UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

River ice conditions in the Nelson drainage system MacKay, Donald Kenning 1962

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1962_A8 M14 R4.pdf [ 6.1MB ]
Metadata
JSON: 831-1.0105663.json
JSON-LD: 831-1.0105663-ld.json
RDF/XML (Pretty): 831-1.0105663-rdf.xml
RDF/JSON: 831-1.0105663-rdf.json
Turtle: 831-1.0105663-turtle.txt
N-Triples: 831-1.0105663-rdf-ntriples.txt
Original Record: 831-1.0105663-source.json
Full Text
831-1.0105663-fulltext.txt
Citation
831-1.0105663.ris

Full Text

BITER ICE CONDITIONS IN THE NELSON DRAINAGE SYSTEM by DONALD KENNING MACKAY B.A., The University of British Columbia, 1959  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in the Department of Geography  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA October, 1962  In presenting  t h i s thesis i n p a r t i a l f u l f i l m e n t of  the r e q u i r e m e n t s f o r an advanced degree a t t h e U n i v e r s i t y  of  B r i t i s h Columbia, I agree t h a t t h e L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and f o r extensive  study.  I f u r t h e r agree t h a t p e r m i s s i o n  c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y p u r p o s e s may  g r a n t e d by the Head o f my Department o r by h i s  be  representatives.  I t i s understood t h a t copying or p u b l i c a t i o n of t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l not be a l l o w e d w i t h o u t my w r i t t e n  Department The U n i v e r s i t y o f B r i t r s h Vancouver 8, Canada.  Columbia,  permission.  ABSTRACT  Hydrological  and meteorological observations related t o i c e con-  ditions on r i v e r s i n the Nelson drainage system are compared s t a t i s t i c a l l y t o determine the measure of agreement between them.  The r e s u l t s  show a high degree o f p o s i t i v e c o r r e l a t i o n . The areal v a r i a b i l i t y o f i c e formation, ice d i s i n t e g r a t i o n , mean length of the i c e - f r e e season, and the standard deviation of f i r s t - i c e and l a s t - i c e are plotted on maps.  The data has been based upon ten-year  mean dates and also the entire 1921 t o 1950 period. The progress o f i c e formation and d i s i n t e g r a t i o n i s examined s t a t i s t i c a l l y both l a t i t u d i n a l l y and also along the major t r i b u t a r i e s of the Nelson River.  Results indicate that f i r s t appearance of ice a f f e c t -  ing discharge generally follows the expected north t o south pattern;  no  systematic progression along t r i b u t a r i e s i n either an upstream o r a downstream d i r e c t i o n i s apparent.  On each major t r i b u t a r y tested, i c e d i s i n -  tegration progresses downstream.  L a t i t u d i n a l progress follows a south to  north pattern with the exception of the southeasterly-flowing portion o f the Assiniboine  River.  Trends and fluctuations i n ice formation and d i s i n t e g r a t i o n are studied by five-point f i l t e r e d series.  Break-up ( l a s t - i c e ) occurred  e a r l i e s t i n the mid-1940 s whereas freeze-up showed no d e f i n i t e trend. 1  F i l t e r e d series of i c e formation dates appear t o exhibit greater cov a r i a b i l i t y than those of i c e d i s i n t e g r a t i o n dates.  The v a r i a b i l i t y of  break-up on headwater streams could be a f a c t o r i n l i m i t i n g the  c o v a r i a b i l i t y between l a s t - I c e records due to the dependency of break-up at downstream s i t e s on upstream conditions. Trends i n the length of i c e - f r e e (open) season f o r 1921 to 1950 are examined using cumulative percentual deviations from the mean.  At most  locations i n the Nelson basin, the length of the ice-free season was shorter than average from 1921 t o the early lgSSO^ and longer than average i n the l a s t h a l f of the record.  Cumulative percentual devia-  tions from the mean i c e - f r e e season are compared to those from the mean annual a i r temperature.  Mean annual a i r temperatures and lengths of  ice-free season do not appear t o be s i g n i f i c a n t l y correlated i n the Nelson basin. The study of factors a f f e c t i n g the formation and d i s i n t e g r a t i o n of r i v e r i c e i n the Nelson basin i s l i m i t e d primarily t o a discussion of the  relationships among a i r temperatures, i c e conditions, and r i v e r d i s -  charge.  The extent and v a r i a b i l i t y of f r e e z i n g and melting degree days  before i c e formation and d i s i n t e g r a t i o n are examined f o r the period 1921 to 1950.  Local a i r temperatures are extremely variable before f i r s t - i c e  and l a s t - i c e dates.  Ice may be reported when temperatures are above the  f r e e z i n g point;  break-up may occur when temperatures are below the  freezing point.  Two possible explanations f o r f i r s t - i c e being observed  under thawing conditions are: upstream moving downstream:  (a) i c e formed under freezing conditions  and (b) the pooling of cold a i r i n en-  trenched v a l l e y s r e s u l t i n g i n i c e formation.  The occurrence of l a s t - i c e  under freezing conditions may be caused by freeze-thaw cycles weakening the  structure of the i c e cover combined with increases i n discharge  r a i s i n g and cracking i t . Mean discharge rates (1921 to 1950) p r i o r t o ice d i s i n t e g r a t i o n are computed and graphed f o r seven locations i n the Nelson basin.  Mean  rates increase three t o f i v e times i n the ten-day period preceding break-up.  Increases i n discharge appear t o be one of the prime f a c t o r s  contributing to i c e disintegration.  ACKNOWLEDGMMTS  The w r i t e r wishes t o acknowledge the advice given by Dr. J . R. Mackay.  Dr. J . D. Chapman, F. A. Cook (Geographical Branch, Department  of Mines and Technical Surveys), members of the Meteorological Branch, Department of Transport, and A. E. Weglo (Water Resources Branch, Department of Northern A f f a i r s and National Resources) have made h e l p f u l suggestions on the manuscript.  TABLE OF CONTENTS Chapter I.  Page  INTRODUCTION Previous studies o f inland i c e conditions  2  The Nelson drainage system  4  A.  The North Saskatchewan River  5  B.  The South Saskatchewan River  7  C.  The Assiniboine-Red River  8  D.  The Milk River  Ice formation and d i s i n t e g r a t i o n data  II.  III.  1  10 10  A.  Freeze-up and break-up data  11  B.  Hydrologic data  11  Objectives  15  Definitions  17  Notation  18  COMPARISON OF DATA  20  Break-up and l a s t - i c e  22  Freeze-up and f i r s t - i c e  33  PROGRESS OF RIVER ICE FORMATION AND DISINTEGRATION Maps  40 40  Chapter  17.  Page Progress of l a s t - i c e  42  Progress of f i r s t - i c e  64  RIVER ICE FORMATION AND DISINTEGRATION TRENDS Trends i n hydrologic data  72  A.  Last-ice  72  B.  First-ice  80  C.  Ice-free season  85  D.  Mean annual a i r temperatures and ice-free seasons  89  Trends i n freeze-up and break-up data  V.  91  A.  Break-up  92  B.  Freeze-up  95  C.  Open season . •  97  ENVIRONMENTAL CONDITIONS AFFECTING RIVER ICE FORMATION AND DISINTEGRATION  ;  Freezing degree days p r i o r t o i c e formation Melting degree days p r i o r t o i c e d i s i n t e g r a t i o n Summed winter temperatures, i c e thicknesses, and i c e d i s i n t e g r a t i o n dates River discharge and i c e d i s i n t e g r a t i o n 71.  71  SUMMARY AND CONCLUSIONS  REFERENCES  98 101 105  I l l 113 117  124  LIST OF TABLES Table 1. 11. 111. IV.  Page F i r s t - i c e records  15  Last-ice Records  15  Numbers of Pairs of Corresponding Observations  22  Means and Mean Differences Between Dates of Break-up and Last-ice  V.  Means and Mean Differences Between Dates of Break-up and Last-ice at Winnipeg, Headingley, and S e l k i r k ....  VI.  Values o f the Wilcoxon Matched-Pairs Signed-Ranks Test..  Vll.  Values o f Sample Correlation C o e f f i c i e n t s , Break-up and Last-ice  Vlll. IX. X. XL.  23  Matrix of Sample Correlation C o e f f i c i e n t s  24 26  28 29  Means and Mean Differences Between Dates of Freeze-up and F i r s t - i c e Values of Sample C o r r e l a t i o n C o e f f i c i e n t s , Freeze-up and F i r s t - i c e L a s t - i c e Progress Along Major T r i b u t a r i e s of the Nelson River  33 38 57  Table  Xll.  XL11.  XIV.  XV.  XVI.  XVII. XVI11.  XIX.  XX.  XXI.  XXII.  Page  Test of Differences Between Records of L a s t - i c e Dates at Latitudinally-separated Stations  65  F i r s t - i c e Progress Along Major T r i b u t a r i e s of the Nelson River  67  Tests of Differences Between Records of F i r s t - i c e Dates at Latitudinally-separated Stations  68  Differences Between Means of Ten-year Break-up Periods . at Winnipeg  94  C o v a r i a b i l i t y Between Paired Break-up Dates  95  Average Number of Cumulated Freezing Degree Days P r i o r to F i r s t - i c e , 1921 to 1950... Years When Mean D a i l y A i r Temperature Was Above Freezing on Date of F i r s t - i c e A f f e c t i n g Discharge, 1921 t o 1950-  102  104  Standard Deviations o f Cumulated Degree Days P r i o r to First-ice  106  Average Number of Cumulated Melting Degree Days P r i o r to L a s t - i c e , 1921 t o 1950  107  Years When Mean Daily A i r Temperature Was Below Freezing on Date of L a s t - i c e A f f e c t i n g Discharge,. 1921 t o 1950..  110  Standard Deviations o f Mean A i r Temperatures and Cumulated Degree Days  I l l  LIST OF FIGURES Figure 1.  2.  Page  F i r s t - i c e Dates; or Longer Last-ice Dates;  S i t e s with, a Record of 10 Years 13 S i t e s with a Record of 10 Years or  Longer  14  3.  Break-up and Last-ice at Prince Albert  30  4.  Break-up and Last-ice at Edmonton  30  5.  Break-up and Last-ice at The Pas  31  6.  Break-up and Last-ice at Saskatoon  31  7.  Break-up and Last-ice at Brandon on the Assiniboine River  32  8.  Freeze-up and F i r s t - i c e at Edmonton  36  9.  Freeze-up and F i r s t - i c e at Prince, Albert  36  10.  Freeze-up and F i r s t - i c e at Brandon on the Assiniboine  11.  , River Mean Dates o f F i r s t - i c e , 1916-25  37 44  12.  Mean Dates of F i r s t - i c e , 1926-35  45  Figure  Page  13.  Mean Dates of F i r s t - i c e , 1936-45  46  14.  Mean Dates of F i r s t - i c e , 1946-55  47  15.  Mean Dates of F i r s t - i c e , 1921-50  48  16.  Standard Deviations of F i r s t - i c e , 1921-50  49  17.  Mean Dates of L a s t - i c e , 1916-25  50  18.  Mean Dates of L a s t - i c e , 1926-35  51  19.  Mean Dates of L a s t - i c e , 1936-45  52  20.  Mean Dates of L a s t - i c e , 1946-55  53  21.  Mean Dates of L a s t - i c e , 1921-50  54  22.  Standard Deviations o f L a s t - i c e , 1921-50  55  23.  Mean Ice-free Season, 1921-50  56  24.  Stations Used t o Check Progress of Ice Conditions  59  25.  Five-Point F i l t e r e d Series of L a s t - i c e ; Edmonton, Prince . Albert, The Pas, Banff  26.  Five-Point F i l t e r e d Series of Last-ice; St. Mary River , International Boundary and Lethbridge, Saskatoon, Medicine Hat  74  76  Figure 27.  28.  29.  30.  31.  Page  Fire-Point F i l t e r e d Series of L a s t - i c e ; Milk RiverTown of Milk River and Eastern Crossing, Millwood, Brandon  78  Five-Point F i l t e r e d Series of Last-ioe; Headingley, Emerson, . Dominion C i t y , Whitemouth  79  Five-Point F i l t e r e d Series of F i r s t - i c e ; Albert, The Pas, Banff  82  Edmonton, Prince  Five-Point F i l t e r e d Series of F i r s t - i c e ; St. Mary River International Boundary and Lethbridge, Medicine Hat, Saskatoon Five-Point F i l t e r e d Series of F i r s t - i c e ; Milk River, Brandon, Headingley, Emerson  84  32.  Locations of Ice Records and Physiographic  33.  Cumulative Percentual Deviations from the Mean Ice-free  Boundaries  Season, 1921 to 1950 34. A Comparison of the Cumulative Percentual Deviations from A,B, the Mean Ice-free Season and from the Mean Annual A i r C. Temperature, 1921 to 1950 D. A Comparison of Open Season and Mean Annual A i r Temperatures 35.  83  Five-Point F i l t e r e d Series of Break-up; Edmonton, Prince A l b e r t , Winnipeg, Selkirk  36. Five-Point F i l t e r e d Series of Freeze-up; Prince Albert, A,B. . Winnipeg. . , C,D. Five-Point F i l t e r e d Series of Open Season; Edmonton, Prince Albert  87  88  90  93  96  Figure 37.  38.  39.  Page  Winter A i r Temperature Regimes and Average Dates of L a s t - i c e , 1921 to 1950  109  Maximum Ice Thicknesses and Break-up Dates on the Assiniboine River, 1943 to 1958  112  Hydrograph of Mean Discharge Rates Preceding and -Following Last-ice Dates, 1921 to 1950  115  CHAPTER I  INTRODUCTION  A r i v e r drainage basin i s a convenient natural region f o r geographic research.  Geographic studies of r i v e r basins may  integration of a l l important phenomena, or they may  deal with the  describe and analyze  the d i s t r i b u t i o n and factors a f f e c t i n g one or more of these phenomena. Studies of the l a t t e r kind, and those dealing with freeze-up and breakup, conservation, land use, sequent occupance, water balance, and so on may  be said to be t o p i c a l o r thematic i n nature but, nevertheless, they  are regional i n scope. Probably the most important comprehensive studies of r i v e r basins i n Canada are the Ontario Conservation Reports (e.g. Ontario Department of Commerce and Development, Spencer Creek Conservation Report, I960).  These reports not only deal with the conservation of r e -  sources but also with t h e i r multiple-purpose development.  In the  Province of B r i t i s h Columbia, the Fraser River basin has been intensively studied with p a r t i c u l a r reference to flood control and h y d r o e l e c t r i c power p o t e n t i a l .  The preliminary report of the Fraser River Board (1958)  discusses various f l o o d control and power generation plans and t h e i r effects on f i s h e r i e s , i r r i g a t i o n , navigation, and the economy of the region as a whole.  Unfortunately, studies of other basins i n Canada  2, have generally been l e s s concerned with comprehensive regional development. There have been numerous publications dealing with the geography of the Canadian P r a i r i e s but r i v e r i c e conditions i n the p r i n c i p a l drainage basin of t h i s area, the Nelson, have not been extensively examined. A number of the t r i b u t a r y basins of the vast Nelson complex have been studied f o r purposes of determining i r r i g a t i o n and hydroelectric potent i a l at s p e c i f i c stream s i t e s , and f o r purposes of f l o o d control.  Al-  though the engineering planning and construction of dams requires an i n tensive examination and a d e t a i l e d knowledge of the physical and hydrometeorological  c h a r a c t e r i s t i c s of the basin area involved, comparatively  l i t t l e attention has been given to ice conditions. It  i s the purpose of t h i s study to describe the a r e a l v a r i a b i l i t y  of r i v e r i c e formation and d i s i n t e g r a t i o n i n the Nelson drainage system; t o analyze the progress of i c e conditions along the r i v e r s i n spring and fall;  and t o discuss the environmental factors a f f e c t i n g the phenomena  of r i v e r i c e formation and d i s i n t e g r a t i o n .  