with two of the cyclonic types and one anticyclonic type; in all cases it is vari- able in area of concentration. ectober is largely dominated by one dry, mild anticyclonic type. The remainder of the month is divided among a wide variety of types with their associated variety of element characteristics. By showing these relationships between tn (1 re frequencies and element occurrences, dynamic-synoptic system weather types give the required quantitative understanding of the climate of the Great Lakes Region, and at the same time provide the necessary missing link for a new, fully integrated framework for climatic synthesis based on the genetic character of the climate. GENETIC CLIMATOLOGX OF THE GREAT LAKES REGION LARRY EDWIN HDDGINS A.THESIS Submitted to the College of Science and Arts Michigan State university of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Geography 1960 6, M 4 67¢ f/¢/& / Climate has traditionally been defined as "average weather“ through the preoccupation of Classical Climatology with mean values of the main weather elements. This definition has long lost scientific acceptance but there must be, none the less, a close relationship be- tween "climate" and daily weather as we experience it. Early climatologists sought to describe climates merely by analyzing norms and means. There was little quantitative attempt to discover co—variances of the elements through synoptic processes, nor even to explore the variations of single elements to greater detail than monthly averages and extremes. In contrast to the impression given by the traditional definition, the close relation between climate and weather was often almost completely lost. Their entire quanti- tative approach was static and in the abstract. In most cases there was no empirical equivalent, in either the arithmetic or "normal" sense, to the averages used to describe climate for variation itself is more frequently the norm. To circumvent these criticisms and to generally make their systems more understandable, writers as a rule attempted to add qualitative explanations and descriptions to their Statistical material. This method was practical considering the data available, but far from satisfactory. Furthermore, it virtually neces- sitated that climatic classifications use other phenomena, such as vegetation, to decide arbitrary, quantitative boundaries rather than be based on the character of climate itself. - ii - The immediate needs arising from the early studies were ob- viously twofold: (l) quantitative explanations and descriptions, empirically verified, of the dynamics of atmospheric circulation; (2) quantitative interpretation of the elements as directly related to actual daily local occurrence. Much of more recent work has been toward these ends. Investigations with relevance to the first include the many studies dealing with energy balances, winds, fronts and air masses; those with relevance to the second include statistical analyses of frequencies, co-variarces, variations, and deviations of the ele- ments. Long term means are still widely used for gross comparisons, but because of the local and short term variations and deviations their direct application to detailed scientific work is limited. Their widest use, therefore, is as a relatively constant datum to which de- viations can be referred. With the rapid development of exploration i to the dynamics of the atmOSphere and into the statistical frequencies and probabilities of elements, climatology has vastly improved. One important aSpect, however, remains to the present with limited attention but which must be afproached in detail before the climatological image is complete. The works have, for the most part, tended to develOp independently along one or the other of the two problems, leaving the seemingly ob- vious gap, that of the qu ntitative connection between them. Although many of the studies of atmosoheric circulation even develop classi- fication systems, detailed cescriptions of the accompanying elements are rare and then almost invariably qualitative. Conversely, studies of the elements often deal entirely with statistical frequencies or probabilities and rarely correlate quantitatively and comprehensively with air masses, fronts, or the general circulation. Needed then is, first, a classification of regional atmos- pheric behavior to which local element occurrences can be directly and comprehensively related, and at the same time, a classification founded in the broader horizon of the general world circulation; second, a detailed and modern analysis of the elements relateo to each characteristic behavior; and finally, a quantitative synthesis of element occurrences and atmospheric behaviors in the interpreta— tion of climates of the world. The approach is entirely genetic. The present study, apart from its local importance, is only a small exploration into the difficulties and possibilities of surh investi— gation, the completion of which shall eventually give the true and logical relation between climate and the weather. At this time I wish to express my thanks to those who have assisted me in the writing of this thesis. I would particularly like to thank Dr. D.H. Brunnschweiler of the Department of GeOgraohy, Michigan State University for his careful reading of the original manuscript, for his many constructive criticisms and suggestions, and eSpecially for his unrestrained encouragement, consideration, and willingness to give of his own personal time to accommodate its completion. I am also indebted to my wife, Nancy for her patience, encouragement, and occasional prodding during the preparation, and for the typing of the manuscript and final cony. I would like to thank Er. A. K. Philbrick of Michigan State University for carto- graphic advice; Dr. F.K. Hare of McGill University for article -iv- reprints, recent climatological ideas, and inspiration; Mr. M.K. Thomas and the Toronto climatological office of the Meteorological Branch, Canadian Department of Transport for bibliographic and stat- istical material; Mr. A.H. Eichmeier, Michigan State Climatologist and the East Lansing Weather Bureau for use of daily weather maps and office space; and the Toronto Meteorological Office Library for pro- longed loan of maps and statistical material. L15 T U? ILIJUS '1'E£TIU1‘; . . . . . . . . . . g . . . . o o O 0 LI: T U11? TABLPE) o o o o o o o O O 0 O o o o o o o o o o o O I:;T|PLQDLICTI’\I“: . . . . . . . . . . . . O . O . . . . Q Q o 0 0 Standard Climatic Descriptions of the Great Lakes hegion . . . . . . . . . . . . . . . . . . . . . . The Need for a Quantitative and Comprehensive Genetic ClimatolOgy of the Great Lakes Region . . . . . . . thnoos Fun ran ahAhYslp of hndluhAL ATmcoPnnnIC BnhAVIoh Ahu anhnhT uCCUmhthmb: Tnmlh aLJAhTadES Il‘u‘U Ulb‘l-E‘Ulvrmvll‘fi'32") o o o o o o o o o o o o o o o o o o 0 Mean latterns . . . . . . . . . . . . . . . . . . . . Point or Areal Frequency Distributions . . . . . . . . Recurrent “Types" . . . . . . . . . . . . . . . . . . Conjunction of Elements and Standard Time Periods, "Complex" Approaches . . . . . . . . . . Air Masses and Lynootic Fatterns, Genetic [ApproaChes o o o o o o o o o o o o o o o o o o o o SYhCFTIC PATTanh hLaTflan Ties ClnsoIFICATlon , , , , , , , THL LlhAnlC_LIhojTIC sisrsn wharnan Tlfno of The bhhaT 14M ‘uL’CYI-VN o o o o o o o o o o o o o o o o o o o o o o C‘JCloniC Tynes O O O O O O O O O C O O O O O O O O O . NitiCI’:Clonj—c nrpes O O O O O O O O 0 O O O O O 0 O O 0 Type Durations and Frequencies . . . . . . . . . . . . M‘dXLlulL \JF ‘nlmi‘tinn TYIL mm-JMT CihflmCTflfilleICb . . Boundaries and Stations Employed . . . . . . . . . . . Data Used . . . . . . . . . . . . . . . . . . . . . . Method of Analysis and Presentation . . . . . . . . . Analysis by Types: an (Northern High North) . . . . . . . . . . . . . an (Northern High bouth) . . . . . . . . . . . . . -v1- H. I ‘ AS 51 TASLs of CchTEh13-Continued aLl (Alberta Low through the Lakes) . . . . . . . . 55 gLe (Gulf Low mast) . . . . . . . . . . . . . . . . 59 le (Montana Low through the Lakes) . . . . . . . . 61 sLl (Southern Low through the Lakes) . . . . . . . . 66 n&eH (Northern and Eastern Highs) . . . . . . . . . 68 nHl (Northern High through the Lakes) . . . . . . . 73 le (Western High through the Lakes) . . . . . . . . 76 aLn (Alberta Low North) . . . . . . . . . . . . . . 79 - .f‘, t .1 ‘ tr: - ‘ " fled Khehr mglanfl (14.811) 0 o o o o o o o o o o o o o o 4 .a I: l~-"T C‘i ‘r‘ -~ (71,” ‘1 m'I -. ,.~ _..-I‘ r-1 T -- ' .‘v'V‘ y. . A {-1.77 '“H {M -v-.j'u umvum ”la-Judo bf‘ inn \J‘ALJJJ'AL ”any Halal-Mu . AL all» i ”Lining " - '. ““ " “' 7‘ 1'“ 15" 1 <.."T'”"-’.‘ Db NBA; 11‘ ;L;.A¢UIA‘JCIW ALL) garland. l... . . . . . . . . . . . . 92 January . . . . . . . . . . . . . . . . . . . . . . . 92 April . . . . . . . . . . . . . . . . . . . . . . . . l0? July . . . . . . . . . . . . . . . . . . . . . . . . . 109 October . . . . . . . . . . . . . . . . . . . . . . . 116 Summary of Seasonal Comparisons and Type Frequency . . 124 :U.J£t;LY full.) Cul\3ClJUk—‘I‘~"N o o o o o o o o o o .‘ o o o o o o 0 1,211 BIBLIVG;LIL;.‘ . o g o o o o o o o o o o o o o o o o o o o o 135 ll. 12. 13. 14. 15. 16. 17. 18. 19. V‘I" 4U. LIST of" IL],Ufi‘iTiufTIgflif} Schematic Diagrams of the Lynamic-Synoptic System Heather Types of the Great Lakes Region Day Frequencies of Great Lakes Weather 1953-57 0 o o o o o o o o o o o o 0 Heather Station Index . . . . . . . . nan (Northern High North) . . . . . . an (Northern High South) . . . . . . aLl (Alberta Low through the Lakes) . gLe (GulfLowEast) ......... le (Montana Low through the Lakes) . sLl (Southern Low through the Lakes) . ndeH (Northern and Eastern Highs) . . nHl (Northern High through the Lakes) aLn (Alberta Low North) . . . . . . . whl (Eastern High through the Lakes) . neH (I‘JeW' England Ill-.511) o o o o o o o 0 January, Normal Temperature and Precipitation April, Normal Temperature and Precipitation July, Normal Temperature and Precipitation . . October, Normal Temperature and Trecipitaticm Departures of kontnlv Average Temperatures from v.7 '7 1.; ! ormal, 1953—5 Nonthly Total Precipitation, lTSJ-L7 . . . . . - viii — 'T‘w pres : h" r. ,7 j r JD 140 91 1.1;; '1‘ cu" I'MJLi-L; Daily Temperature, January 1954 . Daily Precipitation, January 1954 Daily Temperature, April 1953 . . Daily Precipitation, April 1953 . Daily Temperature, July 1954 . . Daily Precipitation, July 1954 . Daily Temperature, October 1954 . Daily Precipitation, October 1954 Page \0 \O 103 107 111 114 118 122 .L:.‘1‘-:cL'UCTIoI‘J The anomalous nature of the climate of the Great Lakes Region '5 as a result of the influence ol the akes themselves and of the con- vergence of the cyclonic tracks of Horth America in that region is generally recognized. Yet, in snite of the great economic, political and social importance of the region, there have been surprisincly few comprehensive examinations of its peculiar climate. In fact, descrip— tions of the climate of the Great Lakes Region as a whole have rarely been attempted. At best, local and continent—wide analyses can be pieced together with these few rare studies for a general and overall understanding. 0 f the Great Lakes Region 1 Standard Climatic DesC“i;tions According to the Koepnen classification, the entire Great Lakes Region is within the belt of D climates, termed humid microthermal o by Trewartha, having January mean temneratures below 96.6 F. and July means above 500 F. Mean annual temperatures range from just over 20 F. north of Lake Superior to near 500 F. south of Lakes Michigan and Erie. Three subtypes of the humid microthermal climates annear, but the dif— ferences are mainly in degree of temperature. Illinois, Indiana and Ohio (Dfa) experience long warm summers with their warmest month above 0 o o o o I r\ o 71.6 F.; Minnesota, Luscon81n, Michigan and Southern ontario (be) have . . . 