Knowledge of r i v e r i c e con-  d i t i o n s i s of p r a c t i c a l value and t h e o r e t i c a l importance not only to geographers but also t o individuals and groups i n such f i e l d s as meteorology, a g r i c u l t u r e , f o r e s t r y , and transportation. t i o n may  Eventually, such informa-  prove h e l p f u l i n freeze-up and break-up p r e d i c t i o n . PREVIOUS STUDIES OF INLAND ICE CONDITIONS Very few studies have been made of inland ice conditions i n  Canada so f a r as freeze-up and break-up are concerned.  In other coun-  3.  t r i e s , p a r t i c u l a r l y the Soviet Union, a considerable amount of research has been done on r i v e r ice conditions.  For example, such writers as  Chernoevanenko (1953), Shebanov (1958), and Sokolov (1955) have discussed ice conditions on various parts of the Don, Duna, Dvina, Southern Bug, Dneiper, and Neva Rivers.  Many a r t i c l e s dealing with i c e conditions on  other r i v e r s of the Soviet Union are l i s t e d i n the abstract volumes of the Snow, Ice, and Permafrost  Research Establishment  of the United States  Army. A few i c e condition studies pertinent t o t h i s t h e s i s have been c a r r i e d out i n the United States.  For example, Hoyt (1913) and Parsons  (1940) have discussed the e f f e c t s of i c e on streamflow;  Shipman (1938)  has analyzed the e f f e c t s of winter temperatures on the closed season of the M i s s i s s i p p i River at Davenport;  and Williams (1955) has described  the f r e e z i n g of the Yukon River, Alaska, i n 1949 and i t s break-up the following spring. In Canada, freeze-up and break-up on the Mackenzie River has been studied by Mackay (1961a; 1961b) who  has analyzed dates of ice f o r -  mation and d i s i n t e g r a t i o n with p a r t i c u l a r reference to the lower section between Fort Good Hope and the Beaufort Sea. Henoch, 1961; Kindle, 1920;  Lloyd, 1943)  Other studies (Brown, 1957;  deal with Mackenzie i c e con-  d i t i o n s or touch upon these phenomena i n c e r t a i n areas or at s p e c i f i c locations.  In the P r a i r i e Provinces, Burbridge and Lauder (1957) have  mapped the average dates of freeze-up and break-up and discussed the meteorological f a c t o r s a f f e c t i n g these phenomena.  Their study does not  u t i l i z e hydrologic data r e l a t e d t o i c e conditions a f f e c t i n g discharge  4.  nor does i t deal exclusively with r i v e r s .  Gurrie (1954) discusses the  factors determining i c e formation and d i s i n t e g r a t i o n on lakes and r i v e r s i n the Canadian P r a i r i e s but these faetors are not related t o i c e cond i t i o n s at s p e c i f i c locations. A number of detailed studies of the environmental f a c t o r s a f f e c t ing dates of i c e formation and d i s i n t e g r a t i o n have been made.  Some o f  the studies pertain to freeze-up and break-up processes i n running water (e.g. C u r r i e , 1954; Devik, 1944; Murakami, 1955) while others deal exc l u s i v e l y with these processes i n standing bodies of water (e.g. Calloway, 1954; Kennedy, 1939). Unfortunately, many t e c h n i c a l papers dealing with i c e conditions that have been published i n other countries are not available to the writer.  However, i t seems reasonable to assume that the a v a i l a b l e l i t e r -  ature provides a good cross-section of the methods and techniques used i n the examination of r i v e r i c e conditions.  Those which seem most a p p l i c a -  ble to the data available f o r the Nelson drainage system have been used i n t h i s study.  THE NELSON DRAINAGE SYSTEM The Nelson River and i t s t r i b u t a r i e s form one of the great d r a i n age basins of North America with an area of approximately 414,000 square miles.  About 375,550 square miles are i n Canada, and the remainder,  38,450 square miles, are i n the United States.  Riverine a r t e r i e s run  eastward and northward from the headwaters of the Bow River i n the Rocky Mountains t o the mouth of the Nelson River i n Hudson Bay, a distance of  5.  1,740  miles.  One branch from Lake Winnipeg stretches 450 miles east to  the Steep Rock area of Ontario, while the headwaters of another t r i b u t a ry, the Red River, are found 500 miles south of Lake Winnipeg i n the north central United States.  The basin extremities are bounded by the  drainage systems of the P a c i f i c coast i n the west, the Mackenzie and C h u r c h i l l Rivers i n the north, Hudson Bay and St. Lawrence River i n the east, and by the M i s s i s s i p p i River drainage i n the south. The area drained by the Nelson River i n Canada includes most of the valuable a g r i c u l t u r a l land of the P r a i r i e Provinces.  The  southern  Canadian P r a i r i e s , however, are not wholly drained by the Nelson River system (Figure 1 ) , f o r the M i s s i s s i p p i River system drains 22,155 square miles of Canadian t e r r i t o r y .  4  The Canadian portions of the Nelson and  M i s s i s s i p p i drainage basins form the focus of t h i s study.  For purposes  of s i m p l i c i t y , the combined areas of these two systems are hereafter r e ferred to as the Nelson drainage system.  The area under consideration  extends over some 397,710 square miles of central Canada (Thomas, 1956; 1958). The p r i n c i p a l t r i b u t a r i e s from which i c e formation and d i s i n t e gration records are drawn are the North Saskatchewan, South Saskatchewan, and Assiniboine-Red Rivers of the Nelson basin as well as the M i l k River of The M i s s i s s i p p i basin.  A.  The North Saskatchewan River The North Saskatchewan r i s e s on the eastern slopes of the Rocky  % i O c a l and inland drainage i n the Cypress H i l l s region of Southern Alberta and Saskatchewan are included i n the 22,155 square miles.  6  Mountains west of Jasper, Alberta, and flows i n an easterly d i r e c t i o n to where i t joins the South Saskatchewan, a few miles below Prince A l b e r t , Saskatchewan.  The snowfields and g l a c i e r s of the mountainous headwater  region provide the greatest portion of the r i v e r ' s water supply.  In t h i s  region, there i s l i t t l e ground water storage because of the rocky t e r r a i n , steep slopes, high gradient, and other f a c t o r s .  Consequently, variable  climatic conditions can produce great ranges i n stage and i n discharge rates over r e l a t i v e l y short periods of time. Between the Rocky Mountains and Great P l a i n s , the North Saskatchewan i s joined by two of i t s p r i n c i p a l t r i b u t a r i e s , the Clearwater and Brazeau Rivers.  In t h i s section, the main stream and i t s t r i b u t a r i e s  flow through deep v a l l e y s with moderate t o steep slopes.  Here, the vege-  t a t i v e cover tends to reduce the surface runoff supplied to the r i v e r s . The North Saskatchewan enters the Great Plains a few miles west of Edmonton and flows east t o i t s junction with the South Saskatchewan. The main stream runs through a well-defined v a l l e y which gradually broadens and f l a t t e n s between North Battleford and Prince Albert. section, the r i v e r i s braided and i t s gradient i s s l i g h t .  In t h i s  The main t r i -  butaries entering the North Saskatchewan between Edmonton and Prince ALbert are the Sturgeon, Vermilion and Battle Rivers.  These t r i b u t a r i e s  supply l e s s than ten per cent of the mean annual discharge t o the r i v e r . Below Prince A l b e r t , the r i v e r gradient steepens and the v a l l e y narrows to the point of junction with the South Saskatchewan and beyond. Downstream from the junction of the North and South Saskatchewan, the combined r i v e r i s c a l l e d the Saskatchewan.  North of The Pasquia H i l l s ,  7  the Saskatchewan flows through an area of marsh and bog where I t s p l i t s into two p r i n c i p a l channels.  These channels, c a l l e d the Old and the  New,  j o i n near Cumberland House and the r i v e r continues flowing i n an easterly d i r e c t i o n past The Pas.  Discharge at The Pas i s regulated by the lake  and bog system to the west;  consequently, the v a r i a b i l i t y i n stage i s  l e s s marked than at gauging stations located f a r t h e r upstream.  East of  The Pas, the r i v e r runs through a series of lakes which f u r t h e r regulate the flow passing into Lake Winnipeg.  The p r i n c i p a l t r i b u t a r i e s entering  the main stream east of the confluence of the North and South Saskatchewan are the Torch, Carrot, and Pasquia Rivers.  B.  The South Saskatchewan River The South Saskatchewan r i s e s i n many branches on the slopes of  the Rooky Mountains. Bow,  I t s main headwater branches are the Bow,  Oldman, Waterton, B e l l y , and St. Mary Rivers.  Little  These streams have  r e l a t i v e l y steep gradients i n t h e i r source areas but gentler gradients as they cross the f o o t h i l l s into the p l a i n s .  The mountainous sections  are varied i n character as steep-walled v a l l e y s and bare rock are i n t e r spersed with areas of heavy forest cover. The f o o t h i l l s - p r a i r i e t r a n s i t i o n a l zone i s sparsely wooded except along the banks of the deeply entrenched t r i b u t a r i e s .  The  streams  meander t i g h t l y as they flow through the r o l l i n g tablelands west of Lethbridge.  The waters of the Waterton, B e l l y , and St. Mary Rivers are  added to the flow of the Oldman River between Monarch and Lethbridge. From Lethbridge eastwards, the meanders tend t o become more widely spaced  8.  as the flow from the L i t t l e Bow, Bow, and smaller streams enters the main r i v e r channel.  At Medicine Hat, the South Saskatchewan veers northeast  through a shallow gorge i n the Middle Sand H i l l s t o the junction with i t s Bed Deer t r i b u t a r y at Empress.  Here, the r i v e r turns east again to the  town of Elbow and then northeast t o i t s confluence with the North Saskatchewan. Just t o the north o f the town of Elbow, the South Saskatchewan dam project i s located.  This project, when the two main dams are com-  pleted, w i l l create a r e s e r v o i r with a storage capacity o f eight m i l l i o n acre feet and a length of some 140 miles.  These dams w i l l have a con-  siderable e f f e c t upon dates of i c e formation and d i s i n t e g r a t i o n both i n the upstream area of the r e s e r v o i r and i n the downstream sections affected by t h i s regulating agency. C.  The Assiniboine-Bed River The Assiniboine and Red Rivers j o i n at Winnipeg t o flow north  into Lake Winnipeg.  As the Bed Biver t r a v e l s only a short distance i n  Canada before i t s confluence with the Assiniboine, the two streams are discussed together. The Assiniboine River r i s e s i n Saskatchewan, east of the Q u i l l Lakes.  I t i s spring-fed i n i t s upper reaches  (Canada, Department of  Northern A f f a i r s and National Resources, Water Resources Paper No. 7), and flows southward through a well-defined v a l l e y t o the Oak Lake d i s t r i c t of Manitoba.  From there, i t flows eastward through the Manitoba  escarpment and across the Manitoba lowlands.  I t i s walled by steep  9.  banks i n i t s upper and middle reaches.  Along i t s v a l l e y bottom, the  meandering course of the r i v e r has resulted i n the formation of numerous oxbow lakes and meander scars.  The p r i n c i p a l t r i b u t a r y streams entering  the r i v e r i n i t s middle and upper reaches are the S h e l l , Qu'Appelle, and Souris Rivers. Hast of i t s junction with the Souris, the Assiniboine  crosses  the f l a t p r a i r i e i n a steep-banked channel i n c i s e d a few feet below surface l e v e l .  Spring floods are a constant threat i n t h i s area.  In an  e f f o r t to decrease f l o o d p r o b a b i l i t i e s , meander necks have been a r t i f i c i a l l y cut through i n some sections.  This serves to increase the e f f e c -  t i v e gradient of the r i v e r by reducing the distance from i t s head to i t s mouth.  Eventually, the r i v e r channel should become more deeply i n c i s e d  as i t moves towards equilibrium.  Changes i n regime, whatever the cause,  w i l l have an e f f e c t upon dates of ice formation and d i s i n t e g r a t i o n . The  Red River enters Canada at Emerson, Manitoba and flows north  past i t s junction with the Assiniboine at Winnipeg to discharge into the southern extremities of Lake Winnipeg.  The Canadian section of the  Red  i s s i m i l a r i n many respects to the lower reaches of the Assiniboine. Its channel, cut but a few feet below p r a i r i e l e v e l , i s sinuous and steep-banked; i t s flow i s sluggish due to the low gradient and the r i v e r frequently overruns i t s banks i n spring.  The northerly d i r e c t i o n of flow  i s a f a c t o r which contributes towards f l o o d i n g during the spring break-up period.  The p r i n c i p a l t r i b u t a r i e s j o i n i n g the Red between the International  Boundary and Winnipeg are the Roseau and Rat Rivers.  10. D.  The Milk River The Milk River r i s e s i n Montana, on the slopes of the Rocky  Mountain F o o t h i l l s , and flows northeast into Canada i n two main branches. These branches j o i n together south of the Milk River Ridge, west o f the town of M i l k River.  The r i v e r winds east through highly dissected topo-  graphy that i s often referred to as •badlands*.  This area lacks p r e c i -  p i t a t i o n and vegetative cover; consequently, the l i m i t e d runoff i s poorly regulated.  Following the spring break-up, streamflow r i s e s to a peak and  then subsides rapidly.  In many years, the smaller streams dry up com-  p l e t e l y before the end o f summer. The Milk River flows southeast through the International Boundary at Eastern Crossing, longitude UO'SS^O" west.  Four of i t s p r i n c i p a l  t r i b u t a r i e s , the Lodge, B a t t l e , Frenchman, and Poplar Rivers, r i s e i n Canada but join the Milk i n the United States. ICE FORMATION AND DISINTEGRATION DATA A lack o f data has placed l i m i t a t i o n s on the study o f i c e f o r mation and d i s i n t e g r a t i o n i n many r i v e r basins.  This lack i s p a r t i c u l a r -  l y noticeable along r i v e r s which have not been u t i l i z e d as primary a r t e r ies of transportation and communication.  On the other hand, where r i v e r s  have been used as waterways and where the open water season has been of some importance to the welfare of communities, a considerable amount of freeze-up and break-up data has been recorded.  These data as w e l l as  hydrologic data of i c e conditions a f f e c t i n g discharge are used i n t h i s thesis.  11. A.  Freeze-up and Break-up Data H i s t o r i c a l records of freeze-up and break-up have been recorded  on a non-professional basis by i n d i v i d u a l s and groups connected various settlement, t r a d i n g , and missionary organizations.  with  With l a n d  routes rather than waterways providing the primary base f o r early transportation and communication i n the southern Canadian P r a i r i e s , l i t t l e interest was shown i n recording freeze-up and break-up dates.  As a r e -  s u l t , few long-term records are available f o r the Nelson drainage system. Freeze-up and break-up observations have been recorded at twenty locations i n the Nelson drainage system.  These data have been compiled  and published by the Meteorological Branch, Department of Transport, Canada (CIR - 3156, ICE - 2, 30 JAN 59).  