0 short summers With their warmes month below 71.6 F.; most of Northern l . o ,, 0 ~ 4—. ‘ Y“ o A {oeppen h., Die glimate oer aroe (oerlin, 1951). -1- - 2 - Ontario is subarctic (ch) with less than four months above 50° F. Annual precipitation ranges from just over 20 inches in the northwest to over 40 inches in the southeast. There is no predominant wet season. Most humid microthermal climates of the world show a summer maximum; but significant areas of the Great Lakes Region, particularly in the east, have surprisingly uniform seasonal distribution or even a slight winter maximum. Trewarthal characterizes the climate of humid microthermal regions as a whole by cold winters, durable snow cover, long frost seasons, and large annual ranges of temperatures. He emphasises that they are largely land controlled and are therefore distinctly contin- ental. Though largely true of the west, this perhaps tends to minimize, first, the great importance of moisture and heat transfer from.the Gulf to year round reliability of rainfall and to summer temperatures; and second, the more local importance of the Lakes themselves. Temperature extremes for the year, however, are accentuated by the dominance in winter of northern continental air and by the monsoonal tendencies of maritime tropical air in summer. Winter is dominated by the non- pcriodic changes of cyclones and anticyclones; diurnal changes are sub— ordinate. Monthly averages are of limited value for the description of winter temperatures because of their great variation from day to day. Particularly noteworthy are strong importations of arctic air known as "cold waves" during which temperatures drOp rapidly in many cases to below 00 F. In summer diurnal changes become relatively more signifi- cant, but air mass control is still important. Of note is the summer lVernor C. Finch and Glenn T. Trewartha, Elements of Geography (New York: McGraw—Hill Book Co., 1949), pp. 185-204. _ 3 - counterpart of the cold wave, that is, the heat wave in which steaoy transportation of trOpical air northward keeps maximum temperatures above 90° F. for some time. The general humid microthermal summer maxi- mum of precipitation, Trewartha attributes to l) the greater reservoir of moisture in the warmer air, 2) the greater prevalence in winter of anticyclonic circulation particularly in subarctic regions, 3) summer convection, A) the tendency to strong inflow of moist maritime tronical air in summer and to the outflow of continental polar air in winter. The increase of winter precipitation in eastern North America, he at- tributes to frontal activity and the absence of a barrier to maritime tropical air. Spring and autumn are characterized by a struggle between winter and summer controls. Mild days are followed by frosts. Spring is famous for its fickleness. Autumn brings some of the nicest days, those known as Indian Summer, with clear skies, warm mid-days and crisp nights associated with anticyclonic circulation; but it also brings, with cyclonic circulation, some of the rawest and gloomiest days. Strahler} in strong contrast, uses a more genetic classification to describe the climate of north central and northeastern United States and southeastern Canada as "humid mid-latitude continental in the battleground of polar and tropical air masses". The bases of the ana- lysis are air masses, fronts, cyclones and anticyclones. Mean element values are secondary and exemplary rather than primary and requiring explanation. The interpretation is noticeably different. The region referred to is intermediate between the source region of polar continental air masses on the north and northwest and the maritime trOpical air masses 1Arthur N. Strahler, Physical Geography (New York: John Wiley & Sons 1110., 1951), pp. 371;.770 - 4 - on the south and southeast. Maximum interaction of the air masses occurs in this region along warm.and cold fronts associated with east- ward moving cyclones. In winter, northern continental air dominates and cold prevails; in summer maritime trOpical air dominates and high temperatures prevail; seasonal contrasts, therefore, are great. Strong air mass contrasts result in much violent frontal activity, highly changeable weather, and ample precipitation throughout the year. In the west, a summer maximum of precipitation and strong continental temperature contrasts and ranges particularly reflect the predominance of trOpical air masses in summer and northern air masses in winter. In the east, maritime air masses have ready access to the region through- out the year; precipitation is more evenly distributed; and temperatures, though still noticeably continental in daily range, have smaller annual ranges. Kendrewl using a regional organization is primarily elemental in approach but stresses the significance of the almost unbroken pro- cession of cyclones and anticyclones in the extremely variable weather and temperature of North America. In referring Specifically to the Great Lakes and St. Lawrence region, he points out that this region has perhaps the most variable conditions owing to the convergence there of the most frequenct cyclone tracks of the continent. Similarly he ac— counts for the "increase in precipitation to more than 30 inches annually in the neighbourhood of the Lakes" by the abundant winter precipitation. In that season not only are the cyclones eSpecially vigorous, but also lW.G. Kendrew, Climates of the Continents (New York: Oxford Clarendon Press, 1953). - 5 - the warmth of the Lakes tends to attract the lows and thus accentuate the convergence of the tracks. Of more specific nature, numerous discussions of the importance of the lakes1 themselves to temperature, precipitation, and pressure are available. For the Great Lakes Region as a whole, analyses of the effects on temperature are the most common. According to Lautzenhiser? the air masses most modified are the very cold, continental arctic masses of winter as they pass over the warmer lake water. In contrast, c? air is little affected in summer, for it is warmed in its travel across land and temperature differences between air and water are small. Conversely, tropical maritime air from the Gulf is little changed in winter as it has already cooled in its northward travel and the lakes are relatively warm. In summer and es— pecially Spring, when the lakes are cool, theair may be many degrees warmer than the water and cooling by the water prevents heat waves from reaching the northern and eastern shores. The same cooling will often cause heavy fog and low stratus to develop if the moisture content of the Gulf air is sufficiently high. Kendrew? on the other hand, merely demonstrates the influence of the Lakes on temperature by the resulting gross pattern of mean iso- terms. It should, however, be noted that this is a mean analysis and 1The lakes warm slowly in summer and cool slowly in winter. Even in the coldest winter they do not freeze over. In summer the sur- face of Lake Superior warms only to the fifties, but Lake Erie and southern Lake Michigan reach water temperatures in the low seventies. 2R.E. Lautzenhiser, "Great Lakes weather", Heatherwise, Vol. VI, NO. 1 (FGb., 1953), pp. 3—5. 3Kendrew, p. 316. - 6 - frequently west shores are similarly modified by easterhywinds. Putnaml in a more local description of Southern tntario climate, further illustrates the effects of the lakes in the diminished difference between day'and night temperatures and the resulting longer frost free season. At Leamingten, the frost fre ee se son is 170 days but inlard at Algonquin Fark, it is less than 137 days. Perhaps the most detailed quantitative study of the effects of the lakes on temperature has been done by Leighly? The modifying in- fluences of the lakes on the annual march of temperatures are examined by a series of iSOpleth maps showing rate of temperature rise in Spring, temperature and date of maximum, rate of fall in autumn, temperature and date of minimum and finally annual range. In Spring he finds the most rapid warming in the northeast, whereas in fall the most rapid cooling is in the northwest. Leighly accounts for this difference by the change in principal source of moist maritime air, in winter the Pacific and in summer the Atlantic, which leads to a great shift of continentality. The explanation is dubious in view of the fact that the Gulf is the pri- mary source of moisture in all seasons. In a second series of maps corresponding to the first, Leighly draws hypothetical isopleths to represent the pattern supposing the ab- sence of the Lakes. Anomaly lines are then drawn by graphic sub- traction. The effects of any one lake are found to vary "directly with its area and inversely with the temperature of the air over the land lPutnam, Donald F. (ed.), Canadian Regions (New York: Thomas Y. Crowell Co., 1952), pp. 22l—25. 2Leighly, J., "Effect of the Great Lakes on the Annual l'arch of Air Temperature in the Vicinity", _ , - ,, Science, Arts and Letters, Vol. XXIII (l9hl), pp. 377- Alt. - 7 - surface about it", except for the effect of Lake Superior in summer which is greatly out of proportion. Regarding pressure, Lautzenhiser expands on the influence pre- viously suggested} In winter a low tends to form over the warm water and in summer a high develops over the cold water. Thus, the tracks of lows and highs tend to be steered to or away from the Lakes area ac- cording to the season. Highs are strengthened in summer and weakened in winter, and for lows the reverse is true. Detailed studies of the importance of the Lakes to the precipi- tation of the region as a whole are virtually nonexistent, though there are many studies of the various winter "snow belts". Lautzenhiser des- cribes as the "most spectacular effect of the Lakes" the movement of cold continental air across the lakes in late fall or early winter. The great temperature difference created by the heating, along with the humidi— fying, of the surface layer of cold air lead to turbulent convection currents and excessive snow to the lee of the lakes. He further suggests that only in winter with the appreciable "lake snows" is the surface water an important source of moisture for precipitation. Slight variations occur in the theory of develOpment of the "lake snows" but most2 agree roughly with that of Lautzenhiser. The most frequent addition to the theory is that of the orographic lift of the potentially unstable air by the highlands which are somewhat inland but experience the heaviest fall. Such is the case in the zone east of Lake Erie to the Adirondacks and in lKendrew, above, p. 5. See, for example, C.L. Mitchell, "Snow Flurries along the Eastern Shore of Lake Michigan", Monthly Weather Review, Vol. XLIX (1921), p. 502, or B.L. Wiggin, "Great Snows of the Great Lakes", Weatherwise, Vol. III, N0. 6 (1950). -8- the uplands of Southern Ontario facing Lake Huron and Georgian Bay. Remickl is unique in his analysis of frictional influences on wind velocity and direction leading to a field of convergence on the right hand portion of the lee side of Lake Erie. Fast moving air from the lake is forced us over the slower land air, is cooled adiabatically, and with s fficient Condensation produces precipitation. Much more rare than studies of the importance of the lakes themselves are studies of the effects of relief bordering the shores. Differences of relief are not great in any part of the region but several areas are of significance. The importance to precipitation of the few relative highlands has already been pointed out. Tutnam.fur+her il- lustrates their importance in the "cold loops" of the northern South— western Cntario highland which reaches in elevation of 1800 feet and of Algonquin Park which is at about 1600 feet. The Need for a Quantitative and Comprehensive Genetic Climatology of the Crest Lakes Region From the foregoing rough synthesis and discussion of various descriptions and explanations of the general climate of the Great Lakes Region, it can be seen that there have been two different and rather dis— tinct approaches, the elemental and the genetic. The first an roach, exemplified by Koeppen, Trewartha, Kendrew, and many local studies both old and new, is fundamentally an organized analysis and presentation of the elements with qualitative eXplanations. The second approach, that of Strahler or Lautzenhiser, is in essence the converse of the first. lJ.T. Remick, "The Effect of Lake Erie on the Local Distribution of Precipitation in Winter", Bulletin of the American Keteorolovical Societ , Vol. XXIII, No. l and 3 (19A2), po.l-h and 111—1?, respectively. _ 9 _ Genetical dynamics of the atmosphere one the oasis of the CCSCTthlOU and resultant elcman; characteristics are secondary. Generalizations are again presented qualitatively and cetails are given in the form of examples. The present study, on the other hand, is an attempt to draw quantitatively the genetic relationships between the dynamics of the atmosohere and actual regional element occurrences. it the same time, it is hoped thvt it helps to fill the more general need for a more com— prehensive study specifically of the climate of the Great Lakes Reeion. Various methods of analyzing regional atmospheric behavior are examined to discover the type of classificatio. to which the great variety of element occurrences of the Great Lakes Region may best be related. The suitability of the "funozmental” basis of the selected type of classi~ ficstion is discussed. The behavior of the atmosphere in the Great lakes negion is classified, and the classification a plied to the ob- servation and analysis of actual daily element occurrences. Finally, the climate of the Great Lakes Legion, its seasonal and yearly variations, and its variations from place to place, are discussed in terms of a synthesis of element occurrences and atmosnheric behavior. METHQDS FJR THE ANALYSIS UK LLEGIUNAL ATMOSPHERIC- RVnHAVIUR any 1.114de UCUULLLLJLLHLLA’: “111111;. FLT. AUVlfl 1A\T._Al m U ULUAUVHL. LuJL‘Jb The methods available for the analysis of regional atmospheric behavior and element occurrences are many and most have been eXpanded in detail elsewhere} It is useful, nevertheless, to briefly summarize the various procedures in order to show why the particular method of this study has been chosen, and to emphasize its capabilities and re- lative advantages. Three general groups may be distinguished: (1) mean patterns, (2) point or areal frequency distributions, and (3) recurrent "types". Mean fiatterns Until recently most attempts to interpret world and regional circulation and climate have been on the basis of mean pressure or pre- vailing wind patterns. The quantitative disabilities and inclination to error of such interpretations need not be reiterated. Fundamentally both patterns are of limited validity. Mean pressure patterns are generally employed to suggest or compute mean wind directions; however, even over short periods of time considerable directional difference lSee eSpecially Wesley Calef and Others, Winter Feather Tvce Frecu uencies Northern Great Plains, Teclinical Report of the Quartermaster Research and Engineering Command, United States Army, through a contract study with the University of Chicago, Regional Environments Research Branch, Natick, dass., August 1957; F.K. Hare, "bynamic and Synoptic Climatology", Annals Assoc. American Geographers Vol. XLV, No. 2 (June 1955‘, pp. 152-162; and Arnold Court, "Climatology: Complex, Dynamic, and Synoptic," . ' r F oer one 5, Vol. ELVII, No. 2 (1957), p. 125. may occur in the real wind and for relatively longer periods the pressure pattern itself is highly variable. For prevailing winds, if streamlines are drawn parallel to the most frequent direction there is usually no possible corresponding pressure distribution; if they are drawn through resultant winds they are perhaps more useful, but it is conceivable that wind in such direction could be either minor or almost nonexistent. To generalize, either mean pressure or prevailing wind patterns represent abstractions that may never exist in actual synOptic cases, and accurate relating of the patterns to element cccur— rences is virtually impossible. Frequency distributions are used primarily to reach a more direct interpretation of oaily synoptic charts over large areas, and to illustrate the reality behind the abstraction of the mean surface circulation or pressure maps. For a grid of sampling points or areas, frequencies of sign of vorticity, frontal passages, Specific air masses, cyclone and anticyclone centres, etc., are recorced and iSOpleth maps drawn. The character of the circulation over the area is thus demon- strated quantitatively and a link is made between the large scale move- ments and specific isolated patterns to which element characteristics may be related. The major weakness of the method for the present purposes arises from the fact that element characteristics may be compared with the occurrences of any one circulation parameter for only one station or areal block at a time. Correlation of elements occurring simultan- eously over large areas with the genetical phenomena as considered is impossible for the occurrences represented on the frequency maps are _ 12 _ based on point recognition and have no direct relationship to any specific areal distribution of elements as recorded on a daily weather map. Comparisons might be made between the frequency maps and mean elenent distributions, but the relations would be necessarily gross and bear he inherent fallacies and inadequacies of mean analysis. Con- ceivably, element frequency maps could be drawn which would show a strong distributional coincidence with the frequency of certain circu- lation phenomena, such as high temperature with mT air or rain wit~ frontal passages. Though extremely useful for illustratiVe purposes, the genetic relationship, that is, whether they actually occurred to- gether, is verifiable only for one station at a time. Furthermore, Specific quantitative element values cannot be logica ly assigned to the parameters independent of location because the associated values vary from place to place. Only if large areal blocks are considered one at a time is simultaneous correlation over even a limited area possible and then the study is primarily that of taxonomy and of "type" frequencies over a given region. Secondly, each of the major parameters, frontal passages, air masses, cyclones and anticyclones, is itself composed of a great variety of types, particularly if considered on a scale of refinement detailed enough for intelligible relation of the accompanying elements. Not only are there the usual taxonomic problems but also, coupled with the variety in nature of the parameters themselves, there is a considerable loss of unity and coherence in the analysis. The method, nevertheless, is an important aid to the isolation of the location of circulation phenomena occurrences, as well as to general synthesis of world climates. For regional climatology it is A primarily a tool for relating isolatec circulation patterns over small areas to the broader atmospheric circulation. Recurrent "Types" The number of theoretically possible combinations of weather elements is infinite. For many practical purposes, however, certain weather states, or at least ranges of states, are sufficiently repetitive for classification. The bases of such "weather type" classifications may vary in scale from zonal circulation, involving three spatial di- mensions and time, to instantaneous element characteristics at a single point; but each classification attempts to consider the totality of weather, rather then single elements, during a short time interval. Four broad categories of "weather types" have been recognized as follows: types based on the conjunction of elements, either genetic or non-genetic, types based on standard time periods; types based on air masses; and types based on total synoptic patterns. 0 'un 'o of T emen s * ‘ '1 - '-‘s "Pomnlex" soroaches Court, in combining these two approaches, has defined "complex" climatolOgy as follows: Each weather type is defined by the simultaneous occurrence within Specified narrow limits of each of several weather elements. In any given system of complex climatology the elements for each type are fixed, as well as the time period to which the typing applies. Different systems use different element limi s and even different elements, and may even use different periods. Element conjunction classifications focus on the percentage frequency over all observations, of the various element complexes; whereas, standard time period classifications give the frequency of 1 Court, Annals Assoc. Amer. Ceog., Vol. XLVII, No. 2, p. 127. -14- hours, days, months, or seasons with given element complexes. "Complex" types can again only be applied to single station analysis or to limited areas over which a single value of an element parameter is valid. Regional synthesis is made especially difficult by a second major drawback, that of a great multiplicity of types. Classifications with hundreds of types are not uncommon. Calef1 in his study of the Great Plains chooses the "werther day" method. *sing only temperature, humidity, wind velocity and sky cover, each arbitrarily divided into a number of ranges, and eliminating such key factors as precipitation and wind directions, the system ends up with 600 element- complex types of days. "weather Day" frequencies are then recorded for a ten year period. .Such statistical probability is perhaps useful for forecasting and for recording detailed information but is cumbersome for climatic description. Indeed, a large percentage of the Calef study is devoted to "more useful generalization" by analysis of the individual elements. Lesser difficulties lie in the handling of "duration" and in the possibility of great changes within the given time units. The great advantage of the method is its entirely statistical, emoerical and ob- jective nature. The overriding objection for the present purposes, however, is the complete absence of genetic relations to the general circulation and therefore, to regional and world climate. The interest of the method is primarily probability, not eXplanation. Conjunction of element types if based first on wind direction or curvature of the isobars are to some extent genetic but still are subject to (l) the same limitations l Calef, p. 7. - 15 _ to a small area which, in this case, must be represented by a single wind vector; (2) the multiplicity of types, and (2) only partial re- lationship to causative factors. Air Masses and Synoptic Patterns, Genetic Approaches The Great Lakes Region is characterized by strongly contrasting and rapidly alternating weather regimes. Under such demonstrative conditions there are few people unaware of certain elementary relation— ships such as the coming of cold waves from the north. Statistical analyses support such conceptions by emphasizing the much higher fre- quency of certain element complexes. One is immediately led to suSpect that there are frequently recurring genetic situations which bring with them.distinctive element complexes. Air mass and synoptic pattern weather type classifications are attempts to isolate such phenomena. Synoptic pattern weather types and air mass weather types are genetic classifications. Both are based on the hypothesis that similar genetic situations occurring at approximately the same time in the calendar year will produce essentially the same conjunction of weather elements, essentially the same duration of weather type, and approximately the same sequence of weather changes. . The conservatism and distinctiveness of the properties of types of air masses give them tangible identity. . . .Thus the climate is describable in terms of sequence and frequencies of air masses of given type each having known agd Specifiable values of the elements important in climate. If a Satisfactory set of synoptic weather types could be designed it would be a nearly ideal system. Not only would it describe the conjunction of weather elements and their duration and sequence; it would also describe this situation for large lCalef, p. 30 2 R.G. Stone, "On Some Possibilities and Limitations of Air Mass Climatology," Annals Assoc. Amer. Geog., Vol. XXV (1935), p. 56. areas, indicate simultaneous occurrences of different weather types at different places, and, provide explication of the at— mospheric processes that produce the weafiun?"types". Air mass climatology averages each element for each air mass type for every month, season or year. Normal characteristics are thus "determined for each air mass and along with the normal frequencies of each air mass they describe the climate in terms at once quantitative O . 2 1 and direCtly relatable to the weather map". Not only are the elements grouped into a limited number of frequently recurring types but also their local origins are described and may be easily related to the gen- eral circulatio on. Unfortunately, numerous problems arise in air mass analysis, (”1" particularly in conflict zones some distance from source regiors. Th Creat Lanes Legion cl earl; ill‘tstrates these difficulties. Eere, the convergence of cyclone tracks on the region means rapid and frequent alternation of air masses. Two or more masses are present over the region a high percent.age of the time and the m ses in combination are not always the same. It is, therefore, almost impossible to classify the entire region by a single air mass. The problem can be handled in either of two ways ; either the classification is expanded to include combinations of air mass es, or the air mass frequencies are calculated for individual stations and isopleths drawn. In he first case, the types thus isolated are in reality closer to synoptic patterns except for the iss ion of explicit reference to fronts and isobar curvature and, therefore, of many important relationships. In the second case the difficulties of relating elements to frequency maps have already lCalef, p. 3. 2. . ' Stone, Annals Assoc. amer. Geon., VOl. )VV, p. m] - 17 - been discussed} In addition, this latter approach has led, for example, to the description of the Great Lakes as a region characterized in winter by northern continental air masses. The Great Lakes Region, however, is certainly characterized as much, or more so, by the frequent passage of cyclones and fronts, and their accompanying element complexes. In both approaches the necessary omission of fronts and cyclones is crucial. Finally, the identification of air masses themselves and the validity of their "normal" prOperties are at best approximate and sub- jective. Considerable modification in a given air mass takes place en- route due to external forces on the mass itself as well as by mixing with air from other sources. At attempt to reduce some of these difficulties has been made by Brunnschweiler2 in his "aerosomatic" (i.e. air mass) study of the northern hemiSphere in which he endeavours to correlate element values with air masses. In so—called "Somograms" (air mass diagrams) and in tables he tries to establish that air masses behave specifically over any area at a given time. From the type station somogram of Chicago can be derived that the individual air masses bring distinct surface weather characteristics to the Great Lakes region, particularly in the winter half-year. Types and frequencies of fronts are also recorded but not related to actual weather behavior. The air mass concept is indeed valuable and numerous references are made to Specific masses in the following analysis, but air mass 1Above, p. 11. 