The records vary i n length,  periods of coverage, and i n the way break-up and freeze-up are defined. The areal v a r i a b i l i t y of i c e formation and d i s i n t e g r a t i o n i n the Nelson drainage system cannot be adequately analyzed from these records and, therefore, additional data must be obtained from other sources, B.  Hydrologic Data Hydrologic data (Water Resources Branch, Department of Northern  A f f a i r s & National Resources, Ottawa, Water Resources Papers, 1912-) dealing with i c e conditions i n r i v e r s and streams may  be used to chart  the progress of freezing and thawing of r i v e r s and to broaden the general understanding of these processes.  Ice conditions are reported at gauging  stations when the r e l a t i o n of the elevation of the water surface to the discharge i s affected by i c e .  This stage-discharge r e l a t i o n can be  12.  affected by various types of i c e and varying amounts of i c e cover.  An  increase i n stage caused by the formation of i c e o r the grounding o f i c e blocks i n the stream channel on s i t e or i n the sections downstream cont r o l l i n g the stage-discharge r e l a t i o n may r e s u l t i n i c e conditions being reported.  The control sections are formed by natural or i n some cases,  a r t i f i c i a l weirs which may act as b a r r i e r s to the f r e e passage of i c e blocks.  I f the passage of i c e i s obstructed i n the c o n t r o l l i n g section,  or sections, a change i n the stage w i l l occur at the gauging s i t e . The extent of hydrologic data related t o i c e conditions a f f e c t i n g discharge i n the Nelson, M i s s i s s i p p i , and Athabasca River drainage basins i s indicated i n Tables 1 and 2.  The d i s t r i b u t i o n of stations with ten or  more years of f i r s t - i c e and l a s t - i c e dates i s noted i n Figures 1 and 2, respectively.  In some instances l o c a t i o n a l dots represent more than one  gauging s t a t i o n .  Athabasca River records were used as aids i n mapping  mean dates of i c e conditions.  15. TABLE I FIRST-ICE RECORDS  Drainage Basin Nelson River  10 or more Tears of Record  Stations with: 20 or more Years of Record  30 or more Years of Record  50  29  18  M i s s i s s i p p i River  4  1  1  Athabasca River  3  2  0  5?  32  Total  19  TABLE I I LAST-ICE RECORDS  Drainage Basin  10 o r more Years of Record  Stations with: 20 o r more Years of Record  30 or more Years of Record  Nelson River  79  44  24  M i s s i s s i p p i River  28  12  5  3  2  0  110  58  29  Athabasca River Total  OBJECTIVES The major objectives of t h i s study are as follows: 1.  t o correlate two kinds o f data, (a) hydrologic data of i c e cond i t i o n s a f f e c t i n g discharge recorded by observers appointed by the Water Resources Branch, Department of Northern A f f a i r s and National Resources and (b) h i s t o r i c a l records o f freeze-up and break-up kept by various i n d i v i d u a l s and organizations on a non-professional basis;  16, 2.  to extend i c e formation and d i s i n t e g r a t i o n data and to increase the coverage i n c e r t a i n areas;  3.  to determine the progress of r i v e r i c e formation and d i s i n t e g r a t i o n i n the Nelson drainage system;  4.  to examine the progress of i c e formation and i c e d i s i n t e g r a t i o n on i n d i v i d u a l t r i b u t a r i e s of the Nelson River;  5.  to describe and compare f l u c t u a t i o n s and trends i n i c e condition dates and i n lengths of ice-free or open season throughout  the  Nelson drainage system; 6.  to examine the c o v a r i a b i l i t y between records of i c e dates i n order to assess the r e l a t i v e importance o f o v e r - a l l c l i m a t i c controls and l o c a l environmental controls i n the processes of ice formation and ice d i s i n t e g r a t i o n ;  7.  to discuss the extent and v a r i a b i l i t y of freezing degree days f o r a given period p r i o r to the onset of i c e conditions a f f e c t ing discharge;  8.  to determine the extent and v a r i a b i l i t y of melting degree days f o r a given period p r i o r to dates of l a s t - i c e a f f e c t i n g d i s charge ;  9.  to indicate the r e l a t i o n s between summed winter  temperatures,  ice thickness data, and break-up dates; and 10.  to i l l u s t r a t e and discuss discharge rates p r i o r t o i c e d i s i n t e gration.  17.  DEFINITIONS Some expressions used i n t h i s t h e s i s have d i f f e r e n t connotations. To avoid ambiguities, these expressions are defined below i n the s p e c i f i c sense i n which they apply to t h i s study.  Meanings of terms that are r e -  l a t e d to hydrologic data are those of the Water Resources Papers p u b l i s h ed by the Department of Northern A f f a i r s and National Resources.  The  terms are l i s t e d alphabetically: "between v a r i a b i l i t y " : - the v a r i a b i l i t y between two series of observations "control or c o n t r o l l i n g section": - the section or sections of the stream channel below the gauging s i t e which controls the stagedischarge relationship at the gauging  site  " c o v a r i a b i l i t y " : - the degree of association between the variance of one v a r i a b l e and the variance of a second v a r i a b l e "drainage system":- the drainage within an area; i t involves more than one r i v e r drainage basin " f i l t e r e d series'* - a moving average with non-equal weights " f i r s t - i c e " : - the f i r s t i c e a f f e c t i n g the normal stage-discharge r e l a t i o n curve i n the f a l l season " l a s t - i c e " : - the l a s t i c e a f f e c t i n g the normal stage-discharge r e l a t i o n curve i n the spring season " r i v e r drainage basin":- a region bounded by other basins; i t includes the whole of the catchment area " r i v e r system":- the r i v e r , i t s t r i b u t a r i e s , and c o l l e c t i n g basins  18.  "stage-discharge r e l a t i o n " : - the r e l a t i o n between the elevation of the water surface at the gauging s i t e and the rate o f flow i n the r i v e r . "stationary  s e r i e s " : - a series without a trend; one with a  constant mean and variance over a given period of time "within v a r i a b i l i t y " : - v a r i a b i l i t y within a series of observat i o n s rather than between two series of observations  NOTATION The notation used i n t h i s study i s l i s t e d below: T T.  t e s t value of the M l c o x o n matched-pairs signed-ranics t e s t T.  0 5 j  f i v e per cent and one per cent significance l e v e l s f o r the  0 1  Wileoxon matched-pairs signed-ranks test z  Fisher's "z" transformation, i e z =  1.15 l o g ( l ^ r ) 1 Q  . r r.05  (1-r)  sample c o r r e l a t i o n V  »QI  f  i  v  e  P  e r  coefficient  cent and one per cent significance l e v e l s f o r the  sample c o r r e l a t i o n c o e f f i c i e n t , E  index of p r e d i c t i v e where K = \j 1-r , 2  Ti  independent  YJL  dependent  e f f i c i e n c y , i . e . , 1! =  100 per cent (1-K)  the c o e f f i c i e n t of a l i e n a t i o n  variable  variable  the cumulated percentile  deviation i n the year i  the length of i c e - f r e e season i n the year i L  the mean length of i c e - f r e e season, 1921 to 1950  19. In summary, research into r i v e r i c e formation and d i s i n t e g r a t i o n i n the Nelson drainage system i s severely l i m i t e d by a lack of h i s t o r i c a l records of freeze-up and break-up.  I f hydrologic records o f i c e condi-  tions can be used t o supplement the freeze-up and break-up data that i s a v a i l a b l e , then some of the geographical aspects of r i v e r i c e formation and d i s i n t e g r a t i o n i n the Nelson drainage system can be examined.  Thus,  the i n i t i a l objective of this study i s to correlate these two kinds of data.  A s i g n i f i c a n t c o r r e l a t i o n between these data would permit the  r e a l i z a t i o n of the other objectives outlined i n t h i s chapter.  CHAPTER I I  COMPARISON OF DATA  At some point early i n the winter season, formation o f r i v e r i c e w i l l cause the stage-discharge r e l a t i o n curve to deviate from i t s normal path.  In hydrologic records, the point of deviation i s recorded as the  date of f i r s t - i c e a f f e c t i n g discharge and no subsequent change i n physical conditions can a l t e r the date of occurrence.  Unlike f i r s t - i c e dates,  freeze-up dates are subject to change i f climatic or hydrologic conditions following the i n i t i a l formation o f i c e cause the i c e t o disappear and r e form at a l a t e r date. In the spring, the d i s i n t e g r a t i o n of the main ice cover at the gauging s t a t i o n may or may not r e s u l t i n the f i n a l date of i c e conditions a f f e c t i n g discharge being recorded.  Jamming of i c e at some point down-  stream may cause the water l e v e l t o r i s e t o the extent that i c e cond i t i o n s may again be reported by the observing o f f i c e r .  In contrast to  dates o f l a s t - i c e , break-up dates are not susceptible t o change i f the phenomenon i s defined as the i n i t i a l movement of i c e downstream i n the spring season (Mackay, 1961b).  I f an a l t e r n a t i v e d e f i n i t i o n which im-  p l i e s complete clearance of i c e i s used t o define break-up, then a movement of i s o l a t e d i c e pans from upstream past the observation s i t e could cause a change i n the date of t h i s phenomenon.  21. From previous statements, i t seems c l e a r that a l l four sets of data are composed of complex variables lacking precise d e f i n i t i o n .  It  i s also c l e a r that i n i t i a l and f i n a l stages of i c e conditions a f f e c t i n g discharge are related fundamentally to freeze-up and break-up.  Any com-  bination of climatological elements that creates optimum conditions f o r f i r s t - i c e w i l l create optimum conditions f o r freeze-up i f those conditions are sustained f o r a reasonable length of time.  A s i m i l a r statement may be  made regarding the r e l a t i o n s h i p between break-up and l a s t - i c e .  I f sets of  data are paired i n sense, and a one-to-one time relationship i s set up between the observations, the differences may be tested f o r t h e i r s i g n i f i cance using appropriate s t a t i s t i c a l methods. Four locations i n the Saskatchewan t r i b u t a r y basin and one l o c a t i o n i n the Assiniboine t r i b u t a r y basin have corresponding records of i c e dates which may be extracted from meteorologie and hydrologic reports. It should be emphasized that the locations are general ones and that the observation s i t e s may o r may not be the same f o r each type o f data. Possible differences i n record due t o the s i t e f a c t o r are not considered here; the s i t e s are assumed to be i d e n t i c a l f o r t e s t i n g purposes.  The  number of years i n which corresponding observations from paired sets of data are recorded at each l o c a t i o n i s l i s t e d i n Table H I . In southern parts of the Nelson basin, no corresponding observat i o n s of i c e conditions recorded at s p e c i f i c locations other than Brandon are available from published sources of data.  In the interests of i n -  creased coverage, ice records from three locations i n close proximity t o each other are examined to ascertain the extent of e x i s t i n g relationships.  22. TABLE I I I NUMBER OF PAIRS OF CORRESPONDING OBSERVATIONS  Location  River  Comparable Years of Record Break-up Freeze-up & First-ice & Last-ice  Edmonton  Saskatchewan  35  27  Prince Albert  Saskatchewan  38  37  Saskatoon  Saskatchewan  14  None  The Pas  Saskatchewan  19  None  Brandon  Assiniboine  13  12  Headingley, on the Assiniboine B i v e r , l i e s 12 miles west of the junction of the Assiniboine and Red Rivers at Winnipeg; S e l k i r k , on the Red River, i s situated 23 miles downstream from Winnipeg.  Dates of f i r s t - i c e and  l a s t - i c e at Headingley are compared with dates o f freeze-up and break-up at Winnipeg; break-up at S e l k i r k i s compared with break-up at Winnipeg and with l a s t - i c e at Headingley. BREAK-UP AND LAST-ICE Means and mean differences of break-up and of l a s t - i c e are l i s t e d i n Tables IV and V f o r ten-year periods and f o r t o t a l length o f the periods covered by the data.  Break-up at Edmonton, Prince Albert, and Winnipeg  refers to dates on which the water was c l e a r of i c e ; break-up at Saskatoon, S e l k i r k , and The Pas refers t o dates of i n i t i a l movement of i c e downstream. Break-up at Brandon ( i ) refers to i n i t i a l ice movement, and Brandon ( i i ) refers to complete clearance of i c e .  TABLE IV MEANS AND MEAN DIFFERENCES BETWEEN DATES OF BREAK-UP AND LAST-ICE  Location  River  Edmonton  N.Saskatchewan  Period 1916-25  1926-35 1936-45 1946-55  1915-55  Prince Albert  N.Saskatchewan  S.Saskatchewan  Brandon ( i )  Saskatchewan  Assiniboine  Assiniboine  «  "  1  4  16  Break-up  2  3  "  20  0  3  "  16  2  2  11  2  3  6  1  Last-ice Break-up Break-up  Apr.23  H  1 4  Earlier Phenomena  ••  1  Apr.  7  it ti  II  it  24  2  3  1926-35  "  20  1  0  II  n  1936-45  "  18  "  20  0  2  II  II  1946-55  "  18  "  21  0  3  II  II  II  II  1946-55  1936-45  Apr.21  "  19  Apr.10 "  9  Apr.29  1946-55  «  24  1936-55  "  26  1946-55 1943-55  Brandon ( i i )  M  Mean Difference in Days  20  1942-55  The Pas  Apr.20 •1 yj  Omissions i n Period  "  1916-25  1915-55  Saskatoon  Mean Dates Break-up Last-ice  1946-55 1943-55  Apr.12 w  10  Apr.19 M  18  "  21  Apr.13 n  1  2  3  2  0  3  0  3  1  1  26 27  0  2  1  1  Apr.13  0  1  0  3  Apr.28 " ••  ..  1  3  Apr.13 "  13  0  6  0  5  Break-up II  it  Last-ice  it n  it it  Break-up n  it  Last-ice  it tt  TABLE V MEANS AND MEAN DIFFERENCES BETWEEN DATES OF BREAK-UP AND LAST-ICE AT WINNIPEG, HEADINGLEY, AND SELKIRK Winnipeg Break-up Apr. 3 " 6 " 10 " 12 " 8 Selkirk Break-up Apr.14 " 14 " 11 " 14 Selkirk Break-up Apr.14 " 14 " 11  n  1 4  Headingley Last-ice Apr.18 " 18 " 16 " 15 " 16 Headingley Last-ice Apr.18 " 16 " 15 " 16 Winnipeg Last-ice Apr. 6 " 6 II M  1  2  8  Years 1916-25 1926-35 1936-45 1946-55 Total Record  Years 1926-35 1936-45 1946-55 Total Record  Years 1926-35 1936-45 1946-55 Total Record  Omissions 1 0 0 0  Omissions 0 0 0  Omissions 0 0 0  Mean Differences in Days  Earlier Phenomena  15 12 6 3 8  Break-up i» tt tl M ti n it it  Mean Differences in Davs 4 2 4 2 Mean Differences i n Davs 8 4 1 6  Earlier Phenomena Break-up II  II  tt tt  tt  If  Earlier Phenomena N/A tt tt it  ro  25*.  There are 106 matched pairs of observations i n the sets of data recorded at the four locations i n the Saskatchewan t r i b u t a r y basin.  The  o v e r a l l average indicates that break-up occurs two days e a r l i e r than l a s t ice.  The largest discrepancy between any matched p a i r of observations i s  20 days.  This difference was recorded i n observations taken at Saskatoon  i n 1946.  In the Assiniboine-Red t r i b u t a r y basin, the computation of a  grand mean difference i s meaningless because of the distances separating observation s i t e s .  The greatest difference between paired observations  was 32 days which separated break-up at Winnipeg and l a s t - i c e at Headingley i n 1922. Mean differences may be greatly d i s t o r t e d by extreme values.  Large  differences between paired observations can a r i s e from a combination of physical factors or, possibly, from an e r r o r i n recording o r t r a n s c r i b i n g the day or month i n which one or the other of the observations took place. The occurrence of such an error could conceivably result i n a n i l mean difference i f the d i r e c t i o n or sign of the difference caused by the e r r o r i s opposed t o the norm.  The Wilcoxon matched-pairs signed-ranks test  (Siegel, 1956, pp. 78-83) was used t o avoid t h i s p o s s i b i l i t y . The Wilcoxon matched-pairs signed-ranks test i s non-parametric; unl i k e standard s t a t i s t i c a l t e s t s of rank, i t determines the s i g n i f i c a n c e of the magnitude as well as the d i r e c t i o n of the difference between paired sets of data.  However, t h i s test i s most e f f e c t i v e when i t i s used to com-  pare sets of data of a f a i r l y high degree of known accuracy.  With data  that may include the occasional error of seme magnitude, the s e n s i t i v i t y of the t e s t i s such that a cautious approach must be adopted i n i n t e r p r e t i n g  26. TABLE VI VALUES OF THE WILCOXON MATCHED-PAIRS SIGNED-RANKS TEST  Location  N  T  Edmonton (a) (b)  30 23  Prince Albert  Significance of Differences  —  -01  T.05  -1.40 42  2.58 55  1.98 73  Not s i g n i f i c a n t Sig. at 1 per cent l e v e l  36  -4.61  2.58  1.96  Sig.  The Pas  19  62  38  46  Not s i g n i f i c a n t  Saskatoon  12  12  10  14  Sig.  Brandon ( i ) (ii)  10 12  11 2.5  3 7  8 14  Not s i g n i f i c a n t Sig. at 1 per cent l e v e l  the results.  T  at 1 per cent l e v e l  at 5 per cent l e v e l  The significance o f the differences between records of  break-up and l a s t - i c e are l i s t e d i n Table VI. The n u l l hypothesis that no difference exists between break-up and l a s t - i c e can be accepted at Edmonton (a), Brandon ( i ) , and The Pas. The r e j e c t i o n of the n u l l hypothesis at Edmonton (b) i s caused by the omission o f the l a s t nine-paired observations from the record. In Table VI, seven rather than nine paired observations are omitted at Edmonton (b) because the i n i t i a l step i n the a p p l i c a t i o n o f the Wilcoxon matched-pairs  signed-ranks test i s t o drop a l l paired ob-  servations with equal scores ( i . e . dates) from the analysis.  Hence,  Table VI indicates the number of p a i r s subjected t o the t e s t and not the number of pairs of corresponding observations i n the records at each l o cation (see Table I I I ) . At Brandon, a change i n the d e f i n i t i o n of break-up from  initial  27. movement of i c e ( i ) to complete clearance of i c e ( i i ) r e s u l t s i n r e j e c t ion of the n u l l hypothesis.  This suggests that dates of l a s t - i c e a f f e c t -  ing discharge at t h i s s t a t i o n are more closely related to f i r s t dates of break-up than to second dates of t h i s phenomenon.  Mean break-up dates  corroborate these test r e s u l t s (Table IV). Test r e s u l t s of records at Saskatoon and Edmonton are cases and i t i s best to regard them with some reservations.  border-line This p a r t i c u -  l a r l y applies i n t e s t s which involve l i m i t e d numbers of paired observations and where those observations may  incorporate errors of one kind  or  another. It i s obvious from t e s t i n g the d i s p a r i t y between l a s t - i c e and break-up records that they d i f f e r s i g n i f i c a n t l y at some locations. However, i t i s also apparent from the computation of mean differences TV) that the time d i s p a r i t y i s r e l a t i v e l y i n s i g n i f i c a n t .  The  (Table  discrepancies  between paired observations at Headingley, Winnipeg, and S e l k i r k were not subjected to Wilcoxon's test because of the distances separating the observation s i t e s . The  degree of relationship between break-up and l a s t - i c e at each  l o c a t i o n was measured by the product-moment c o r r e l a t i o n c o e f f i c i e n t .  In-  dividual values of " r " and a pooled estimate of " r " f o r the locations i n the Saskatchewan t r i b u t a r y basin are l i s t e d i n Table VII. timate was The  derived by using Fisher's  M  The pooled es-  z " transformation.  " r " values indicate that a stochastic relationship  exists  between break-up and l a s t - i c e i n the Saskatchewan t r i b u t a r y basin. measure of positive c o r r e l a t i o n i s highly s i g n i f i c a n t when i t i s  The  28. TABLE VII VALUES OF SAMPLE CORRELATION COEFFICIENTS, BREAK-UP AND LAST-ICE (Saskatchewan Tributary Basin)  Location  r  r  »05  r ,  01  Significance  Edmonton Prince Albert Saskatoon The Pas  .574 .905 .737 .929  .324 .314 .497 .433  .418 .403 .623 .549  Sig. Sig. Sig. Sig.  Pooled Estimate  .825  .190  .250  S i g . at 1 per cent l e v e l  at at at at  1 1 1 1  per per per per  cent cent cent cent  level level level level  considered that break-up dates recorded by various individuals and organizations at one l o c a t i o n have d i f f e r e d i n some years by as much as two weeks (Mackay, 1961b).  The general index of predictive e f f i c i e n c y E, i s  43.5 per cent although individual values of the index run from 18.1 cent at Edmonton to 63 per cent at The Pas.  per  These indices show that the  problem of estimating the value of one phenomenon from a given value of the other phenomenon i s a d i f f i c u l t  one.  In the Assiniboine-Red t r i b u t a r y basin, Brandon i s the sole l o c a t i o n with records of break-up and l a s t - i c e .  Results at Brandon are  s i m i l a r to those obtained i n the Saskatchewan basin. l a s t - i c e dates with f i r s t dates of break-up (r = dates with second dates of break-up (r = .826) one per cent l e v e l .  Correlations of  .792), and l a s t - i c e  are s i g n i f i c a n t at the  The indices of predictive e f f i c i e n c y are 39 per  cent and 44 per cent, respectively. Correlations between records of i c e d i s i n t e g r a t i o n from Headingley, Winnipeg, and Selkirk i n the Assiniboine-Eed t r i b u t a r y basin are placed i n matrix form (Table V I I I ) .  29. TABLE VIII MATRIX OF SAMPLE CORRELATION COEFFICIENTS >  Headingley Last-ice  Winnipeg Break-up  Selkirk Break-up  Headingley Last-ice Winnipeg Break-up  .469  Selkirk Break-up  .718  .341  The r e l a t i v e l y high measure of agreement between l a s t - i c e at Headingley and break-up at S e l k i r k as compared to the l e s s tenuous r e l a t i o n s h i p apparently e x i s t i n g between break-up at Winnipeg and  compara-  ble phenomena at the other l o c a t i o n s indicates that break-up at Winnipeg i s somewhat of an anomaly.  This could be caused by differences i n break-  up c r i t e r i a , t r a n s c r i p t i o n errors, warming e f f e c t of sewage, and so on. Estimates of the regression of l a s t - i c e on break-up are noted on the scatter diagrams (Figures 3, 4, 5, 6 and 7) used to i l l u s t r a t e the relationship between the two phenomena at each location.  Break-up i s r e -  garded as the independent variable because t h i s event, i n most instances, occurs p r i o r to l a s t - i c e . In view of the l i m i t e d amount of data available f o r comparisons and the occurrence of occasional extreme differences between paired observations, the use of median values of differences seems most appropriate f o r purposes of adjustment, i f adjustment i s considered desirable or  30.  BREAK-UP  Days from Mar. 31st  14  18  22  26  30  34  BREAK-UP Days from Mar. 31st  BREAK-UP  38  42  Mar. 31st  Days from Mar. 31st 24  1V>  20 16  12  5 FIGURE 7  BREAK-UP AND LAST-ICE AT BRANDON ON THE ASSINIBOINE RIVER 1 Paired observation  -4  4  Days from Mar. 31st  12  8 12 16 20 BREAK-UP (Initial ice movement)  16  20  24  28  BREAK-UP (Complete ice clearance)  •  24  Days from " Mar. 31st  j  Days from  ^~*"Mar. 31st 32  33. TABLE IX MEANS AND MEAN DIFFERENCES BETWEEN DATES OF FREEZE-UP AND FIRST-ICE  Location & Period  Mean Dates Freeze-up F i r s t ;-ice  Mean Difference i n days  Omissions  Earlier Phenomena  Edmonton 1926-35  Nov. n  1936-45 1946-55  II  1915-55  «  2 4 3 3  F i r s t - ice  Oct. 2 9 Nov. 4  4  2  0  2  13  10  3  4  1  14  4  6  1  1  13  1  tt  ti  n it  n  it  it  tt  Freeze -up tt  M  Prince Albert 1916-25  Nov. 1 0  1926-35  n  14  1936-45  •i  1946-55  it  Nov. tt  F i r s t - ice  4  5  1  it  12  5  1  tt  Nov. 1 2  Nov. 1 5  3  1  9  15  6  1  8  1  6  1  9 17  n it  Brandon ( i ) 1946-55  ti  1943-55  tt  Freeze-up it  II  Brandon ( i i ) Nov. 23  1946-55  it  1943-55  necessary.  21  Nov. 1 5 II  15  F i r s t - ice tt  n  Median values o f differences are as follows: four days at  Edmonton, two days at Prince A l b e r t , two days at Saskatoon, two days at The Pas, and two days at Brandon.  The median value of the difference  between break-up and l a s t - i c e f o r the f i v e records combined i s two days. FREEZE-UP AND FIRST-ICE Table IX l i s t s means and mean differences between f i r s t - i c e and freeze-up f o r ten-year periods and the whole period of record at each location.  Freeze-up at Edmonton r e f e r s t o the i n i t i a l formation o f i c e ;  freeze-up at Prince Albert i s defined as the time the r i v e r was completel y frozen over at the observation s i t e .  Freeze-up dates at Brandon are  34. recorded f o r both the i n i t i a l formation o f i c e ( i ) and complete freezeover ( i i ) . At Edmonton the d e f i n i t i o n of freeze-up i n use i s such that t h i s event generally occurs before f i r s t - i c e a f f e c t i n g discharge i s recorded at the gauging s t a t i o n .  At Prince A l b e r t , r e s u l t s are considerably  d i f f e r e n t ; mean dates indicate that f i r s t - i c e i s observed seven days earl i e r than freeze-up as defined at that station.  At Brandon, dates of  f i r s t - i c e a f f e c t i n g discharge are generally l a t e r than f i r s t freeze-up dates but e a r l i e r than second freeze-up dates.  Differences i n d e f i n i -  tions of freeze-up make i t unreasonable to compute an o v e r - a l l mean f o r t h i s phenomenon, and, thus, to determine a grand mean difference between freeze-up and f i r s t - i c e . The Wilcoxon matched-pairs signed-ranks t e s t was used t o t e s t the magnitude and d i r e c t i o n o f the difference between freeze-up and first-ice.  This t e s t indicates that no difference exists between freeze-  up and f i r s t - i c e at Edmonton; at Prince Albert, however, dates of f i r s t ice are e a r l i e r than dates of freeze-up.  Besults at Brandon indicate  that there i s no difference between dates o f f i r s t - i c e a f f e c t i n g discharge and f i r s t freeze-up dates, but second freeze-up dates are s i g n i f i c a n t l y l a t e r than f i r s t - i c e a f f e c t i n g discharge.  I n t u i t i v e l y , i t would seem that  dates of f i r s t - i c e a f f e c t i n g discharge and dates o f freeze-up defined at the f i r s t formation o f i c e are more comparable than dates of complete freeze-over and f i r s t - i c e dates.  Results of Wilcoxon's t e s t add weight to  t h i s statement despite the great 'within v a r i a b i l i t y ' exhibited by both sets o f data at these locations.  35. Scattergrams i l l u s t r a t i n g the relationship between freeze-up and f i r s t - i c e at Edmonton, Prince A l b e r t , and Brandon are included i n the text as Figures 8, 9, and 10.  P l o t t e d values that deviate strongly from  the mean of either variable have been noted on the scattergrams according to the year of t h e i r occurrence.  At Edmonton (Figure 8) i t i s obvious  that a few values markedly lower the degree of p o s i t i v e correlation.  In  those years i n which a marked difference between the dates of freeze-up and f i r s t - i c e have been noted, i t i s possible that recording of e i t h e r phenomenon may be i n error.  However, the date of f i r s t - i c e a f f e c t i n g  discharge i n the year 1933 i s more than three standard deviations from the mean date of f i r s t - i c e f o r the period 1921 to 1950.  I t seems reasonable  to assume that t h i s date i s an observational o r t r a n s c r i p t i o n a l error.  At  Prince Albert (Figure 9} the scattergram indicates that a few large d i s crepancies between dates of f i r s t - i c e and freeze-up have again greatly r e duced the measure of p o s i t i v e c o r r e l a t i o n .  The plotted value f o r the year  1947 i s 3.4 standard deviations from the mean date o f f i r s t - i c e and may be assumed t o be i n error. Some of the other years that are noted, may also incorporate errors of one kind o r another i n the recorded dates of one or both phenomena.  At Brandon (Figure 10), the d i s p a r i t y between f i r s t - i c e  and freeze-up dates i n 1944 has greatly reduced the measure of agreement between the two sets of data. The product-moment c o r r e l a t i o n c o e f f i c i e n t was used to measure the relationship between freeze-up and f i r s t - i c e at Edmonton, Prince A l b e r t , and Brandon.  The measure of c o r r e l a t i o n between second dates of freeze-up  and dates of f i r s t - i c e a f f e c t i n g discharge at Brandon i s the sole value of  36  Days from Oct. 3 1 s t  FIGURE 8  FREEZE-UP AND FIRST-ICE AT EDMONTON 1 Paired observation 2 Paired observations  • •  1933 Omitted in regression computation  -20  -16  -12  - 8  8  12  16  20  24  FIRST-ICE  28  32  Days from Oct. 3 1 s t  Days from Oct. 3 1 s t 42  • 1912  38  34  30  26  |  22 • 1949  J  18  h  "1930  14  10  • 1947  h  FREEZE-UP AND FIRST-ICE AT PRINCE ALBERT 1 Paired observation 2 Paired observations  -18  -14  -10  10 FIRST-ICE  14  26  30  34  • •  Days from Oct. 3 1 s t  37.  FREEZE-UP (First appearance of ice)  -4  0  4  8  12  16  20  24  FREEZE-UP(Complete ice coverage)  28  32  36  38. TABLE X VALUES OF SAMPLE CORRELATION COEFFICIENTS, FREEZE-UP AND FIRST-ICE Location  n  Edmonton  27  .547  Prince Albert  35  Brandon ( F i r s t dates) Brandon (Second dates)  r  01  Significance  .367  .470  Sig. at 1 per cent l e v e l  .416  .325  .418  Sig. at 5 per cent l e v e l  12  .616  .532  .661  Sig. at 5 per cent l e v e l  12  .427  .532  .661  Not s i g n i f i c a n t  r  '05  r ,  a l l those obtained i n correlating break-up with l a s t - i c e dates and freezeup with f i r s t - i c e dates that i s not s i g n i f i c a n t at the f i v e per cent l e v e l . However, i f the year 1944 i s omitted from the period of record, the coe f f i c i e n t of c o r r e l a t i o n level.  (r = .710}  i s s i g n i f i c a n t at the one per cent  The computed value of the c o r r e l a t i o n  c o e f f i c i e n t between f i r s t  dates of freeze-up and dates of f i r s t - i c e a f f e c t i n g discharge (r =  .850)  also becomes s i g n i f i c a n t at the one per cent l e v e l . In the Assiniboine-Red tributary basin, no s p e c i f i c l o c a t i o n other than Brandon has records of both freeze-up as well as f i r s t - i c e  affecting  discharge that cover periods of s u f f i c i e n t length to j u s t i f y measuring the degree of relationship  between the two phenomena.  