2D.H. Brunnschweiler, "Die Luftmassen der NordhemiSphaere", (in German, with English abstract), Geographica Helvetica, Heft III (1957), pp. 164-195 . _ 13 - frequencies and average prOperty values cannot be relied upon too un- critically. SynOptic pattern weather type classifications, on the other hand, attempt to overcome the major weakness of air mass weather types by in- corporating into the classification entire synoptic patterns including isobar curvature, circulation, multiple air masses, and fronts. Again, "normal" characteristics can be determined for each type and, along with the frequencies of each type, used to describe the climate. The advantages of the synOptic pattern method have already been suggested} The method, however, is not without difficulties. Primary among these is the formulation of an objective classification of the patterns and the subsequent "typing" of individual synoptic charts. No two weather maps, and eSpecially no two sequences, are exactly the same. The problem is the usual scientific taxanomic one of classifying a con- tinuum. Every climatic classification is faced with the same difficulty; but, because of the reliability of the general world circulation? the majority of synoptic situations are sufficiently distinctive and re- petitive to be recognized as "types". With careful refinement and re- vision, and with particular reference to the controlling influences of the upper atmosyhere, it is believed a satisfactory classification can be attained. The synOptic pattern weather type method, therefore, has been selected for this study of the Great Lakes Region. For the pre- sent the method cannot be as quantitative as might be desired because 1Above, p. 15, see second quotation from Calef. 2See, for example, F.K. Bare, “The Westerlies", Geographical Review, (to be published July, 1960). -19- .. . 1 of the limited knowledge of these controlling factors; however, the re- quired details are, indeed, close at hand. Secondary difficulties arise in the type of data available and in the recording of pattern frequencies, but these will be handled in subsequent discussion. lUpper-level flow patterns, and moisture, heat, and momentum transfer. SYrofi‘TIC PaTTEEuu' hm’fiflini TYPE (EALbIFI CATIQN The potential advantages of synoptic pattern weather types have been generally recognized for some time, particularly by meteorologists interested in "analogue" forecasting. Consequently, a considerable aneunt of effort has gone into this kind of classification. Numerous schemes from various countries have been proposed, all with fundamentally the same purpose but varying somewhat in basis, scale, number of types, and presentation. It is surprising that in almost all cases the classi- fications are preoccupied with the establishing of "types", whereas the accompanying climatic elements are described only incidentally and quali- tatively. Older classifications are based entirely on surface pressure anomalies, that is, on the movement of cyclones and the position, orien- tation, and expansion of the semi-stationary polar and subtropical highs. Newer classifications bear more relationship to Upper air flow patterns and are, therefore, more fundamental and more closely associated with the broader world circulation. A.few ex'nrles will suffice. By the early thirties the importance of the upper air flow was beginning to be realized though it was still far from being understood. In 1933, Blair1 devised a system of weather types based solely on pres- sure anomalies, but he did rec0gnize that there was some correlation be— tween the relatively longer trends of weather sequence and the "general 1Thomas A. Blair, "Weather Types and Pressure Anomalies", Monthly weather Review, Vol. LXI, No. 7 (1933), pp. 196-198. _ 2o _ - 21 - circulation". In 1935 Dejordjo} a Russian, developed a classification of weather types for Central Asia. Weather types are each taken as one brief phase of the continuing general synOptic process and are characterized by a natural combination in the sequences of weather phenomena and the pre- valence of some definite kind of weather, such as dull rainy weather during cyclonic intrusion. By Werld War II, the importance of the general circulation to local climate had been fully realized. Under the impetus of military demands, analysis of the upper air flow developed rapidly and many new and greatly improved weather type classifications appeared based on the mean upper air flow patterns rather than the traditional surface features. The Germans develOped a detailed classification of "Grosswetterlagen" in northern Europe, and the Americans "extended" the classification to the Mediterranean? In both cases, the classifications were based on “zonal circulation index". Of particular interest, is a classification of weather types of North America develOped and tested during the war years by the California Institute of Technology. Elliot has described the nature of the classi- fication and its types. One characteristic of the majority of the older schemes is the use of a single typical synOptic chart to represent a given weather type. In contrast to this the guiding principle of the California 1V.A. Dejordjo, "Weather Types of Central Asia", (In Russian with English Summary), Geophysics, Vol. V, No. 2 (1935), pp. 163—200. Summary in R.G. Stone, "A_Modern Classification of Weather Types for SynOptic Purposes", Bulletin of the American Meteorological Society, Vol. XVI (1935), pp. 324—26. 2 . . . . , UniverSity of Chicago, Institute of Meteorology, A Report on SynOptic Conditions in the Mediterranean Area, (Chicago, August 1943). -22.. Institute of Technology weather types is that a typical series of daily synOptic charts represents each type. For example, one of the types in Eastern North America is re- presented on the surface synoptic charts by the progression and develon- ment of a cyclonic disturbance from just north of the Gulf, across the Great Lakes, and into northern Quebec. These successions of surface patterns are, however, only reflections of the all important upper air flow. Various arrangements of the large upper-level waves, meridional flow patte ns, and different degrees of expansion of the ring of strongest westerlies, zonal flow patterns, form the basic frame- work upon which the weather types are based. Twelve meridional flow types are differentiated by the longi- tudinal position of wave crests and troughs, four of which are peculiar to eastern North America; and four zonal flow types are distinguished by the latitudinal position of the strongest upper—level westerlies. This important and extensive work on the broader continental scale for all of North America having already been done, it is profitable in more local analyses to be able to make at least a rough correlation to the types then established. Furthermore, the significance of the upper air flow patterns, upon which the continental types are based, cannot be overemphasized. Not only do these patterns represent the fundamental genetic structure of world circulation and climate; but also, the extreme conservatism of the Upper air westerlies and their waves make them an ideal basis for recurrent weather types. The weather types of the present study, therefore, have been devised as adaptions of the broader continental types to the needs of lR.D. Elliot, "The Weather Types of North America", Weatherwise, Vol. II, (1949). 21bid. _ 23 - the more local application to the Great Lakes Region. Because the pri— mary purpose of weather types in this study is a means of relating element occurrences to atmospheric behavior rather than an aid to analogue forecasting, they are largely differentiated by surface features which, for the present, are more easily and directly associated with the elements. They are, none the less, based in the upper-level flow patt- erns and, therefore, may also be easily and directly related to the fundamental world circulation. In devising the weather types of the Great Lakes Reg ion, the surface reflections of the continental types were first examined in order to determine the typical synoptic pattern sequences which are significant to the region. It was found that several of the types could be ignored or combined with others because their major differences occur in parts of the continent other than over the Great Lakes. heather systems represented in the remaining sequences were then differentiated first on the basis of their gross circulation patterns, that is, whether they were cyclonic or anticyclonic, and secondly by the approximate directionsto their points of origin. For each system the gross circu- lation pattern is used as the major surface reflection of the upper air flow pattern and also to give a general suggestion of the nature of its total synoptic form. The approximate direction to its origin gives an indication of the initial character of the air masses involved upon entering the region, and also an idea of the trajectory of the system due to the steering effects of the upper air flow. The types thus dis- tinguished are small in number and reasonably distinct because of the conservatism of the upper air flow patterns. -24- In regional application, however, where a clear relation to actual element occurrences is sought, and particularly in a non-uniform region such as that of the Great Lakes, the exact local tracks of the systems are extremely important and must be a criterion in the final differentiation. The different tracks determine he positions of the synoptic srstcms with respect to actual ground locations and, therefore, also determine how Specific areas within the region will be affected by a given type. Fortunately, because of the local modification of the cyclone and anticyclone tracks due to the seasonal influences of the Great Lakes} the most frequent tracks either pass distinctly over the Lakes or well to the north, south, or east of the Lakes. Therefore, ‘. . . 2 . . . . the problem of subjectiVity is again limited, and only a very small number of types is added to the classification. These three criteria, origin, circulation, and local trajectory, form the basis for differentiation of the major weather types of the Great Lakes Region. Because the types are represented by synOptic systems following definite tracks which are related to the general circulation, they are truly dynamic and can be best referred to as "dynamic-synoptic system weather types". Such a classification produces a limited number of types with remarkably recurrent element-ran e complexes. The attraction of an infinite nimber of types and statistics is avoided because the element- range compleXes are natural bro pings by origin rather than by arbitrarv 1See above, p. 7, Lautzenhiser. In a more uniform area cyclone and anticyclone tracks might not be quite so distinctly divided, but then the need for local tra— jectory differentiation is not so great. -25... divisions. Variations within a given type can be readily described and explained, rather than necessitating entirely new types. Furthermore, both cyclonic frontal preci itation and non-frontal anticyclonic pre— cipitation are CJVUred logically; temperature contrasts across fronts ‘I may be demonstrated; and isotherms can be drawn that vill show not only the temperature distribution within the different air masses, but also the extent to which the air masses reach across the region. Finally, the limited number of types provides a directly usable and natural framework within which specific problems may be attacked. ""‘\ A \ l'1"‘ ‘ ‘V'T‘ T u -' v, ‘ 4“" " Bib; LTi~n'd--IC_EJl"i;-o?