In order to i l l u s t r a t e  that there i s some measure of agreement between dates of freeze-up and f i r s t - i c e at other locations,  the record of f i r s t - i c e at Headingley i s com-  pared with that of freeze-up at Winnipeg.  The sample value (r =  .565) of  the c o e f f i c i e n t of c o r r e l a t i o n i s s i g n i f i c a n t at the one per cent l e v e l .  39. The comparisons of data show that differences between dates of freeze-up and f i r s t - i c e are considerably l a r g e r than those between dates of break-up and l a s t - i c e .  A l l four ten-year mean differences at Prince  Albert, at l e a s t one ten-year mean difference at Edmonton, and one at Brandon support such a conclusion.  However, the differences i n freeze-up  d e f i n i t i o n s may magnify the discrepancies between freeze-up and f i r s t - i c e to a greater extent than those i n break-up d e f i n i t i o n s enlarge the d i s c r e pancies between break-up and l a s t - i c e . In the majority of cases, the 'within v a r i a b i l i t y ' i n freeze-up and f i r s t - i c e records exceeds the 'within v a r i a b i l i t y ' i n break-up and l a s t - i c e records.  The standard deviations of f i r s t - i c e and l a s t - i c e sup-  port these findings (Figure 16 and Figure 22).  There i s also a greater  •between v a r i a b i l i t y ' which i s r e f l e c t e d i n the lower measures of p o s i t i v e c o r r e l a t i o n between freeze-up and f i r s t - i c e . The s t a t i s t i c s show that there i s a high degree of c o v a r i a b i l i t y between hydrologie records of i c e conditions a f f e c t i n g discharge and h i s t o r i c a l records of freeze-up and break-up.  Moreover, the time d i s p a r i t y  between paired sets of these data i s limited. hydrologie data may  These f a c t s suggest that  be used t o analyze the areal v a r i a b i l i t y and rate of  progress of- r i v e r ice formation and d i s i n t e g r a t i o n i n the Nelson drainage system.  CHAPTER I I I  PROGRESS OF RIVER ICE FORMATION AND DISINTEGRATION  The progress o f r i v e r ice formation and d i s i n t e g r a t i o n i s studied f i r s t by means o f maps showing the areal variations and second by s t a t i s t i c a l techniques.  A basic ten-year period i s used both f o r maps as w e l l  as f o r test purposes  (Wilcoxon matched-pairs  signed ranks t e s t ) .  In some  instances, i t was necessary t o adjust the length of the period to compensate f o r omissions i n s t a t i o n records and t o maintain the number o f observations at ten f o r each test. A f i v e per cent s i g n i f i c a n c e l e v e l i s used as c r i t e r i o n f o r acceptance or r e j e c t i o n of the n u l l hypothesis of no difference between i c e dates.  I f acceptance of the n u l l hypothesis was caused by an extreme  difference between one p a i r o f observations, a further year of record was compared t o substantiate or negate the previous test r e s u l t s .  MAPS Mean-first-ice and l a s t - i c e dates are plotted f o r stations recording four o r l e s s omissions i n each ten-year period under consideration. Those stations with two t o four omissions i n each ten-year period are used as reference points only when no other information i s available i n the immediate.area.  In areas where clustered mean dates d i f f e r i n value, the  most representative ( i . e . modal o r mean) value i s used f o r p l o t t i n g purposes.  Isopleths of mean dates f o r ten-year periods are shown on Figures  41.  11, 12, 13, and 14 f o r f i r s t - i c e conditions, and on Figures 17, 18, 19, and 20 f o r l a s t - i c e conditions. Means (Figures 15 and 21) and standard deviations (Figures 16 and 22) of f i r s t - i c e and l a s t - i c e f o r the period 1921 to 1950 are mapped and included i n t h i s t h e s i s .  A map of the mean length o f the ice-free season  (Figure 23) i s also included f o r the period 1921 t o 1950. climatic observations  Averages of  based on t h i s period are regarded as "normals"  (Meteorological Branch, Department of Transport, Canada, CIR-3208 CLI-19, 3 Jun. 59, p. 11). S t a t i o n records covering the period 1921 to 1950 are l i m i t e d i n number; 26 stations have l a s t - i c e records with fewer than s i x omissions, and only 22 of these have f i r s t - i c e records with the same c h a r a c t e r i s t i c . As no obvious progression o f standard deviations, and mean dates of f i r s t ice f o r the period 1921 to 1950 i s apparent within the Nelson drainage system, dot i n t e n s i t i e s are used t o overcome the spurious impression of abrupt t r a n s i t i o n that may be created through the use of cross-hatching. Isopleths of mean f i r s t - i c e dates f o r ten-year periods  (Figures  11, 12, 13 and 14) bow south i n the west central area of the Nelson drainage system with some moderation o f t h i s tendency apparent i n the period 1946  to 1955 (Figure 14).  The period 1946 to 1955 also indicates a marked  s h i f t t o l a t e r dates of i c e formation throughout most areas of the Nelson drainage system.  Variations exhibited by the isopleths of f i r s t - i c e i n  the ten-year periods are modified by the process of averaging to produce a l e s s complicated  pattern f o r the period 1921 to 1950 (Figure 15).  In contrast t o isopleths of mean f i r s t - i c e dates f o r ten-year  42.  periods, those of l a s t - i c e dates (Figures 17, 18, 19, and 20) bow north i n the west central part of the Nelson drainage system.  This tendency  i s modified to some degree when the o v e r - a l l period, 1921 to 1950 (Figure 21), i s considered.  A v i s u a l comparison of isopleths of mean l a s t - i c e  dates, 1921 to 1950, with isotherms of mean monthly temperatures f o r March, May (Currie, 1954), and A p r i l (Kendrew and Gurrie, 1955) suggests that dates of l a s t - i c e and a i r temperatures may be correlated.  The o r i -  entation of isopleths o f mean l a s t - i c e and isotherms o f mean monthly temperatures f o r spring months i s s u b s t a n t i a l l y the same, northwest t o southeast. The mean length of i c e - f r e e season f o r the period 1921 to 1950 (Figure 23) varies from 232 days i n the southwest corner of the Nelson drainage basin to 196 days i n the northeast sector between Prince ALbert and The Pas.  Rates of change i n the mean length of i c e - f r e e season d i f f e r  substantially from area to area.  A rapid decrease i n length occurs from  west t o east i n the t r a n s i t i o n a l zone between the Rocky Mountains and the Great Plains physiographic regions.  A minimal rate of decrease occurs  northward from the International Boundary to the upper reaches of the Assiniboine River in the central section of the Nelson drainage system. PROGRESS OF LAST-ICE The expected progress of i c e cover disintegration i n northwardflowing r i v e r s suggests that open-water areas w i l l i n i t i a l l y appear i n the upper reaches and spread downstream u n t i l the r i v e r becomes ice-free. Destruction of the i c e cover most l i k e l y occurs i n downstream surges  43.  rather than by any continuous process of d i s i n t e g r a t i o n (Burbridge and Lauder, 1957).  Snow melt and spring rains swell the r i v e r i n i t s upper  reaches and increase the pressure beneath the i c e surface, f o r c i n g the ice to r i s e , a f t e r which cracks appear, and the i c e cover breaks up. The i c e blocks then move downriver u n t i l jamming requires the process to repeat i t s e l f . An examination of the t r i b u t a r i e s of the Nelson River should show the effect of r i v e r o r i e n t a t i o n on the progress of l a s t - i c e .  The North  Saskatchewan River i s predominantly a west-to-east flowing r i v e r throughout i t s entire length; the South Saskatchewan flows northeasterly to i t s junction with the North Saskatchewan, 30 miles past Prince ALbert; the Assiniboine flows southeasterly u n t i l i t crosses the Manitoba-Saskatchewan border where i t turns southward as f a r as the Virden area and then eastward to i t s junction with the Red River at Winnipeg.  A major westward-  flowing t r i b u t a r y , the Winnipeg River, cannot be examined f o r l a s t - i c e progress because the data available are inadequate f o r t h i s purpose. Progress of l a s t - i c e along the North Saskatchewan, the South Saskatchewan, and the Assiniboine Rivers i s l i s t e d i n TableXI.  Locations of gauging  s i t e s used to chart the progress of l a s t - i c e i s shown i n Figure 24. The d i s i n t e g r a t i o n of i c e a f f e c t i n g discharge on the North Saskatchewan progresses downstream from i t s head-waters to i t s junction with the South Saskatchewan.  The fact that no s i g n i f i c a n t difference  exists between dates of l a s t - i c e at Rocky Mountain House and Edmonton and, f a r t h e r downstream, between dates of the same phenomenon at Frenchman's Butte and Prince Albert tends to substantiate the b e l i e f that d i s i n t e g r a t i o n of the i c e cover occurs i n sections.  o  CI CJ1  57.  TABLE XI LAST-ICE PROGRESS ALONG MAJOR TRIBUTARIES OF THE NELSON RIVER NOTEs I f 'T* >8,  the difference between l a s t - i c e dates i s not  I f 'T'< 8, the difference i s s i g n i f i c a n t .  significant.  See Figure 24 f o r station locations.  A. North Saskatchewan River; base period, 1944-55 Stations Compared  VTf  Significance  Rocky Mountain House-Edmonton  13  Not s i g . at 5 per cent l e v e l  Edmonton-Frenchman's Butte  0  Sig. at 1 per cent l e v e l  Frenchman's Butte-Prince Albert  13  Not s i g . at 5 per cent l e v e l  Prince Albert - The Pas  Stations Compared  0  IIJII  Cowley-MacLeod  0  MacLeod-Lethbridge  22  Sig. at 1 per cent l e v e l  Significance Sig. at 1 per cent l e v e l Not sig., at 5 per cent l e v e l  5  S i g . at 5 per cent l e v e l  Medicine Hat-Saskatoon  0  Sig. at 1 per cent l e v e l  Saskatoon-The Pas  0  Sig. at 1 per cent l e v e l  Lethbridge-Medicine  Hat  C. Headwater Stations on Easterly-Flowing Streams. Base periods: St.Mary River, 1945-55; Bow River, 1937-46; Athabasca River, 1919-31, Significance  Stations Compared ( i ) St.Mary River International BoundaryLethbridge  0  S i g . at 1 per cent l e v e l  ( i i ) Bow River Banff-Calgary  3  S i g . at 1 per cent l e v e l  ( i i i ) Athabasca River Jasper-Entrance Entrance-Athabasca  10 0  Not s i g . at 5 per cent l e v e l S i g . at 1 per cent l e v e l  58. (TABLE XI continued) D» Assiniboine River; base period, 1944-55 Stations Compared  |T^  Sturgis-Kamsack  19  Not s i g . at 5 per cent l e v e l  Kamsack-Millwood  20  Not s i g . at 5 per cent l e v e l  Millwood-Brandon  4  Brandon-Headingley  15  E. M i s s i s s i p p i Drainage Basin; Stations Compared International BoundaryMilk River Milk River-Eastern Crossing  Significance  Sig. at 5 per cent l e v e l Not s i g . at 5 per cent l e v e l  Milk River; base period. 1946-55 VT* 15 2  Significance Not s i g . at 5 per cent l e v e l S i g . at 1 per cent l e v e l  D i s i n t e g r a t i o n of the i c e cover on the South Saskatchewan River begins on i t s head-water t r i b u t a r i e s i n the southwestern f o o t h i l l s of Alberta.  E a r l i e s t dates of l a s t - i c e occur at the MacLeod and Lethbridge  gauging stations on the Oldman River and at a gauging s i t e on the St. Mary River near Lethbridge.  Ice d i s i n t e g r a t i o n progresses along these  t r i b u t a r i e s and thence down the South Saskatchewan as the r i v e r runs i t s north-easterly course across the Great Plains.  Break-up of the i c e  cover continues i n a downstream d i r e c t i o n past the junction of the North and South Saskatchewan, past the gauging s t a t i o n on the main Saskatchewan at The Pas, and on t o the r i v e r ' s outlet i n Lake Mnnipeg. A feature of break-up on the upper reaches of two of the South Saskatchewan's headwater t r i b u t a r i e s which i s a t y p i c a l of other easterly-  60. flowing streams r i s i n g i n the Rockies, i s the lateness of break-up dates. Break-up at Cowley on the Oldman River and at the International Boundary gauging s i t e on the St. Mary River i s l a t e r than i t s occurrence at the downstream stations of MacLeod and Lethbridge.  The s i t u a t i o n i s reversed  on the Bow t r i b u t a r y to the north; dates of l a s t - i c e at the Banff gauging s t a t i o n are e a r l i e r than those at Calgary some distance downstream. Farther north, on the North Saskatchewan, dates of l a s t - i c e at Rocky Mountain House exhibit no s i g n i f i c a n t differences from dates downriver at Edmonton.  S i m i l a r l y , on the Athabasca River s t i l l farther north, no s i -  gnificant difference e x i s t s between Jasper and Entrance records although l a s t - i c e dates at Entrance are e a r l i e r than those downstream at Athabasca. Thus, where records o f gauging stations along easterly-flowing streams i n the Rockies and t h e i r f o o t h i l l s can be tested f o r differences, dates of l a s t - i c e are e a r l i e r at upstream s i t e s on two occasions, e a r l i e r at downstream s i t e s on two occasions, and exhibit no difference between upstream and downstream s i t e s on two occasions. The r i v e r s on which i c e cover d i s i n t e g r a t i o n occurs e a r l i e r at downstream gauging s i t e s r i s e i n the southwestern corner of the Nelson drainage basin, an area i n which FBehn o r Chinook winds may have a considerable e f f e c t upon spring run-off.  These warm dry winds are heated by  compression as they descend the l e e side of the Rockies.  In the spring  they reach t h e i r frequency peak and can cause snow and i c e t o disappear very r a p i d l y .  ALthough some i c e and snow i s undoubtedly removed through  the process of sublimation, some run-off w i l l occur and the melt-water may, i n some instances, be s u f f i c i e n t to break open l o c a l r i v e r s and streams.  61. As the Chinook winds are heated on descent at the dry adiabatic lapse rate of 5.5°F per 1,000  f e e t , differences of a few hundred feet  i n the elevation of gauging sites can produce a change from freezing cond i t i o n s at an upstream s i t e to thawing conditions at a downstream one. I f these conditions are sustained f o r a number of hours and, i f the Chinooks have a high frequency of occurrence i n a given period of time, the cumulative effect may r e s u l t i n break-up occurring i n i t i a l l y  on the  lower reaches of streams. In the southern f o o t h i l l s of Alberta, the elevation of the Oldman IfcLver at Cowley i s approximately 1,000  feet above i t s elevation at  Lethbridge; the St. Mary River i s some 1,500  feet higher at the  International Boundary gauging s i t e than at the Lethbridge s i t e .  Differ-  ences between the elevations of upstream and downstream gauging s i t e s of t h i s size could have a considerable influence upon the i n i t i a l point of break-up. On the Canadian side of the border, the strength and frequency of Chinooks generally decrease northwards.  Calgary i s l e s s affected by these  winds than Lethbridge, Edmonton l e s s than Calgary, and so on.  I f Chinooks  a f f e c t r i v e r i c e d i s i n t e g r a t i o n , the r e s u l t s should be most obvious i n the southern f o o t h i l l s area.  Figures 17 to 23 i n c l u s i v e support t h i s statement.  The Assiniboine River which flows i n a southeasterly d i r e c t i o n f o r more than h a l f i t s length was tested to determine whether ice cover d i s i n tegration progresses downstream i n a manner s i m i l a r to the progress of break-up on the North and South Saskatchewan Rivers.  Five gauging s i t e s  62. with records suitable f o r comparisons of l a s t - i c e dates are situated on the Assiniboine at f a i r l y even i n t e r v a l s along i t s length.  Results of  Wilcoxon's test indicate that ice cover d i s i n t e g r a t i o n progresses downstream from Millwood i n Manitoba, located near the r i v e r ' s middle reaches, to Headingley, 12 miles west of the junction of the Assiniboine and Red Rivers i n Winnipeg.  From Millwood to Sturgis i n Saskatchewan, an approx-  imate distance of 150 miles up-river, there i s no s i g n i f i c a n t difference i n l a s t - i c e dates.  Similar r e s u l t s obtained from t e s t i n g records f o r the  North and South Saskatchewan Rivers make i t reasonable to assume that lack of differences between records can be interpreted as meaning that break-up of the i c e cover i s not a continuous process but, rather, a spasmodic one with varied-size sections breaking up at i r r e g u l a r i n t e r vals.  The sections do not necessarily break-up i n the same order on  each t r i b u t a r y examined, but generally speaking, the progress of i c e cover d i s i n t e g r a t i o n begins on a r i v e r ' s upper reaches and proceeds downstream towards i t s mouth.  At l e a s t , t h i s i s the case i f records from gauging  s i t e s along the North Saskatchewan, South Saskatchewan, and Assiniboine Rivers are an i n d i c a t i o n of break-up on other r i v e r s i n the Nelson drainage  system. Three s e r i e s of l a s t - i c e records from stations that are l a t i t u d i -  n a l l y separated along north-south bands i n the west, c e n t r a l , and east sections of the Nelson drainage system were tested f o r differences. Figure 24 i l l u s t r a t e s the l o c a t i o n o f each s t a t i o n and Table XII l i s t s the results of Wilcoxon matched-pairs signed-ranks t e s t of t h e differences. The base period used f o r t e s t i n g purposes varied with each s e c t i o n d u e t o  63. the omission of some l a s t - i c e dates from a number of s t a t i o n records. The l a t e s t ten-paired  dates from records are compared i n each t e s t .  Ice conditions are expected to remain longer at northern gauging s i t e s than at southern s i t e s during the spring break-up. mean l a s t - i c e dates indicate that t h i s i s the case.  Isopleths of  Furthermore, s t a t i s -  t i c a l t e s t s of differences tend to substantiate the general  south-north  pattern of l a s t - i c e progress. In the western section, the record of l a s t - i c e at Lethbridge i s e a r l i e r than other l a s t - i c e records i n t h i s section.  Each of the other  s t a t i o n records from the western section that was tested was same as, or e a r l i e r than i t s nearest northern neighbour.  e i t h e r the  In the central  section, records of l a s t - i c e exhibit differences between a l l stations subjected to Wilcoxon's t e s t .  The differences substantiate the b e l i e f that  ice conditions tend to p e r s i s t f o r a longer period i n spring at the s t a t i o n with the more northerly l o c a t i o n .  Stations i n the east section follow the  same general pattern established by those i n the c e n t r a l section with one exception, the two  southern most stations, Wawanesa and Tantallon, have  records that exhibit no s i g n i f i c a n t difference between them.  The same re-  s u l t s were obtained from two t e s t s between stations i n the west section. It may  be assumed that either the removal of ice conditions can be expected  to occur concurrently at stations where no s t a t i s t i c a l l y s i g n i f i c a n t d i f f e r ences exist or that the acceptance of the n u l l hypothesis of no difference i s i n error.  I f differences are not s t a t i s t i c a l l y s i g n i f i c a n t , i t does not  mean that r e a l differences between dates of l a s t - i c e are non-existent;  it  means, rather, that on the basis of Wilcoxon's t e s t , r e a l differences are  64. not meaningful from a s t a t i s t i c a l point of view. The rate of disappearance of i c e conditions a f f e c t i n g discharge varies i n d i f f e r e n t areas of the Nelson drainage system.  The d i s i n t e g r a -  t i o n of i c e conditions a f f e c t i n g discharge, 1921 to 1950,  progresses  northward at an average rate of 16 miles per day i n the west and central sections, and at an average rate of 22 miles per day i n the eastern section.  The average rate of progress measured at right angles to i s o p l e t h  orientation i n the central section i s approximately 17 miles per day. The minimal average rate of l a s t - i c e progress, four to five miles per day, occurs west to east i n the Rocky Mountain and f o o t h i l l s region between Banff and Calgary.  PROGRESS OF FIRST-ICE Records of f i r s t - i c e dates from four l o c a t i o n s along the North Saskatchewan-Saskatchewan River system were tested and the r e s u l t s i n f e r that no differences exist between them.  On the South Saskatchewan River,  three widely-separated locations exhibit no differences between t h e i r records.  However, Saskatoon,does display s i g n i f i c a n t l y l a t e r dates of  f i r s t - i c e than The Pas, on the main Saskatchewan.  Three s t a t i o n records  on the Assiniboine River show no s i g n i f i c a n t differences between dates of f i r s t - i c e formation (See Table X I I I ) . The procedure used to check the north-to-south progress of i c e formation i s the same one that was used to examine the progress of ice d i s i n t e g r a t i o n i n the opposite d i r e c t i o n .  The Nelson drainage system  was again a r b i t r a r i l y divided into a western, a c e n t r a l , and an eastern section, and t e s t s of differences were again conducted between  65.  TABLE XII TEST OF DIFFERENCES BETWEEN RECORDS OF LAST-ICE DATES AT LATITUDINALLY-SEPARATED STATIONS NOTE* I f *T*>8, the difference between l a s t - i c e dates i s not s i g n i f i c a n t . I f 'T' <8, the difference i s s i g n i f i c a n t .  See Figure 24 f o r s t a t i o n locations.  *T' VALUES A. Western Sections base period, 1936-45 St.Mary R. at International Boundary  Oldman R. at Lethbridge  Bow R. at Calgary  Red Deer R. at Red Deer  North Sask. at Edmonton  St.Mary R. at International Boundary Oldman R. at Lethbridge  2  Bow R. at Calgary  8  2.5  Red Deer R. at Red Deer  6  0  North Saskatchewan R. at Edmonton  19.  15.5  B. Central Sections base period. 1924-40 Frenchman R. at International Boundary Frenchman R. at International Boundary Swift Current Cr. at Swift Current S. Saskatchewan R. at Saskatoon N. Saskatchewan R. at Prince Albert  Swift Current Cr.at Swift Current  S.Sask.R. at Saskatoon  N.Sask.R. at Prince Albert  66.  (TABLE XII  continued)  C. East Section:  base period, 1927-55 Souris R. at Wawanesa  Q,u*Appelle R. at.Tantallon  Swan R. at Swan River  Saskatchewan R. at The Pas  Souris R. at Wawanesa Qu'Appelle R. at Tantallon Swan R. at Swan River  13.5 4  5  Saskatchewan R. at The Pas  G  l a t i t u d i n a l l y - s e p a r a t e d stations.  O  O  Unfortunately, locations that have recorded  f i r s t - i c e dates are few i n number.  Eor example, the central section i s l i m i t e d  to a comparison between records of f i r s t - i c e at Saskatoon and P r i n c e Albert. The r e s u l t s of the t e s t s are l i s t e d  i n Table XIV.  On the basis of available data, the western s e c t i o n may be divided, according to t e s t s of the time of i n i t i a l i c e formation a f f e c t i n g discharge, into two l a t i t u d i n a l segments with the demarcation point f a l l i n g between Calgary and Red Deer.  Tests show that no difference exists either between r e -  cords from Lethbridge and Calgary o r between those from Red Deer and Edmonton. I f the s t a t i o n records are assumed to be representative ones, then there i s some basis f o r making the assumption that r e l a t i v e homogeneity i n terms of the time of i c e formation exists i n the area between Lethbridge and Calgary, and i n that between Red Deer and Edmonton.  Demarcation may thus be envisaged as a .zone  of indeterminate width separating areas that exhibit differences i n the time of i n i t i a l i c e formation a f f e c t i n g discharge.  67. TABLE XIII FIRST-ICE PROGRESS ALONG MAJOR TRIBUTARIES OF THE NELSON RIVER NOTE: I f *T >8, the difference between f i r s t - i c e dates i s not s i g n i f i c a n t . T  If 'T' <8, the difference i s s i g n i f i c a n t .  See Figure 24 f o r station locations.  A. North Saskatchewan River: base period, 1921-54 Stations Compared  VP  Significance  Rocky Mountain House-Edmonton  12.5  Not s i g . at 5 per cent l e v e l  Edmonton-Prince Albert  18.  Not s i g . at 5 per cent l e v e l  Prince Albert-The Pas  24.5  Not s i g . at 5 per cent l e v e l  B. South Saskatchewan River; base period, 1941-54 Stations Compared  *_T^  Significance  Lethbridge-Medicine Hat  21.  Not s i g . at 5 per cent l e v e l  Medicine Hat-Saskatoon  20.5  Not s i g . at 5 per cent l e v e l  Saskatoon-The Pas  4.5  S i g . at 5 per cent l e v e l  C. Assiniboine River; base period, 1925-54 Stations Compared  JJTJ  Significance  Millwood-Brandon  13.5  Not s i g . at 5 per cent l e v e l  Brandon-Headingley  25.  Not s i g . at 5 per cent l e v e l  68. TABLE XIV TEST OF DIFFERENCES BETWEEN RECORDS OF FIRST-ICE DATES AT LATITUDINALLY-SEPARATED STATIONS NOTE: I f 'T'> 8, the difference between f i r s t - i c e dates i s not s i g n i f i c a n t . I f »T«< 8, the difference i s s i g n i f i c a n t .  See Figure 24 f o r s t a t i o n locations.  'T' VALUES A. Western Section; base period, 1936-45 St. Mary R. at Lethbridge  Bow R. at Calgary  Red Deer R. at Red Deer  N.Sask.R. at Edmonton  St. Mary R. at Lethbridge Bow R. at Calgary  21.5  Red Deer R. at Red Deer  6  N. Saskatchewan R. at Edmonton  3*5  14.5  B. Central Section; base period. 1936-45 A t e s t of the two stations with available records, Saskatoon and Prince Albert, suggests that no s i g n i f i c a n t difference exists between them.  The *T* value i s 14.5  C. Eastern Section; base period, 1921-31 Souris R. at Assiniboine R. Swan R. at Sask. R. Wawanesa at Millwood Swan River at The Pas Souris R. at Wawanesa Assiniboine R. at Millwood Swan R. at Swan River Saskatchewan R. at The Pas  23-. 5 12.5  10  5  8  11  69.  The central section has no records suitable f o r comparison purposes south of the Saskatoon gauging station.  Saskatoon's record was t e s t -  ed against Prince Albert's and the result showed that there was no difference between them. In the eastern s e c t i o n , the e f f e c t s of i c e upon discharge are f i r s t noticeable at the most northerly s t a t i o n , The Paa.  To the south, r i v e r s  generally remain ice-free u n t i l l a t e r i n the f a l l season.  At f i r s t glance,  test results do not necessarily convey t h i s impression; the apparent  incon-  g r u i t y of Swan River's agreement with the other stations on the time of f i r s t - i c e development, and The Pas* disagreement with stations south o f Swan River, confuse the issue.  Tests show, however, that s t a t i s t i c a l d i f -  ferences between Swan River's record and the others are border-line cases. In each of these cases, a decrease i n the magnitude o f the difference between one p a i r of observations would have resulted i n a r e j e c t i o n of the n u l l hypothesis of no difference.  Mean values f o r the period of the t e s t  bear out the b e l i e f that ice formation at Swan River occurs l a t e r than the formation of the same phenomenon at The Pas, and e a r l i e r than i t s formation at Millwood.  E s s e n t i a l l y , the t e s t r e s u l t s support the isopleth map of  mean f i r s t - i c e dates, 1946 to 1955. Southward progress of i c e formation a f f e c t i n g discharge based upon mean dates of f i r s t - i c e , 1921 to 1950, occurs at a rate of 20 to 25 miles per day i n the western and central sections, and at a rate of approximately 100 miles per day i n the eastern section.  The minimal rate of f i r s t - i c e  progress, 15 t o 20 miles per day, i s west to east i n the t r a n s i t i o n a l zone between the Great Plains and Rocky Mountain physiographic regions.  70. In summary, ice i n i t i a l l y appears on the r i v e r s and streams i n the north central part of the Nelson basin and spreads r a d i a l l y west, south, and east.  The progress of freeze-up i s primarily c o n t r o l l e d by  l a t i t u d e and continentality although other factors may formation dates i n some areas.  influence ice  Spring break-up begins on the headwater  t r i b u t a r i e s of the South Saskatchewan River i n southwestern Alberta. The general trend of r i v e r ice d i s i n t e g r a t i o n i s eastwards, past the Cypress H i l l s , and then northeastwards towards the mouth of Nelson River i n Hudson Bay.  Break-up normally progresses downstream from a r i v e r ' s  headwaters t o i t s mouth although some v a r i a t i o n from t h i s pattern may expected.  be  For example, chinooks may upset the pattern on streams flowing  through the Rocky Mountain F o o t h i l l s .  As the progress of r i v e r i c e f o r -  mation and d i s i n t e g r a t i o n varies from year to year, an examination of trends i n i c e dates w i l l contribute to the broader understanding of these phenomena i n the Nelson drainage system.  CHAPTER IV  RIVER ICE FORMATION AND DISINTEGRATION TRENDS  There i s ample evidence of climatic change i n the northern hemisphere during the l a s t f i f t y to one hundred years.  A i r temperature records  from a number of regions indicate a warming trend over the f i r s t three of four decades of t h i s century.  For example, Longley (1953) and Crowe (1958)  show that these trends are evident i n a i r temperature records from Western Canada.  In the Canadian P r a i r i e s , Longley indicates that the r i s i n g trend  i s more marked i n Saskatchewan than i n Alberta or Manitoba during t h i s period.  The extent to which trends i n the length of i c e - f r e e season r e -  f l e c t those i n mean annual a i r temperatures  i s examined i n t h i s t h e s i s .  One of the factors increasing the d i f f i c u l t y of determining rel a t i o n s between trends i n a i r temperatures  and r i v e r i c e conditions i s  the imposition of man-made or a r t i f i c i a l disturbances upon the natural variations i n climate.  The majority of meteorological stations with  long-term records are located i n urban areas or i n close proximity to them.  Urban growth, which may  affect the exposure and environment of the  instruments at meteorological stations, may ations of a i r temperatures conditions.  induce trends i n the observ-  which may not be r e f l e c t e d i n those of i c e  In addition, there may be errors i n c a l i b r a t i o n , recording  and t r a n s c r i p t i o n , which add to the d i f f i c u l t i e s of assessing relations between a i r temperature and ice condition trends.  