‘flC tl’o'fsl-i Buglaljlrflgt ii; no or -11; M... as”: '..,i‘J-_:-LL.1'»J Eleven major dynamic-synoptic system weather types of the Great Lakes Region have been differentiated on the threefold basis discussed above. Each of the criteria for a given type is represented by a letter in the type symbol as follows: (1) approximate direction of system origins - by points of the compass or first letter of a geo- graohic location; (2) cyclonic or anticyclonic circulation - by L or H reapectively; (3) relation of system tracks to the Great Lakes - by points of the compass. Cnce explained, types will be referred to in subsequent discussion by their apprOpriate three-letter symbol. Enclonic Types aLn (Alberta Low North) At the surface the well known Alberta cyclone moves from northern Alberta (a) almost directly eastward, with the centre of the Low passing across Hudson Bay or James Bay (n). The associated fronts are usually occluded as far south as the northern Great Lakes; over the southern Great Lakes they may be either occluded or separate. The warm front is often poorly developed, depending on the importation of Gulf air which is in turn primarily dependent on the preceeding type. The cold front is usually followed by an anticyclone from the west with moderate temperatures or by another cyclone. The mean upper—level flow pattern is usually characterized by l . . . . See Fig. 1. For each cyclonic type tnree succeSSive phases of the system are combined on a single map. _ 26 - - 27 - .3)". tic aegion .‘ nic—oynop . rifle S. .r' '- e ~reat +4 «u u... e 0 2n... .3 2; oi" .) 6.. H1. 8 . \J OJ ’\‘ of. . 9‘ .v ‘.:A J. . .t UJ-C I m - L '1 Fig. l.--Schf -23.. a fairly smooth west to east zonal flow concentrated somewhat.farther north than for other types. Occasionally there is a slight trough in the east, in which case the surface fronts, though not the centre of the low, may sweep quite far south. This type is most common in summer and early fall with the north— ward shift of the general circulation, but may occur in any season.. all (Alberta Low through the Lakes) At the surface this type is again represented by a cyclone from Alberta (a), but in this case, it moves southeastward acrOss the Great Lakes (1). Associated fronts occluce gradually as the system progresses eastward and may reach the Great Lakes at any one of three stages: well developed, with Open warm sector; partially occluded; or entirely occluded. The cold front is usually well deveIOped, but the warm front is occa- sionally unidentifiable. The centre of the low usually moves directly over the Lakes at a latitude approximately that of Upper Michigan. Occasionally in the occluded and partially occluded cases, it passes just slightly to the north of Lake Superior but the main frontal devel- opment is still over the Lakes Region. The mean upper—level flow pattern is characterized by a marked trough in about the samelrmgitude as that of the Great Lakes. This type is roughly the winter equivalent of the aLn type. It does occur rarely in other seasons, but it shows a marked winter concentration. le (Montana Low through the Lakes) In this type, a cyclone from Montana (m) moves directly east- _ 29 _ ward across the Great Lakes (1), and out the St. Lawrence. Two frontal forms are characteristic: either a single cold front; or a cold and a warm front, usually partially occluded, but with a broad open warm sector beyond the occlusion. The mean upper air flow is again remarkably west-east in trend across most of the continent but farther south than in the case of aT . East of the Great Lakes it swings north to a crest over the Atlantic. Over the year, le is by far the most common type. It occurs often in all seasons, but has a maximum in spring and minimum in fall. 811 (Southern low through the Lakes) A low just to the northeast of the Gulf (s), which either has reformed after passing over the American Rockies or is a new centre along the frontal zone of another low, draws in warm moist air from the Gulf, intensifies, and moves north through the Great Lakes Region (1) to northern Quebec. Fronts and warm sectors are strongly develOped and bring heavy precipitation to the entire region and high temperatures to the southeast. Generally the fronts begin to occlude over the Lakes and disappear by the time the centre of the low reaches northern Quebec. The mean upper-level flow pattern is Characterized either by a very deep trough over the western states, reaching almost to Mexico, and a slight crest over the eastern seaboard; or occasionally, by a west to east zonal flow, south of normal, over the central states. Although highly concentrated in spring, sLl is also reaponsible for the major severe thaws of winter. - 3o _ gLegfiGulf Low East) Intense cyclones originating in the Gulf region (g), move northeastward, along the Appalachians, following a track somewhat farther east than that of sLl systems, and pass to the east of the Lakes (e). Fronts and warm sectors are well develOped, but usually do not extend far enough west to reach the Great Lakes. Because the centre of the Low is usually quite far east, this type is often accompanied over the Great Lakes Region by the beginning of another type in the west. A very pronounced upper-level trough in the east reaches as far south as the Gulf and then rises sharply to a crest iust east of the continent. This is almost excusively a winter and fall type. The occas- ional fall hurricanes which reach the Lakes region are of this tyne. . . l Ant_cyclonic Types nhnngorthern High North) At the surface the centre of an anticyclonic outbreak from northwestern Canada (n) moves south and then east, passing to the north (n) of the Great Lakes. The mean upper—level flow pattern is characterized by a dis- placement of the band of strongest westerlies and their associated disturbances far to the south and out of reach of the Great Lakes. This type has a strong maximum in winter with the southward shift of the general circulation, but does occur occasionally in all seas OTIS o 1See Fig. 1. Open arrows indicate trajectories of the systems preceding the phases shown; single line arrows, those following. - 31 - an (Northern High South) This type is much the same as the an type except that the out- break (n) is much more severe, and the centre of the high passes to the west and south (8) of the Great Lakes. There is an extreme diSplacement of the band of strongest upper- level westerlies to the south. Again concentrated in winter, an is also fairly common in Spring and occurs occasionally in summer and fall. le (Western High through the Lakes) Anticyclones moving eastward from about the international border (w) generally pass directly across the Great Lakes (1). The high is us- ually so intense that its track is little affected by the lakes. Occasionally, however, the centre of a Western High passes to the south of the Great Lakes; in such cases, the type has been sub- classified as st. The mean upper-level flow pattern for this type is unusual in that the band of strongest winds is Split into two over the western part of the continent; one part forms a crest and the other a trough in about the same longitude. Cyclonic disturbances often accompany each of these bands, and pass to the north and south of the eastward moving high. / This type reaches a strong maximum develonment in fall. In other seasons it occurs only infrequently. In winter it is rare and then invariably of the st type. neH_Ljew England High) Anticyclones of the neH type are unusual in that they do not - 32 - follow either of the usual trends from west to east or from south to northeast. As other anticyclonic types, particularly le, move to the east of the Lakes they are able to draw in much warm mT air, and tempera- tures rise rapidly by ten to fifteen degrees. Over New England (e) or New Jersey the highs become semi-stationary and extend in a curved ob- long form, first toward the south, and then west. "his unusual pattern often becomes quite persistent and comes to dominate much of the eastern United States until it is destroyed or converted to a nieH by an aLn or le. It is actually the end product or remnant of other anticyclonic types; but it is greatly modified by the influx of mT air and has a re- markably distinctive pattern deveIOpment. The mean upper-level flow pattern is, likewise, quite distinctive. A crest forms in the east, and a closed anticyclonic centre appears at the upper-level over the central states, generally from.about Iowa to eastern Tennessee. Over the Plains a slight trough forms 81d allows cyclones to attack the neH from the northwest. This type occurs primarily in the fall with the highest fre- quency of le and when the land mass between the Lakes and the Atlantic is relatively cool. It occurs rarely in other seasons, but then is somewhat modified in form and position, and is not as persistent. n&eH (Northern and Eastern Highs) Also unique, n&eH is composed of two anticyClonic cells: a re- latively cool dry anticyclone centred anywhere from northern Quebec and Hudson Bay to the Prairies (n); and the large, warm and humid, Bermuda High centred off the southeast coast of the united States (e). Separ- ating the two strongly contrasting systems is a linear frontal zone _ 33 - running from southwest to the northeast roughly over the Great Lakes Region. This pattern is almost invariably introduced by the cold front of an Alberta or Montana Low, and most frequently by the aLn type. As the cold front moves eastward it is blocked in the south by the large Bermuda High and is forced to swing to its southwest to northeast align- ment. The front then generally migrates slowly toward the southeast as a quasi-stationary front. very small waves sometimes form along the front and move toward the east. The details of the pressure pattern vary, but there are always two anticyclonic centres and a tendency to lower pressure in the frontal area. The southern high is usually not as intense, but it covers a large area and is very distinct. The exact position of the front it- self, which is very important for local conditions, is also variable: when introduced by an aLn, the front generally passes completely across the Lakes but when introduced by an aLl or le it may reach its initial alignment even slightly to the south of the Lakes. The most distinctive feature of the mean upper-level flow pattern is a closed anticyclone to the southeast of the continent which may per— sist for very long periods. Over the Lakes the flow is from.west to east. Although remarkably concentrated and common in summer, nheH does occur in other seasons, eSpecially Spring. nHll (Northern High over the Lakes) In this type, a weak high from the northwest (n) moves south- 1The very approximate correlation of the weather types of the Great Lakes Region to the continental weather types of the California Institute of Technology as outlined by Elliot are as follows: aLn - B, '\ b) -‘ eastward into the Lakes area (1); there it persists and intensifies over \ the cold water of Spring and summer. occasionally, it joins along a ridge with another centre over Hudson Bay. The best nevelOpments of this type show a definite closed anticyclone over the Lakes; tut others, in spring may be represented nerely by a pronounced dip of the isob: s southward. Th: mean upper—level flow pattern is usually characterized by a fair y smooth, moderate, west to east flow just south of the Lakes, but surface lows are oeflecteo arcund the high lying over the Lakes. This tynp occurs only in spring and summer. ijpe Durations and Frenrencies The analysis of type durations and frecuencies is essential to an understanding of the relative contributions of the Various types to seasonal and yearly climate. The following observrtions are the l . o I J. D 3 o 1 ~ ‘ results of a detailed examination oi daily weather maps of hortn America for the Lonths of January, April, July and hctober of the years l}53 to 19)? inclusive. The sighificahce of these observations when Cnubined with the element characteristics of the various types to the seasonal and yearly climate of the region will be discursed in a sub— sequent section. ‘Ihe duratlcr :f'znuranie'weather't;"n; over tlu;"rea+ Takes 7“. himhly variable. especially is this true if a series of occurrences IJr‘s—a: all - Bn-h Bn—c BF° mTl — 2° ell - A 11° ”Le - Ga- nPn - 3x a J 3 9 3 ’ 3 - 3 , n ‘ " TY '7 -. -' ‘ '- 'v \ '71 T--"‘ 5n; hug _ Lb” wag: _ C; rwa,..1,3 equIvalsrt; n;€l — .’; ?1_L - ma. l v'( j ‘\ ., . A .Y ,. a , . ,. T\,.'1 - . Pr 1-_~ '.. - v, o.o., Lept. of Comm., Leathel Bureau, bail; table It 1 for 11,, on QC’) :r'r“ bub JCWFS l///*lj)/. .. 