72. TRENDS IN HYDROLOGIC DATA  Records of i c e conditions a f f e c t i n g discharge cover l i m i t e d time periods.  The longest hydrologie records available f o r t h i s study span a  range of 48 years.  None of these records i s complete; a l l have a number  of observations missing.  The more complete records are examined and com-  pared f o r s i m i l a r i t i e s and differences of movement or change with respect to time. Five-point f i l t e r e d series of representative s t a t i o n records from major t r i b u t a r y basins in_ the Nelson drainage system are graphed and analysed i n t h i s study.  A f i l t e r e d type of series with non-equal weights  was chosen over an equally weighted type to reduce the auto-correlation and increase the weight of the central plotted value.  Weights are a l l o t t e d  i n the manner i l l u s t r a t e d below: Year 1  Year 2  Year 3  Year 4  Year 5  6%  25%  3&fo  25%  6%  The f i l t e r e d value i s plotted at Year 3« One c h a r a c t e r i s t i c of any five-point f i l t e r e d series that should be noted i s the l o s s of two values at each end of a l l sequences of observations.  As the records cover r e l a t i v e l y short periods, and as a number  of observations are missing from records, r e a l values, denoted by dots, are plotted f o r the i n i t i a l and f i n a l two years of each sequence.  A. Last-ice  A t o t a l of 15 f i l t e r e d series of l a s t - i c e records are presented  73. i n graphic form of which eight are drawn from the Saskatchewan  trib-  utary basin, f i v e from the Assiniboine-Red t r i b u t a r y basin, and two from the M i s s i s s i p p i drainage basin.  The records are chosen with a view to  presenting as complete an areal picture of the Nelson drainage system as possible.  Where a choice i s available, the record with the longest  sequences of observations i s chosen over a record with more observations but l e s s continuity. Last-ice series from upstream s i t e s are plotted on the graphs of adjacent downstream series to give some v i s u a l i n d i c a t i o n of the degree of c o v a r i a b i l i t y between them.  On occasion, l a s t - i c e series from different  t r i b u t a r y basins are i l l u s t r a t e d on one graph f o r the same purpose. dashed l i n e ( - — )  A  represents a series from an upstream l o c a t i o n , and a  dotted l i n e (....) represents one from another t r i b u t a r y basin.  In some  instances, the bases of superimposed series from other t r i b u t a r y basins are adjusted f o r comparison purposes.  Ordinate values of superimposed  series can be read from the graph of the o r i g i n a l s e r i e s . V i s u a l inspection of l a s t - i c e series f o r Edmonton, Prince Albert, and The Pas (Figure 25) on the North Saskatchewan-Saskatchewan  River sug-  gests the following* 1. o s c i l l a t o r y movement of records from Edmonton and Prince Albert i s i n phase during the early years of record.  The two series  show l i t t l e agreement from 1935 to the end of t h e i r records; 2. short-term fluctuations at Prince Albert and The Pas are out of phase with the possible exception of the l a t e r years of record; 3. l a s t - i c e generally occurs l a t e r at a downstream s i t e than at an  74.  Days M a r c h  from  Edmonton  31  30-  20  Days M a r c h  Days M a r c h  Prince Albert.. Edmonton..  from 31  The Pas Prince Albert  from 31  1915  Days March  1945  1950  1945  1950  Years  Banff The Pas  from 31  2010-  1915  1920  1930  1935  1940  FIGURE 25 FIVE-POINT FILTERED SERIES OF LAST-ICE  1955  Y»ar»  upstream one but, occasionally, the reverse s i t u a t i o n holds true; and 4. there i s a s l i g h t trend towards e a r l i e r dates of l a s t - i c e from 1920 to the mid-1940*s. F i l t e r e d series of l a s t - i c e are graphed f o r f i v e s i t e s located along various parts of the South Saskatchewan system.  Three of the s i t e s  are located on headwater t r i b u t a r i e s and two are situated on the main channel.  The i n t e r n a t i o n a l Boundary and Lethbridge s i t e s on the St.Mary  River (Figure 26) have l a s t - i c e records which show some measure of cova r i a b i l i t y : o s c i l l a t i o n s d i f f e r i n amplitude but are e s s e n t i a l l y The St. Mary River i s one of the headwater t r i b u t a r i e s  i n phase.  on which l a s t - i c e  normally occurs e a r l i e r at the downstream s i t e than at the upstream one. The l a s t - i c e series from the Bow River at Banff tends to p a r a l l e l the movement of the series from the International Boundary s i t e on the St. Mary River.  A d i s t i n c t change i n the r e l a t i v e time of l a s t - i c e at the  two stations over the length of record i s evidence that the series are out  of phase where missing observations i n the mid-1930's cause each  series to be discontinuous.  Series from Saskatoon and Medicine Hat  (Figure 26), located along the main channel of the South Saskatchewan, show some agreement i n movement but the omission of a number of l a s t - i c e dates l i m i t s comparisons to a period of approximately twenty years. There i s some i n d i c a t i o n that l a s t - i c e series from the South Saskatchewan and i t s t r i b u t a r i e s trend towards e a r l i e r dates i n the 1920 to mid-1940 period.  From the mid-1940's to the end of record, a swing towards l a t e r  l a s t - i c e dates seems to be developing.  76.  St M a r y R i v e r — International B o u n d a r y Banff  Days from March 31  1920  1925  1930  1935  1940  1945  1950  -| 1955  Years  St. M a r y R i v e r — L e t h b r i d g e  Days from M a r c h 31  St M a r y Ri ver — Inter na t l o n a I Bdy..  1 1955  Days from March 31  Years  Saskatoon T h e Pas . .  20-  -n 1950  1915  Days from March 31  Medicine Hat Saskatoon  10-  O-  -10-20 1915  1930  1935  1945  F I G U R E 26 FIVE-POINT F I L T E R E D SERIES O F  LAST-ICE  1950  1955  Years  77.  In the M i s s i s s i p p i drainage basin, f i l t e r e d series of l a s t - i c e records (Figure 27) display c h a r a c t e r i s t i c s that are s i m i l a r t o f i l t e r e d series of records from the South Saskatchewan basin.  The swing to e a r l i e r  l a s t - i c e dates through the period 1920 to mid-1940 s, and the reversal of ,  swing from the mid-1940's to the end of record i s apparent i n both records. These series show a f a i r l y high degree of c o v a r i a b i l i t y . Last-ice series from the Assiniboine-Red basin (Figures 27 and 28) appear more stationary i n character than those from western parts of the Nelson drainage system.  A f a i n t trend towards e a r l i e r dates i s apparent  at Brandon and Headingley but records from Millwood, Emerson, and Dominion C i t y give no such i n d i c a t i o n .  Series from Brandon and Headingley appear  to swing i n unison with the exception of a short period i n the 1940 to 1950 range.  Series from Emerson and Headingley show some measure of cov-  a r i a b i l i t y with obvious differences i n t h e i r records occuring i n the period 1920 to 1932. The series from Dominion City on the Roseau River, a t r i b u t a r y of the Red River, does not fluctuate as greatly as the series from Emerson on the Red, but basic movements follow a similar pattern i n the l a t t e r h a l f of record. In order to determine whether some degree of c o v a r i a b i l i t y exists between l a s t - i c e dates throughout the Nelson drainage system as a whole, the f i l t e r e d series of l a s t - i c e from The Pas i s superimposed on those from Banff, Milk River, Saskatoon, and Brandon. show considerable agreement  Banff and The Pas (Figure 25)  i n o s c i l l a t o r y movement.  Saskatoon and The  Pas (Figure 26) have s i m i l a r trends i f absolute values (dots) are considered.  78.  M i l k R i v e r (town) — M i l k R i v e r T h e Pas  Days from M a r c h 31 20 IO  OH -io —i 1920  i  1  1925  1930  1940  1945  —I  j  1950  1955  1945  1950  1955  1945  1950  1955  Years  Milk R i v e r — E a s t e r n C r o s s i n g Milk Ri ver (town) — M i I k R i v e r  Days from M a r c h 31 20IO-  1920  1925  1940  Oays from M a r c h 31  1910  M i l l wood  1915  Brandon T h e Pas..  Days from M a r c h 31  1910  1915  — i 1920  1  1  1  1  1  1  1925  1930  1935  1940  1945  1950  FIGURE 27 FIVE-POINT FILTERED SERIES OF LAST-ICE  T 1955  Years  79.  H e a d i n g ley Brandon  Days from March 31  20-  1 1915  Years  1955  1920  Emerson H e a d i n g ley  Days from March 31  30  20-  1 1955  Days from M a r c h 31  Roseau River — D o m i n i o n City Emerson  1935  1940  FIGURE 28 FIVE-POINT FILTERED SERIES OF LAST-ICE  Years  80.  Milk River and The Pas (Figure 27) conform as to basic trend over the range of data.  Series from Brandon and The Pas (Figure 27) agree i n trend  but show considerable d i s p a r i t y i n o s c i l l a t o r y movement. The majority of records from the Nelson drainage system suggest there i s a trend towards e a r l i e r dates of l a s t - i c e i n the period from 1920 to the mid-1940's.  In most f i l t e r e d  i s apparent towards the end of record.  series, a swing t o l a t e r dates  Some s e r i e s , p a r t i c u l a r l y those  drawn from the southeastern part of the Nelson drainage system, fluctuate l e s s markedly than others and appear to be more stationary i n character. Fundamental agreement i n trend and reasonable l e v e l s of agreement i n o s c i l l a t o r y movement between l a s t - i c e series from most areas of the Nelson system suggest  that early i c e disintegration on headwater t r i b -  utaries i s an i n d i c a t i o n that early dates may be expected throughout the Nelson basin.  Later-than-normal i c e disintegration dates i n any t r i b -  utary basin would suggest that dates may be l a t e i n the other t r i b u t a r y basins of the Nelson system.  The graphs also indicate that i c e d i s i n t e -  gration dates at downstream locations are, to some degree, dependent upon upstream dates.  B. F i r s t - i c e  Thirteen f i r s t - i c e records are f i l t e r e d of which three are drawn from the North Saskatchewan-Saskatchewan, f i v e from the South Saskatchewan, one from the M i s s i s s i p p i , and three from the Assiniboine-Red River systems. F i r s t - i c e series from upstream channel s i t e s are superimposed  on graphs of  downstream channel s i t e s , and i n some instances, a series from one t r i b -  81.  utary basin i n superimposed on that from another t r i b u t a r y basin.  Super-  imposed f i r s t - i c e series are d i f f e r e n t i a t e d i n the same manner employed in treating last-ice series. F i r s t - i c e series from Edmonton and The Pas (Figure 29) show marked s i m i l a r i t i e s i n movement.  The series from Prince Albert, located  between Edmonton and The Pas, shows l i t t l e agreement with series from either of these l o c a t i o n s . The reasons f o r the disagreement between the series from Prince Albert and those from the other two l o c a t i o n s are not obvious.  v  F i r s t - i c e records from the North Saskatchewan-Saskatchewan River  show no marked trend, although there i s some suggestion that f i r s t - i c e occurs l a t e r towards the end of record. F i l t e r e d series of f i r s t - i c e dates from locations on the South Saskatchewan River system (Figures 29 and 3°) indicate a trend to l a t e r dates over the range of t h e i r records.  O s c i l l a t o r y movement of series  from downstream locations on the St. Mary and South Saskatchewan compares favorably with movement i n upstream series over the period 1935 to  1955.  There i s l i t t l e agreement between upstream and downstream series on either r i v e r p r i o r to  1935.  Brandon, Headingley and Emerson, i n the Assiniboine-Red basin, exhibit marked s i m i l a r i t i e s between f i r s t - i c e records (Figure 3l)»  Com-  parisons of f i r s t - i c e series show that major o s c i l l a t i o n s are i n phase with only minor discrepancies i n movement and amplitude.  The  first-ice  series from Milk River on the Milk River t r i b u t a r y of the M i s s i s s i p p i drainage basin shows some measure of c o v a r i a b i l i t y with those from the Assiniboine-Red basin (Figure 3l)«  A trend towards e a r l i e r dates of  82  Days from  Edmonton  The Pas ...  o c 3i  1910  Days Oct  from  31  1915  1920  1925  1930  1935  1940  1945  1950  Years  The Pas Prince Albert  -10  —i  1  Banff  Years  1955  1945  Days from O c t 31  1955  The Pas  3020-  Years  1940  1945  FIGURE 29 FIVE-POINT FILTERED SERIES OF FIRST-ICE  1950  1955  83.  St.Mary  Days from O c t 31  River — InternationaI  Boundary  3020IOO-  -IO-  Oays from Oct 31  St.Mary  River — Lethbridge  St  River — International  Mary  Bdy.  302010-  o-IO-  —r-  — i —  1955  1915 Days from Oct 31  Medicine T h e Pas  Years  Hat.  302010-  1925  1930  1945  1935  [  1955  Years  Saskatoon. . . M e d i c i n e Hat  Oays from O c t 31  2010-  o-20-  1910  —i 1920  1 1925  1  1930  1  1935  1  1940  1  1945  FIGURE 30 FIVE-POINT FILTERED SERIES OF FIRST-ICE  —I  1955  Years  84.  _„  ,  Doys Oct  Milk R i v e r (town) — Milk R i v e r ...  from  T h e Pas  31  1 I9IO  1915  1920  1925  1930  1935  1940  1945  Brandon  Days from Oct 31  Milk R i v e r (town) —Mi Ik R i v e r  20-  10-  O-  Days from O c t 31  Headingley. B r a n d o n ....  20-  10-  — i — 1915  Days from Oct 31  1925  Emerson. Headingley.  30  20  IOH  FIGURE 31 FIVE-POINT FILTERED SERIES OF FIRST-ICE  1950  1955  Yeor»  85.  f i r s t - i c e from the beginning of record to the l a t e 1930's i s evident i n both the Assiniboine-Red as well as the M i s s i s s i p p i basins.  The period  from the l a t e 1930's to the end of record gives some i n d i c a t i o n of swing to l a t e r dates of f i r s t - i c e . The f i l t e r e d series from The Pas i s superimposed on those from Edmonton, Banff, Medicine Hat and Milk River. the  As mentioned previously,  series from Edmonton and The Pas exhibit a r e l a t i v e l y high degree of  covariability.  Banff and The Pas are markedly similar from 1930 to the  end of record.  Series from Milk River, Medicine Hat and The Pas swing  in unison from 1940 to 1955 but p r i o r to t h i s period there i s l i t t l e evidence of c o v a r i a b i l i t y . The comparison of f i r s t - i c e series from different t r i b u t a r y basins suggests that some measure of c o v a r i a b i l i t y exists between f i r s t ice records.  In f a c t , the measure of agreement between f i r s t - i c e series  appears to be stronger than the agreement indicated between l a s t - i c e series.  In other words, climatic controls operating throughout the  Nelson basin exhibit a greater influence upon dates of f i r s t - i c e than on dates of l a s t - i c e ; l o c a l environmental factors appear to be of less importance i n the determination of f i r s t - i c e dates.  C. Ice-free Season  Cumulative percentual deviations (Crowe, 1958; Kraus, 1955) based on the mean length of i c e - f r e e season, 1921 to 1950, are computed and graphed f o r records from s i x stations located i n the Nelson drainage system.  The cumulative percentual deviation curves are derived from  86. the  equation,  c  L  -  m £ (L  L  -L)  The cumulative percentual deviation curve passes through the o r i g i n at the beginning of the year 1921,  and at the end of  1950.  