35 _ of the same type is considered as a unit. If only single occurrences are considered, the approximate durations for individual types are as follows}! sin and aLl, 1—2 days; mil, sLl, and gLe, 2-3 days; the northern Highs and le, 2-h days; n&eH and neH, both very variable, from l—5 days. The problem of quantitatively analyzing type frequencies is both more important and more difficult. Occurrence frequencies could be recorded, but because of the variability of durations they would have limited significance. Aoproximations of the comparative total times that the Great Lakes are dominated by each of the given types, however, are of great value; such approximations can be obtained by typing the dynamic-synoptic systems in the region2 for each day or shorter standard time period. This has been done for each day of the twenty months covered by this study by comparison of the 1:30 a.m. and 1:30 p.m. charts of a given day and 1:30 a.m. chart of the following day. The accompanying chart shows day frequencies in the Great Lakes Region of the various weather types (Fig. 2). A fairly accurate esti- mate of the actual number of occurrences of any one type may be obtained by dividing the day frequencies by the apprOpriate average duration. In typing both for an entire region and for a standard time period as great as a day, certain difficulties immediately arise. Obviously type durations do not coincide exactly with days. Further- more, type separation itself is difficult. However, individual types are remarkably distinctive even if their exact boundaries are not; and this flexibility of their boundaries can be utilized to obtain 1See for examples Figs. 21 to 29. 2Compare frequency map and element—complex methods where in- dividual stations must be typed; see above, pp. 11 and 1A. ~36- vmummoa .meab, , . mesh.“ .l......._...,.c .3 mowocesoopm .mflTIN .mwm 9% m>43... Izmmd >m<32<fi _ 37 - closer coincidence of type durations and days. In most cases, days can be awarded to a certain type with enough validity that the final frequency generalization is of an order of accuracy at least in line with that justified by the short five year observational period. Tran— _sition days must either be awarded to the type dominant for the longest period of the day, or be divided by half day periods between the two types. Greater accuracy can only be obtained by using shorter standard time periods, more frequent synOptic charts, and a longer total ob- servational period. As in the case of a forecaster's prediction, in the classification of patterns which fit a given type only moderately well, or in any natural classification, a certain amount of subjectivity cannot be avoided. In the accompanying frequency chart it will be noted that sub— types, some of which have already been suggested, have been included. Most of these were discovered, during frequency or element analysis of , the major types, on the basis of different element-complexes, slight shifts in oosition of the patterns, or different frontal arrangements. Finally, it should be noted again that the observations cover only a five year period. The frequencies obtained should not be taken as absolute figures of probability. However, the seasonal frequency distributions appear sufficiently repetitive that their major features can be accepted as valid as long as they are considered in relative rather than absolute terms. For individual types, the variation in their frequencies from year to year gives some indication of their reliability. Annual; or ‘w'fgaTI-{Eiii "1‘le annr CHAmU‘h-LitTICS Boundaries and Stations Employed Although numerous studies have been made of the Great Lakes Region, none have suggested significant boundaries. For detailed anap lysis of the element complexes and distributions for the various weather types, however, some workable boundary is required in order to limit the number of stations considered and to keep the discussion relevant to the Great Lakes. The most logical boundary would be the one that delimits the area whose weather elements are perceptibly influenced by the lakes themselves. In determining this boundary it is justifiable to use only work already available. A detailed study of the extent of influence of the lakes is not the present purpose, and boundaries are required merely for convenience and to concentrate interest on the core area. Leighly's article on the effects of the Great Lakes on air temperature1 forms an excellent basis for temperature limits; but unfortunately, no equivalent study of precipitation has yet been made. An investigation of the actual extent of influence of the lakes on precipitation would be enormous and very difficult in itself. The effects of the lakes cannot be distin- guished by a simple hypothetical isopleth method because of the compli- cations of bordering relief and also because of the Variable distribution of frontal precipitation independent of the Lakes. Temperature alone, 1See above, p. 6. _ 39 - therefore, has been used for the determination. Although Leighly's study is one of mean values, it is nevertheless adequate. For the pre- sent practical purpose, those areas which show an influence great enough to be represented in mean values are most likely to be the most signi— ficantly affected. A.more detailed analysis such as by weather tyoes, would tend to extend the limits rather than restrict them. Leighly's analysis of the various seasonal effects of the lakes on temperature has already been outlined in some detail. In a final map he summarizes the cumulative summer and winter effects by Combining the individual phenomena anomalies. Five arbitrarily chosen units are used to draw iSOpleths indicating the total relative modification due to the lakes in each of the two seasons. Toward the outer limits and the zero line the effects of the lakes are limited, especially in summer, to one or two of the phenomena considered. The boundary for the present study is drawn to correSpond roughly with Leighly's "1" unit line for winter but is somewhat smoothed out to include all areas immediately adjacent to the Lakes even if they are not apparently influenced to the same degree (see Fig. 3). The winter isopleth is chosen because the effects of the lakes are greatest in that season. The "1" unit line is selected so that all areas are at least influenced at some time. Beyond this boundary the lakes may still be significant in some cases, but the additional analysis for these extensions would merely add detail to the periphery and would not alter the basic conclusions for the core area with which we are primarily concerned. In addition, detailed mean monthly precipitation maps suggest not only that this x85 coasmsm a”...,u.aao.._......p, .mfi {uh ’ N MD \ 8\O\ COO/z u 0M0 . oN—M . \ z o _ .o z a X10 _ / /// 32:33 .33.: III . I E. as E. 9.. / x / >223 330 . .4. an... .o. . a a 0/ ,. 8:853: 555.5 .. .. I .II tan”: 1 De .1. 1— _ a. *0 ~60?” hmurzg III 2.4:... E». my a ,,/ . 4,. - m? allude an... )9... cm .3 cu _m 8. .. o... I / .2. 4m. 92 .2 .. , . O I . . .. . 2 _ . a om. .0... s. :1 3..” 8:. ,, , o a .4 u . ,. _I l I 018.009”... _ EOJ 0—H 3. 0% O _ l :9. ’% .V :0 b (O ”I // ._. II a h- OVOA l‘ m r as. ’11 / L KI—I Junk-C A! I. \ n . o I, a .o i , i on , w 24 3 _ \auo \ to. r / o .P,,z_mzoum.3 , a?» ,, r .w a x, u / / _ 3 >0. at. Aoz_>>o.._._on_ >wv: 228.: $32.. Emma a. mzoEEw $12“; Swim _ 41 - Fig. 3.—-Continued, Stztion Key 7' .g “:1 v1" 1,1 Ugh-Io Linian‘m-J LLH Yogi HinoiLJ.xiA Lag Saginaw AT C 12 Alb Albion N 6 Tr Troo1 kville nT 12 SC Saint Charles C 6 Alx Alexandria L 6 o Corry 6 CtJ 82 .int Johns C 6 All Allegany St.Tk.N 6 Lr Erie “TA? 12 Cd Sanduskyf ;5 Ar Arcade h 6 Ti Tittsburg 1.T?T 2 8t (tandish CL 6 Au Auburn E 12 TR Three Livers SN 5 Ba BataVia h 6 lanéfi LlUHIGAR Tr Traverse City A“ N 12 Eu Buffalo '.LAT 3 12 Ad Adrian SE 12 RT Willow Run 53 12 D Derby N 8 Alm Alma C 7 El Elmira E 12 Alp Alpena RB N 12 UgPLi LlCHlen G Gouverneur B At Atlanta N 6 Re Peechwood H 5 ii Hilton U 6 PC Tattle Creek AT So CJ Champion Van Lew Lewiston H 6 T1 Bloomingdale £1 60 Tipper Tk. W 5 Lo Lockport h 8 CSH Caro Et.Hosp. CL 5 Lb EScanaba W 12 La Nassena AT E 12 Ca Cadillac N 12 CH Grand Farais !T E 12 Tu Pulaski E 7 Ch Charlotte 8 6 Ho Houghton AP h 12 Ho Rochester NPAT N 12 D DetrOit TTAT ' 1? K Kinros A.F.P ase E 12 So Sodus W 7 SJ East Jordan N 6 KW' Nanistique Ktrwks. E 6 Sy Syracuse TTfi“ E 12 LL 39s, Lansing FT S 12 kq Narquette KT W 12 Tu Tupper L. L 8 ET11ast Tawas C30 ’ Lun Kunising E 7 Ut Utica AT E 12 L Lau Cla ire S” 6 NEH Newberry St.Hosn. E 12 Na hatertown A_? E 12 Ev Lvart C 5 On Ontanagon h 5 Erie I'Iestfie 1d 11! 8 FL. Fife Lake r: 6 SCI-I Sault Ste. Narie Fl Flint 1.T1‘ SE 12 ETA“ E 12 tHIU G1 Gladan AT C 12 St Stenhenson N 7 AC Akron Canton E 12 CA rlen Arbor N 5 s Ashtabula E 6 CH Grand Haven C 6 ILLIALA C18 Catawba Is. N 6 v3 Grand Rapids LTAT 0N1? Al Albion 6 Ch Chardon E 6 Gr Greenville C 5 FT Ft.Uayne VTAT 12 Cl Cleveland “TAT E 12 HT Hale Five Channels LT La Torte 6 Co Columbus 1mfifn“ h 12 Lam N 12 JD Ogden Dunes 7 Da Layton 1TiT J 12 HB Harbor Beach CL 7 TT lemouth Tower 6 E Elyria E 7 He Hesperia CW 6 EB South Tend VTAT l2 Fi Findlay AP W 12 HL Higgins Lake N 6 hi hinamac Ge Geneva L 6 Hi1 Hillsdale E 7 Hi Hiram E 6 Ja Jackson A? S 12 ILLIKqu No Norwalk N R? Nanistee Tower N 1? Ant Antioch 7 Pa Painesville E 12 hi hilford GK. SL 12 AC Aurora College 7 T1 lemouth H'7 , PH Nio Hydro N 12 Ch Chicago VTAT 12 Sa Sandusky 1 12 LC Nount Clements 55 12 Ha Kankakee 6 T” Tiledo Fewage W 12 Lsk Fuskecon NTAT CT 12 at Uttawa 7 To Toledo LTAT “’12 NC Lewaygo Croton C” 12 TF Tark Forest 12 Yo Youngstown ‘T1TLT 12 T' Taw Taw SW"? Ro Rockford LT 12 Tel Fellston PT N 12 Ta 1aukegan 7 1Tarticulars are given in the following orcer: map symbol, station name, area of state, final hour (all n.m.) of obserVational day. 1“ denotes Airoort; 77‘] 17, Yeather Tureen PfetiOfis. C- :1- n As C P .1 .n. 81 1.! 14. LC FL Gr CB Gu Ke | C" \.~ 034 0 (Continued) "ul t. C\ .1148 110 Looleton E 12 Aflflamrl6 Tavfield N 6 Prule 15. NE Crivitz High Falls b1 6 Eau Claire AP H Fond du Lac E 6 Grantsburg AT W 12 Green Bay MFA? E a Gurney N 6 Kenosha SE l2 12 12 \ CLLALIAN oTnTionb At FF PxJ ‘-q (4 t"‘*—‘JO CIT-1! ULL b=lL AU BI MHLYlefii Atikokan Fort Frances THUNDER BAY Armstrong AT Cameron Falls Ft. William AF Longlac ALGc-im Franz Sault White Ste. Marie River LUUBULY Biscotasing Chapleau Coniston Gore Bay AP Ruel Turbine Til-llbf’uwui‘xG Cochrane Earlton A? Kanuskasing AT Kirkland Lake NeW'Liskeard Timmins NIPlsbING Algonquin Park Bear Island Tr Ki / L D Hu La .’ (I) [0 t‘1’13 CD C) l ‘5 'N I ’ ‘ . I — ' - Dr. T._— Continued, Ftation K”? Lake Geneva 53 Radison FEAT F Nanitowac E 6 Lilwaukee STAT oshkosh E 8 Tark Falls U Plymouth E 6 Tort Hing N 6 Sturgeon Lay LE 6 Townsend EL 6 Naukesha SE 12 Nausau AT W'12 West Tend SE 7 l? ' «4 SL 12 1? Crystall Falls Fadawaska North Bay AT 131 Tina».- chTi‘uJu Brockville Killaloe A? Kingston Cttawa AT Gadfilflvbfil Deeton Durham Huntsville Eagnetawan Muskoka AP Orillia mmnSmmd Stayner Wiarton AP LAKL Uhph Brucefield Forest Lucknow S arni a. Southampton Falkerton IJgililb ISET—fiib Delhi Grimsby Grimsby Rock Chapel Leamington. VD Ya. 101 h 113 'P11 ...'.L Vi (0 :33 ’; hi *3 s :‘7‘ :t“ '1" Him '43 “3'33 t"! ‘ y—‘w NW 4 C?§36CDDH.IQP FVU‘ SAWS:- t—q "C1 11 7isconsin Dells W 1? his. Rapids L'l? 14.11.11.126‘38 Babbitt 5 Duluth RTAT 12 Grand Larais 9 International Falls iTRT l2 Leadowlands 6 Tokeqama Lam 5 Two Harbours 6 Virginia 5 T‘elee Island Dort Dover Ridgetown St. Thomas Wallaceburg Welland Vfindsor AP hLbT CLLTLAL Brantford Fergus Shand Tam Glercoe London A? Monticello Stratford Kitchener Mic; UI‘ETMLIU Agincourt Georgetown Hamilton Malton Orono Stirling R. Toronto Trenton Tweed Uxbridqe LAST CLL”HAL Ansley Fenelon Falls Gilmour Haliburton Feterborough -53.. boundary includes at least those areas most obviously influenced in all seasons; but also that a detailed map of the outer extent of in- fluence, if it could be prepared, would again merely provide for an ,extension. Within this boundary stations used in the elemental analysis of the types have been selected primarily on the basis of the final hour of their observational day. Major stations generally end their day at midnight, but it was found necessary to use the majority of stations with final hours as early as 6 p.m. to obtain a fairly tight network and a reasonably uniform distribution. The dangers of using stations with earlier final hours in determining daily regional dis- tributions is obvious. A.few major stations have been selected be- yond the boundary to the "0“ line to give a more generalized coverage of this peripheral zone. Data Used In the present study, only temperature, precipitation, and occasionally wind direction1 are considered. Other elements, however, such as cloud cover and insolation, could easily be included because the method is comprehensive in contrast to classifications which allow for only a limited number of elements, usually temperature and pre- cipitation alone. The inclusion of these other elements would not complicate the classification or require an increased number of types; it would only add detail to the analysis because the classification lNote: Wind direction is not taken as given in detail by the synoptic pressure pattern. Where required, exact directions are observational for the individual cases and not computed. - 44 - is based on genetic aspects, not on the resultant element values. Temperature and precipitation are selected here because they are the most frequently recorded and generally the most demonstrative of the elements. Wind direction is sometimes considered because of its im- portance with regard to modification by the lakes. Furthermore, the preper information for other elements is simply not available for many stations and if it were, it would be two Voluminous for hand methods. Only four months of each year, and only five years, 1953—1957, are considered in order to limit the volume of statistics and analysis. January, April, July and October are selected as representative of the four seasons. General conclusions should again be regarded in the light of the short observational period. The statistical observations themselves are based on three pri- mary sources: (1) Climatological Data} published each month for each 2 state in the United States; (2) Monthly Record, the monthly publication of element statistics for all Canadian stations; and (3) the Daily Weather Map of North America? The use of statistical data for detailed regional analysis of this kind has many limitations and dangers. Some of these arise from the great importance of local environment, but most are the result of varying observational conditions and reliability. Statistics for any lU.S., Dept. of Comm., Weather Bureau, Climatological Data (Vol. varies for different states), 1953-1957. 2Canada, Dept. of TranSport, Meteorological Branch, Monthly Record: Meteorological Observations in Canada, 1953-1956. 3U.S., Dept. of Comm., Weather Bureau, Daily Weather Man, 1953-1957. -145- one station particularly for any one type, may be quite different than for surrounding locations and in such cases certainly cannot be used for generalizations. The trends, therefore, of numerous stations, (usually 3 or more) must be examined in order to make valid conclusions. Even then the validity is questionable if a conclusion is drawn from only one occasion. However, when comwarisons are made for several oc- currences of the weather type and similar element range patterns are found to occur, generalizations may be made with a degree of accuracy roughly proportioned to the number of cases compared. Indeed, if several occurrences are considered and a similar pattern emerges, local conditions may be considered as part of the "norm" itself. Method of analysis and “rescntation For the following elemental analysis, daily statistical data and synOptic charts were used to discover the character and Spatial dis- tribution of the elements as associated with Specific cases of type oc- currences; and thus, by comparisons, to determine normals and variations for each of the eleven dynamic—synOptic system types. For each type, an initial survey of statistics and synoptic charts was first made to discover significant differences in either its synOptic patterns or its associated element occurrences; subtypes have been established where necessary. Representative days from the season of the type's peak frequency, were then selected, and various isopleth maps have been drawn for these days. Mean temperature maps demonstrate the cumulative regional dis- tribution for periods of a d'y; that is, they are examples of the general character of the type which may be directly related by frequency -45.. to monthly averages} The influence of the lakes, especially for anti- cyclonic types, is usually shown by temperature range isopleths. For cyclonic types, the primary interest lies in the warm sectors because temperatures outside the fronts are generally related to preceding and following types and, also, because the warm sectors are responsible for many above normal temperatures. Temperatures within a warm sector, contrasts across fronts, and the extent of a warm sector are shown either by syn0ptic examples or by special isotherm maps. For the latter, means are calculated from maxima and minima which may occur on different days depending on the regional position of the warm sector. The appro- priate figures, those that represent warm sector temperatures, are se- lected either on the basis of which are highest, or by correlation with the 12 hour synoptic charts. Unfortunately, the map times do not coin- cide exactly with observations; however, with subjective allowance the 1:30 a.m. chart is taken to roughly correSpond to the time of minimum temperature and the 1:30 p.m. chart to the time of maximum temperature. CorreSponding temperatures beyond the reach of the warm sector are valid regardless of when they are taken. Maps of total precipitation encompass logically all types of precipitation, frontal or non-frontal. They again give the general distribution and illustrate the values which are significant in month- ly totals. For many purposes a more detailed quantitative breakdown 1In subsequent maps and quantitative discussions "mead'refers to a figure intermediate between two extremes, usually maximum and minimum temperature for a single day; "average" refers to the result obtained by dividing a sum by the number of quantities addeo, for example, the average temperature for a given month; "normal" refers to standard values for long periods of time, for example, long term temperature averages for January. - A7 _ of origin would be useful, but this is impossible using only daily statistics. Such an analysis would require knowledge of the exact local conditions at the moment of fall for individual stations. De- tail of this order is far beyond the scope of this study; but never— theless, the dynamic-synOptic weather systems provide the preferred framework. For the present, the distribution of precipitation can be considered only as related to the movement of an entire synoptic system for a given day. For this reason, the influence of the lakes in cy— clonic types is difficult to detect. For anticyclonic types, the in- fluence of the lakes may be inferred either from the total precipitation map, or from a sequence of synoptic charts showing observed wind dir— ections and areas receiving precipitation. Distributional and quantitative normals and variations for both temperature and precipitation were then established for each type by comparison of all their occurrences over the five years in their maximum frequency month. These are presented either in terms of a comparison with the selected isOpleth maps or with long term monthly normals (temperature only}, or in terms of range values. Finally, a comparison is made of occurrences of the reSpective types in the months other than that of their peak frequency. Types are presented, primarily, in order of their major analy- sis month, and secondarily, in order of highest frequency. It should be noted that the types analyzed for any one month are not necessarily the most common types of that month. -43.. by Types F! a U) lagys nhn (Northern High North) This type has a strong maximum frequency in Jennery and it i no also the dominant type of that month (see Fir. 2). It brings the a £4 :me as the "cold anes" (’1 usual cold weather of Minter but is not as extr of the nhs type. Winds are northerly only in the initial stages and then become easterly and southerly. January 22, 195h (see Fig. 43), with winds north to northeast, represents close to the extreme of cold attained by the an type in the initial stages; January 23, 195A (see Fig. AD), on the other hand, is typical with winds more east and southeast. maximum and minimum temperature for either day may be ap roximated by comparison of the mean and range isopleths. For all January occurrences, temperatures in all areas are well below long term normalsl during the initial stages. Even maxima are often below the mean normals. In Northern Ontario minima may he in the minus forties. is the centre of the anticyclone moves east temperatures rise throughout the region but eSpecially in Wisconsin, Minnesota, and Thunder Bay which eXperience the greatest shift in wind direction from north to south. In final stages, as the centre eon— tinues to move still farther east, temperatures slightly above normal occur first in Wisconsin and then move east (e.g. coppare January i2 and 23, Figs. 53 and AD). Usually a new type develops before the whole region is above normal. Temperatures to the lee of each lake are always relatively 1For all subsequent references to normal temperatures see Figs. 15-18, Normal Temperature: January, April, July, and october, reopectively. -hg- 3982 swam Eofiézv $27.3 .mE oocom £2.54 V .0“ . a £00. 2 .0 a moo 825 D -I .308 ). 09.3... .33.? V o E. Z .a i :3. .fl 0 .s < a. m a. _ h. 5.825 a. _ Stuns: oocomi. 5:025 r. _ _ 83+ v2.23. h 2:3 o ,.N goo. 5:0 I as! .0 _ _ .oo 2 «out. D 8:233»: 3. D .5253: once: E00. LU>O I £00. 0+ 00 a slab :25 .U 2:3 .0 mocom «Sand V . ‘ ---A s _. J new. 2 I 5 v3. .mm .24., mank .66 On.— .Ed on; N. .Z(_. - 5h - and New York had higher minima on the twelfth, but these were due to the more easterly winds associated with the preceding Low. By mid- afternoon winds were generally more westerly and temperatures were un- modified throughout Ontario and from Minnesota to eastern Indiana. Higher maxima, however, were recorded to the lee of Lakes Michigan and Erie in Lower kichigan, Ohio, and Fennsylvania. New York was exposed to winds from Lake Ontario, but modification here was less significant than the more southerly direction of the winds on the thirteenth. On January 13, northwestern Ontario and Minnesota eXperienced both their highest maxima and highest minima with southerly winds plus the modi- fication by Lake Superior. In Wisconsin maxima were higher with the more southerly winds, but minima were lower because of the greater modification of minima on the twelfth. Illinois, unmodified on the twelfth, recorded higher temperatures throughout the day. Minima in Indiana and Ohio were lower on the thirteenth with unmodified morning westerly winds, but by afternoon winds were southerly and higher maxima were recorded in Indiana. Ohio maxima, like Wisconsin minima, were still lower in spite of southerly winds because of lake modification on the twelfth. In eastern Ontario and New York minima were lower with unmodified westerly winds, but temperatures in southwestern Ontario were modified by Lake Huron; by noon winds were southerly in these areas and higher maxima were recorded. Precipitation associated with the an type is even more clearly dependent on the lakes. January 12 (Fig. 5C) illustrates the distri- bution. To the lee of the lakes the most exposed areas generally re- ceive .20 to .hO inch and occasionally up to .50 inch; less exposed areas _ ,5 - receive .03 to .20 inch. No precipitation occurs to the windward. In April temperatures are similarly well below normal. Maxima and minima in Chicago vary from 420 to 500 F., and 250 to 35° F., res— pectively. In Armstrong, Northern Ontario, minima are in the low twenties and occasionally down to 100 F. United States means are generally 350 to ADO F., though they may get as low as 250 in Chicago or 150 in International Falls. As the winds become more southerly, temperatures in the west and north may rise 50 to 100 F.; minima, nevertheless, are still generally below 32° F. Precipitation is slightly less than in January. Though rare in July and October the type has similar relative characteristics. Temperatures are well below normal, particularly in October, and precipitation is associated entirely with the lakes. In July the lakes are cool and their importance is at a minimum: temper- atures are little modified and precipitation is much lighter and more scattered than in January. In October the lakes are again relatively warm.and modification to temperatures and precipitation to the lee is significant. Minima are below, and averares close to or slightly above freezing. The occurrence of an in any season is extremely important be- cause of its very low temperatures. Numerous occurrences in any one month will strongly reduce average temperatures. In winter an is significant in the formation of snow belts to the lee of the lakes. aLl (Alberta Low through the Lakes) This cyclonic type is primarily a winter type but is not as common as le (see Fig. 2). The variety of the frontal patterns -55.. associated with aLl has already been suggested} TWO major subtypes are distinct: (a) well developed fronts and warm sector; and (c) entirely occluded fronts. Two other subtypes may also be noted: (b) partially occluded fronts with warm sector reaching only to southeastern areas; and (d) only the cold front distinguishable. Subtypes (a) and (b), or (b) and (c) may occur in the same sequence if the process of occlusion takes place over the Lakes (see Fig. 6A). For subtype (a), temperatures of the warm sector, which usuall covers most of the region except Northern Ontario, are most important. Temperatures beyond the warm front and behind the cold front depend on preceding and following types. Mean temperatures of the warm sector are about freezing, varying from 250 to 350 F., or 50 to 100 above nor- mal (e.g. Fig. 6B). Maxima, all above freezing, may reach into the low forties. Temperatures for subtype (c), which has no warm sector, are entirely dependent on the preceding and following types, but the cold front is usually followed by a an or an and the apprOpriate below nor- mal temperatures. The partly occluded subtype (b) brings warm sector temperatures as in (a) but the sector is much narrower and influences only the south- eastern areas. The cold front subtype (d) has poorly deveIOped warm sectors and lower temperatures than subtype (a). Precipitation for all subtypes is fairly heavy in certain areas and widespread in lesser amounts. Nith the Open warm sector subtype (e.g. Fig. 6C), most areas receive over .03 inch. Greater amounts occur 1Above, p. 28. -57.. AMOXS 03¢ SMSOLSJ 30....” NQLODHCV HdelloO omfih 0 in < .Nm 2 1.. m H. o .2 < .«q _ Lia m . .mmlJ 3.205.230»; ¢ _ _ 89:. .F 3.2. 6.32 __ _ _ .o¢ .25 I 59: «.022... _ _ ON. .25 I 69: 22:85 a _ . be. 2 now. \ J _ . F s 50—.- N_ «O .r/ . _ . . 30...:Mcowhcwwime —. .0“. 2..»0. D 30.. .o 22.00 ,\ ._\ . . . - now 0. ano D .. .no. .225 D . x... . . . . ._ .no..8.5 D . li'llll: .8. use 3\33 .. . 39.3 .25. . .... .r , 39523.23. .8 ZO_._.<._._n=ommn_ m. . . . . an... ZO_._.<._._a_ommn_ . 4, . . t. , . C . Motogusalhos am em .3 VA... on: _ endow”. a.m....“fifi ‘ . a... :25 D 55.....qu 825 2% .mN u.~m 2 .2 D man .26 I seen; HJH- 82.» .23. -\._\ _ K? 23.23 mom. 6 .24.. .. .m. _ $2.4meth z..E r. nmm_.o~-¢~.ma<_. .8 2:22.: 50:9: 08¢ taxis“ ...mm:o.: mzmuzhom. z42. em my... I .E.o 0m: .00 $0 zoagoomhz. .6... 0m". nmm. mu >42. I. an. -71.. 750 F., maxima from 800 to 850 F., and minima from.55o to 650 F. hodification by the lakes, north of the front, is quite im- portant. Temperatures, particularly maxima, are much lower to the lee of the lakes. hith northeast winds, stations in the north may even have higher maxima than modified stations in the south. For example, on July A, 1956, maxima and minima respectively at various stations were as a . . . 0 0 .. ,.o . 0 follow.: bault :te. Marie, 71 and A9 F.; Chicago, 09 and 65 F., Cleveland, 710 and 670 F., and International Falls, 750 and 520 F. Jhen the front is over the Lakes, most areas within the pre- cipitation belt receive from a trace to .23 inch (e.g. Fig. low), but various areas may get up to 2 inches and rarely up to 3 inches. When the front is south of the Lakes the belt, lying across the southern lakes, receives slightly less precinitation, but often there are also showers to the lee of the lakes. A close resemblance of this latter case to nfin except for the additonal belt of Precipitation may be noted; however, the overall pattern is definitely the nfieh tyne. Lhen the front is north of the Lakes, there are, at the most, very scattered showers. These occur particularly in the northwest associated with the front. Cn the other hand, when the front is hushed far enough south by the northern Ligh so that it brings no precioitation to the area and the centre of the high is able to move over the Lakes, the day has been classified as nHl. Such a situation is usually accompanied by a weak- ening and disapnearance of the hieh to the Southeast. In october, naeh is fairly rare, but element characteristics ‘ o are much the same as in July. South of the front temperatures are 10 r‘ O 9 ’ O I \ .‘\O o v to 25 above normal with means from o: to be F. in the south. North _ 72 _ of the front temperatures are about normal or slightly above, with means 0 O . J -. ,-.O ,0 T. . . . from 40 to 50 F. in the north and Lo to 60 r. in tne south. Dre- 1? ‘ cipitation occurs as in July. variations 'n the form of the sistem, i... however, are significant. often the southeastern high is not as well developed and tends to centre more over hest Virginia and Cape Hatteras rather than off the southeast coast. The upper-level anticyclone is also more poorly developed and lies over the southeastern states. As a result, the positions of the surface highs vary considerably from east to west and, thus, alter the angle of the fronts across the Lakes. Some,imes the front runs almost north-south or even from northwest to southeast with the northern high over Quebec. This ,ype rarely occurs in January. In April, tanneratures north of the front are slightly above normal. South of the front means run from 500 to 650 F. or about 100 to 150 above normal; maxima vary from 650 to 800 F. and minima from 0 .o a a . . . . .- . . L0 to 50 r. :reCipitation lS conSicerable but varies from not quite as heavy as in July to quite light depending on the flow of maritime tropical air from the south. The liehter p'ecipitation occurs when the high extends farther to the southwest drawinc in cantinental air, or when the flow from the Gulf is blocked by a front south of the high centering ff Georgia and North Carolina. In April there is a tendency for the southeastern high to be more dominant than it is in July and it may gradually push the front, once formed in its east—west trend, to the north rather than allowing it to move south. Furthermore, the type is more frequently introduced by le in this month and, therefore, the front is Lore often fire" oss the entire Lakes in '1 established to the south of the Lakes and may c the reverse direction. As the front is pushed north, fairlv heavy pre- cipitation occurs first in southern areas and then gradually farther north. Temperatures in the south .ehind the front in these c,ses ma; b r0 o 0 ° - a” 6 o r0 n - e l) to L0 above normal, that is, means from 5 to 7) r., maXima 0 r0 ....: .' , E‘ O CO“ from 75 to 8) F. and minima from )5 to 6; x. Type need is a leading source of summer precipitation, and this precipitation is wideSpread because of the fluctuation of the quasi- stationary front. In string amounts are afain considerable but only secondary to those of nLl and sLl types. In summer, average temnera- tures are also noticeably influenced: because of the high dominance of the tvne the relative freeuencies of the frontal ositions may make the pi ) A u difference between a hot, humid season and a comnaratively cool season. nHl (Northern High through the Lakes) This type occurs only in Spring and summer when the lakes are relatively cool. In summer it is usually second in frequency only to the dominant nieH tyne, which it often follows in sequence (see Fig. 2 and 10A). In spring it is much less important and often noorly devel- oped. With nHl is associated cool, cry weather. Temperatures in July for the entire reeion are usually below normal and noticeably colder than regions either to the west or east. Areas in the east with north winds tend to be quite cool, while areas in the west with south winds are markedly warmer. The resulting iso- therm pattern is much the some as the normal .ul‘ nattern, only several _ r) O decrees cooler. Settern mea_s are aporoximately 10 to 12 below normal rith temperatures in the sixties reaching only to haw York or s utheastern -71“. Ontario. Western means are 40 to 70 below normal, with sixty degree temperatures into northwestern Ontario, and they have a remarkably small latitudinal gradient (e.g. Fig. 110). Temperatures to the lee of the lakes, particularly maxima, are significantly lowered-—usually 30 to 60. Modified western areas may be as cool as unmodified eastern areas; but conversely, areas in the east which are modified are much colder than areas in the west which are unmodified. Maximum.temperatures are often as cold or colder in Cleveland than in International Falls and lower in New York than in Minnesota and Uooer Michigan (e.g. Fig. 11A). As the high moves east temperatures rise sliqhtly. There is virtually no precipitation with this type except around the fringes of the region, where it is associated with other attacking systems, or rare showers below .10 inch in lee areas. April temperature patterns are much the same though they are a few degrees warmer, compared to normals. The east is generally some- what colder than the west and areas to the lee of the lakes are asain relatively cooled. Examples are indicated on the accompanying mad for April 28, 1954 (Fig. 118). Unmodified western areas are up to 100 above normal, having means from A00 to 550 F., maxima from 500 to 650 F. 1 . . O 0 ,. . . and minima from 30 to A5 F. M0dified southwestern areas and eastern . ,o + o . 0 areas are about normal with means from A) to 50 F., makima from A5 0 . . . A 0 “0 . . . . . . . to 55 F., and minima from 55 to 45 F. PreCioitation is n11 as in ghilyr. The nhl type is significant primarily in July averaee temperatures. eastern temperatures tend to be lowered and western temperatures at least maintained. The pattern, however, is very close the normal pattern; hence :I. . ['IIII .0u& \. >33... it... >vw. N0 as _,_.a.. swan“... . i > a 4 {.ml m. s. 2.92 angummficmm ..Ek m. Gun 31: era}. \ >33... (.3. 00.... _u._omu \ >993... it... 00.... N. .muu 403?-3 03.2322. Two 93. 403?- 3 00:23:03 two 93. — - a d b ._. .NQSQ. Unmet—4.0.4.02 . 00.... N. _®mw Ac r :\S D can: .3. D :32 «3...... A >33... 326.. D so. 3 .00.. _ g am. 3 m3". . . m... 2 (:3 I .oo. 3 .no. . . . 93.. who." _ _ a--. .. 0.30:2. . . a _ a. _ I 931.8. . 3 cm. _ am. 9». i d. O i. N am. m». t. _ we. 0 oun wwm. Hurting 303.303. Em: $5.0ch $5 H.383 man—um. HMO-I'mg A>HUO39 b0: 70.». . _gL- -75- its importance to monthly averages is difficult to detect from year to year. le (Western High through the Lakes) The anticyclonic type le is by far the dominant system of fall; in October it shows the highe t frequency in a single month of any of the eleven types. In other seasons it is of limited importance. The subtype st occurs only rarely in any of the seasons. In cctober, temperatures are all close to normal. Daily means 0 o for the entire region are between 35 F. and 50 F., and there is very little latitudinal gradient. Areal variation is largely denendent on the degree of modification by the lakes: the accompanying map of mean temperature isotherms for October 20, 195A (Fig. 138) is representative. Minimum temperatures, in particular, may be raised by as much as 10 to the lee of the lakes. For examole, on October 25, 1953, with winds north to northeast, Fort Wayne recorded a 1:30 a.m. temperature of A90 0 F. and Armstrong in northwestern entario only 35 F.; but on the follow; ing day winds were more southerly and Armstrong jumped to tho F. while Fort wayne drOpped to 410 F. Thus, in early stages of the type, when winds are northerly, here is a considerable range between southern and northern minima; but later, when winds are southerly, minima are re- markably uniform or even lower in the south. Maxima, on the other hand, generally remain about the same, or are even lowered slightly by the lakes. The accompanying maps showing range isopleths for a representa- tive occurrence in October of 1953 (Figs. 13C and 13D) clearly demon- strate these differences and chances in modification with changing wind directions. Northern maxima are generally in the high forties and southern maxima in the high fifties or low sixties. Northern unmodified .ig#-|. _.a 4 .....mo 2 .9. a ..o a.m.. :25 D «I; 0 Ln _.\ h. J 1.. .am 0 1... a. .«l. _ .m. 4 Ma. £0N .66 On; :. .. ......xx .....W won. 506:: I £0N .56 On”. 0L“— mom_ 500:: a 5:025 ..- $8 2 .m. D Etta $8 2 .m. B 033.35.?) .....ON .25 a 2:3 .8534 4 meow .30 D \hlmN ...... . . . ...... II .. an. . n . x... N . a: ..x...... /. .\ \ . ._ _. s . . Van..- 3:: .8 . 2:3 . a. vmm. " x . . 1 eam. l . 8.50 2:22.: . 8.50 5:22.: . . a: .mu . . 0.2 i . . . _ mNFoO ZDZ_X<2 . .VN POO IDS—X42 mm . .3 133.5. oou 233.30 . . moz<§ m Clevelandmx *7 nulj Erie mON Lr\(‘*\ mm “\C \I has C ~\O ..‘NW GO U\ (0 C'\ bud" (‘3 9 57 CO 6 an A ‘1/ t 21 “O 72 75 l 53 5h 30 8! 96 4 6 51 A7 36 40 A5 1 2 64 72 7 é? \f.‘ 4T Durham (To Own [rim (\flfi (, ~(’\ NO 1;“ C) {Vi-N Cw?) -filu hm P—L’K 1 ((‘\ \’ :Lr\ \CL.’\ C 'r—4 I“ ‘) C-OJ bxo C(T‘: nLq x—L u J F‘KC (If C) C tn '\ ('3 \. IA :1 ‘ -> \2 '\) itk‘x C ~L\ if(\ C-lfl \OC.‘ e\{) C)" C .... C \0 \U i; C’Yfl mn St.Thoma°. .Penn.0hio Buffalo trxr A"; tab. (7 x’Jx L‘dfl -erm (; \IJ C‘\-" . (two In: - L" \L: OH f,‘ - kt \Ly Mr“ CC‘IN ~Tf'r-l CC-N’D O’f (’L\ C"~\O C -Lr'\ b -L.’\ 0’ CH C‘- U\ (“\C I r ’ 1 'Z‘. 5. 72+ 75 rt» 5: E mx v 1an 61!; C)“ Roche ster mxfi, - lll - A k/ \m 0 0 , if D I ‘ . {1‘ Ln 0 Ln CO d N mn6 (CNO tzfv.‘ f»-z:\ nun I: \D C‘ KO ("MD C NO {CAD cr~ Cf 1 L."\ C\[‘~ (ALA O\\_’7 C-L’\ tnr+ cex -Tr°f [\JJ\ ("\C - Cwl ‘\ F-C \ C \-U'\ " C) O'CNO \OC‘W cw r4\’.‘ if \ '1 D~~C"\ [\-LW\ LAC“ bui C I C‘ “I. C «O rm~ C' «O \O (1'? f:\f‘ (m9 O\Lf\ sr4 C *L’\ {fr—J. b-LN .36) FAA (”‘02 \LC L'\ \C‘ H \F\O NM \T b-L.\ r413 his“ LAO CW 7) O C D N r? 1': ' Cf) L\ (\2 U\ (GAO Eéé 1: Ni) rum Geo L1'\l."\ OLA Lfib~ cm (vi: can t0 h‘\ LBJ“) rem "TO bum {,3 LL\ \O-Q C «0 (”NJ t3\3 C‘.» Ln D «C: O «9 t’, O CC C»- Eéé and Clevel 7e 6 <21? \v MO (ONO r— f" C\Lf'\ :‘ L.\ MN C \\O H H C; ‘\\ ~ : b L."\C"'\ C’\(“\ UNC‘W mic! L\C(\ L;P\O <14: FL“ H (‘3 ( ‘:_‘\ \CC'\ CN—I u\<1 (V‘ (W L1'\ 03 r-{\O Mum O\(\ ml”) 41-. \C\.. U\(\ L"\L{\ (“\C.’ ...: ":‘x <‘("\ t" b. L!\(’\ O\(-’\