Cumulative percentual deviations are used to f a c i l i t a t e comparisons between station records with means that d i f f e r over the period 1921  to 1950.  The cumulative percentual deviation curve r i s e s during  the years i n which the ice-free season i s consistently longer than the mean i c e - f r e e season.  The curve f a l l s during the years i n which the i c e -  free season i s consistently shorter than the mean ice-free season. yearly changes i n the length of i c e - f r e e season may  Small  only change the slope  of the curve without appreciably a f f e c t i n g the general trend (see Crowe 1958,  p.3; Kraus, 1955).  In those years i n which dates of l a s t - i c e con-  ditions are missing, the ice-free season i s assumed to be average i n length. Records are selected from stations that give some representation to the varying physical conditions that are found i n the major physiographic regions of the Nelson drainage system.  Neither the Canadian  Shield nor the Hudson Bay Lowlands are represented because of lack of data.  The l o c a t i o n of stations with respect to physiographic boundaries  i s shown i n Figure 32. Figure 33 indicates that the length of i c e - f r e e season at each of the s i x stations d i f f e r s considerably from year to year.  The  curves,  with the exception of the one representing Brandon's record, are b a s i c a l l y  co •o  Edmonton  Percent  I9IO  1915  1920  1925  1930  X = 201 d a y s  1935  1940  1945  1950  1955  |9|Q  1915  1920  1925  1930  1935  1940  1945  1950  1955  CO  FIGURE 33 CUMULATIVE PERCENTUAL DEVIATIONS FROM THE MEAN ICE-FREE SEASON, 1921 to 1950  00  89. valley-shaped. seasons.  This suggests that the trend i s towards longer i c e - f r e e  At Banff, Edmonton and The Pas, the ice-free season i s generally  shorter than average i n the period 1921 to the early 1930's; the season i s generally longer than average at these stations from the early 1930's to the end of record.  At Saskatoon, the i c e - f r e e season tends to remain  shorter than average u n t i l the 1940's and then swings to a longer than average season i n the l a t t e r portion of the record.  The curve f o r Leth-  bridge i s b a s i c a l l y W-shaped which indicates there are two periods of shorter than average ice-free season and two periods of longer than average i c e - f r e e season.  The curve f o r Brandon suggests the i c e - f r e e season i s  neither longer nor shorter than the mean f o r more than four years i n a row.  As the length tends to fluctuate f a i r l y rapidly with no tendency to  remain consistently shorter or longer than the mean, the i c e - f r e e season at Brandon may be regarded as more stationary i n character than the i c e free season at the other locations.  D. Mean Annual A i r Temperatures  and Ice-free Seasons  Cumulative percentual deviations from the mean i c e - f r e e season and from mean annual a i r temperatures, 1921 to 1950, are i l l u s t r a t e d i n Figure 34 f o r three locations i n the Nelson basin.  The graphs do not  indicate that positive and negative deviations from mean annual a i r temperatures and lengths of i c e - f r e e season are i n any measure synchronous. However, trends i n the deviations from the mean annual a i r temperatures, 1921 to 1950, show marked s i m i l a r i t i e s at the three locations.  If air  temperature trends are b a s i c a l l y s i m i l a r across the breadth of the Nelson  B.  I920  1925  1930  1935  1940  1945  1950  1900  I9IO  1920  1930  1940  1950  I960  Mean ice-free season M e a n a n n u a l air t e m p e r a t u r e .  FIGURE 34 A, B, and C. A COMPARISON OF THE CUMULATIVE PERCENTUAL DEVIATIONS FROM THE MEAN ICE-FREE SEASON AND FROM THE MEAN ANNUAL AIR TEMPERATURE D. A COMPARISON OF OPEN SEASON AND MEAN ANNUAL AIR TEMPERATURES  basin, as records from Edmonton, Saskatoon and Brandon i n d i c a t e , there i s l i t t l e reason to expect that the use of a i r temperatures from upstream locations would improve the l e v e l s of agreement i l l u s t r a t e d i n the graphs. Figure 34 also shows movements i n mean annual a i r temperatures (absolute values) at Prince Albert superimposed on the f i l t e r e d series of open season (break-up to freeze-up) f o r the same l o c a t i o n . show l i t t l e agreement i n o s c i l l a t o r y movement.  The two series  The comparison at Prince  Albert points to the same conclusion reached following the examination of cumulative  percentual deviations; namely, that varying lengths of open  or ice-free season cannot be d i r e c t l y r e l a t e d to changes i n mean annual a i r temperatures.  TRENDS IN FREEZE-UP AND  BREAK-UP DATA  A few locations i n the Nelson basin have longer freeze-up break-up records than f i r s t - i c e and l a s t - i c e records.  and  Edmonton, Prince  Albert, Winnipeg and S e l k i r k have break-up records that are 60 to 80 years long.  Of the four stations mentioned, only Edmonton and Prince  Albert have freeze-up records of a s i m i l a r duration.  These records are  f i l t e r e d and graphed i n the same manner used i n t r e a t i n g records of  first-  ice and l a s t - i c e a f f e c t i n g discharge. Break-up records from Edmonton and Prince Albert are throughout t h e i r ranges.  continuous  The freeze-up record from Prince Albert i s the  only series of observations that i s continuous  throughout i t s range.  As  Edmonton's freeze-up record has a number of gaps, graphical methods of comparison with other records are of questionable value.  Both freeze-up  92.  and break-up records from Winnipeg have gaps of 14 years, from 1893 to 1907.  Break-up observations are missing at Selkirk from 1916 to 1920  inclusive. A. Break-up Break-up records from Edmonton and Prince Albert (Figure 35) show marked s i m i l a r i t i e s i n o s c i l l a t o r y movement.  A l l major o s c i l l a t i o n s are  i n phase with some obvious differences i n amplitude.  Some minor variations  i n swing are apparent but the records are e s s e n t i a l l y i n agreement throughout t h e i r comparable ranges.  The measure of agreement tends to substantiate  the accuracy of break-up observations taken at each l o c a t i o n .  I t also dem-  onstrates the dependency of downstream break-up dates upon the time of i c e d i s i n t e g r a t i o n upstream. There i s an obvious trend towards e a r l i e r break-up at Edmonton and at Prince Albert.  The basic trend i n these two records supports conclus-  ions reached following the examination of trends i n l a s t - i c e observations at the same locations. Break-up records at Winnipeg and Selkirk (Figure 35) show some agreement where comparable data are a v a i l a b l e .  The general o s c i l l a t o r y  movements are i n accord from 1929 to the end of record.  There i s some  disagreement i n the 1922 to 1928 period but a graphic comparison of records from the Red River with those from the N r t h Saskatchewan River shows that 0  Selkirk's record i s i n almost complete agreement with those from Edmonton and Prince Albert during t h i s period (Figure 35).  In f a c t , throughout i t s  entire range, the f i l t e r e d series from Selkirk has a s u f f i c i e n t l e v e l of  93.  Edmonfon Prince Albert. Selkirk  Days from M a r c h 31. 40  30  20  ion  1900  Days from M a r c h 31  I9IO  Prince  Albert  30  20-  Days from M a r c h 31  Winnipeg  30-  20-  1890  Days from M a r c h 31  1930  I960  Selkirk..  30H  20H 10  I960  FIGURE 35 FIVE-POINT FILTERED SERIES OF BREAK-UP  94. TABLE XV DIFFERENCES BETWEEN MEANS OF TEN-YEAR BREAK-UP PERIODS AT WINNIPEG NOTE: I f t>2.88, the difference i s s i g n i f i c a n t at the 1 per cent l e v e l . I f t<2.88, the difference i s not s i g n i f i c a n t at the 1 per cent l e v e l . •f 1921-30  VALUES 1931-40  1941-50  1951-60  1921-30 1931-40  5.71  1941-50  3.66  1.96  1951-60  5.89  0.32  1.94  agreement with those from Edmonton and Prince Albert to suggest that Winnipeg' s record i n the 1922 to 1928 period i s anomalous. Student's ' t * test of the difference between means of ten-year periods indicates that the 1921 to 1930 period at Winnipeg d i f f e r s s i g n i f i c a n t l y from subsequent periods.  This tends to corroborate the results obtained by graphic com-  parisons.  Table XV shows the ' t * values obtained.  The c o v a r i a b i l i t y between break-up records from Edmonton, Prince Albert and S e l k i r k i s measured by Pearson's product-moment method of correlation.  The correspondence  matrix form (Table XVI).  between paired break-up dates i s l i s t e d i n  The results suggest that there i s a considerable  degree of covariation between break-up dates across the f u l l width of the Great Plains physiographic region.  The degree of c o v a r i a b i l i t y substant-  iates conclusions reached following the examination  of l a s t - i c e records.  The values obtained are s t a t i s t i c a l l y s i g n i f i c a n t at the one per cent  95. TABLE XVI COVARIABILITY BETWEEN PAIRED BREAK-UP DATES Note*  There are 51 p a i r s ; the same years are used i n determining each value. A l l values are s i g n i f i c a n t at the 1 per cent level. Edmonton  Prince Albert  Selkirk  Edmonton Prince Albert  .729  Selkirk  .468  level.  .480  Moreover, the r e s u l t s i l l u s t r a t e quantitatively the greater degree  of dependency between records from stations located on the same r i v e r than between those from separate tributary basins.  B. Freeze-up The freeze-up series from Prince Albert i s superimposed on that from Winnipeg (Figure 36B). reasonable l e v e l of agreement  A v i s u a l comparison suggests there i s a o s c i l l a t o r y movement with the exception  of the period 1922 to 1928. The period of disagreement i s the same i n which the break-up series from Winnipeg disagrees with those from Selkirk, Edmonton and Prince Albert. Both freeze-up series indicate', there are two dominant movements over the period 1900 to 1958. F i r s t - i c e series from the Assiniboine-Red basin have similar i n c i p i e n t trends although those from other locations i n the Nelson drainage system are not necessarily i n accord with these findings.  I'X'J 96,  Doys from O c t . 31  A.  Prince  Albert  _.  30 20IO  1870  Days from O c t 31  laao  I9SO  1920  Winni p e g  D  Prince  Albert.  20  1890  L e n g t h of open season (in d o y s )  E d m o n t o n.  A \J\y--  1900  L e n g t h of open season (in d a y s )  Q  1910  1920  1930  1940  1950  1940  1950  I960  t  Pnnce  Albert.  220  210-  1920  FIGURE 36 A and B. FIVE-POINT FILTERED SERIES OF FREEZE-UP C and D. FIVE-POINT FILTERED SERIES OF OPEN SEASON  —I  I960  Years  97. C. Open Season  The only long, continuous record of open season that i s available i n the Nelson basin i s from Prince Albert; Edmonton's record i s interrupted f o r extensive periods (Figure 36 G and D).  A dashed l i n e j o i n i n g absolute  values of open season i n the major interrupted period at Edmonton may  be  used as a rough guide to the o s c i l l a t o r y movement within t h i s period. Swings i n the series at Edmonton and Prince Albert tend to p a r a l l e l each other. 1912  At Prince Albert, the length of open season i s decreasing from  to 1928;  from 1928 to 1958,  the length of open season i s increasing.  In summary, f i l t e r e d series and cumulative percentual deviations of r i v e r i c e formation and d i s i n t e g r a t i o n data provide evidence of climatic fluctuations i n the Nelson drainage system. indicate  For example, l a s t - i c e series  that break-up occurred e a r l i e r between 1920  and the mid-1940's  than i n the previous or subsequent periods; f i r s t - i c e series show a trend to l a t e r dates of i c e formation over t h e i r range.  Again, the length of  the i c e free (open) season has tended to increase i n most parts of the Nelson system from the 1920*s into the 1950's.  Although yearly f l u c t 7  uations i n the lengths of i c e - f r e e season do not exhibit marked covaria b i l i t y with those i n mean annual a i r temperatures, an increasing trend i n the former i s accompanied by a r i s i n g trend i n the l a t t e r .  Some of  the environmental conditions a f f e c t i n g yearly variations i n ice formation and d i s i n t e g r a t i o n dates are discussed i n the following chapter.  CHAPTER V  ENVIRONMENTAL CONDITIONS AFFECTING RIVER ICE FORMATION AND DISINTEGRATION  Water flowing i n a r i v e r i s affected by the varying climatic conditions encountered along i t s path of flow.  Therefore, the formation of  ice at any given point on a r i v e r i s dependent  upon both the l o c a l en-  vironmental conditions as well as on those that exist upstream.  Similarly,  i c e d i s i n t e g r a t i o n i s affected by both upstream and l o c a l environmental factors. Freeze-up can be related to upstream a i r temperatures by estimating the rate of progress of a given mass of water and determining the a i r temperatures a f f e c t i n g i t along i t s flow path.  Flow-travel times i n con-  junction with a i r temperatures from enroute weather stations have been used to estimate freeze-up dates on the Mackenzie River (Mackay, 1961a: 1961b).  However, flow v e l o c i t y data f o r Nelson River t r i b u t a r i e s are not  available and estimation of t h e o r e t i c a l mean v e l o c i t i e s i s beyond the scope of t h i s t h e s i s . The d i s i n t e g r a t i o n of a r i v e r ' s i c e cover i n the spring i s a complex process.  Over-all increases i n the sum of direct and diffused  r a d i a t i o n received i n spring months are r e f l e c t e d i n the upward trend of a i r temperatures.  Above freezing temperatures occurring at an observation  s i t e w i l l cause the r i v e r ' s i c e cover to become p i t t e d and s t r u c t u r a l l y weakened.  In the upstream drainage areas, above freezing temperatures  99. increase run-off and flow rates i n the t r i b u t a r i e s and main channel of the r i v e r system.  I f upstream conditions produce s u f f i c i e n t run-off to  l i f t and crack the i c e cover, break-up and complete i c e d i s i n t e g r a t i o n may be expected to proceed r a p i d l y . One important f a c t o r a f f e c t i n g the time of i c e d i s i n t e g r a t i o n on a r i v e r i s the thickness of i t s i c e cover.  Ice thickness i s dependent  upon such factors as winter temperatures, time and depth of snow cover, current speeds and t u r b i d i t y , and so on. Ice thickness data are l i m i t e d i n extent.  Observations covering 16 years are available f o r the Assiniboine  River at Brandon but data from other r i v e r stations i n the Nelson basin i s l i m i t e d to one or two years.  In any event, the ice thickness determined  by borings at one l o c a t i o n may not necessarily be a good estimate of the thickness of a r i v e r ' s i c e cover.  The undersurface of the i c e cover w i l l  vary considerably depending upon such factors as the position of the main flow channels, the rates of flow, and the v a r i a b i l i t y i n the depth of the snow cover. "The thickness of the i c e cover has been estimated i n some areas by summed winter temperatures.  This method does not take into account the  development of a snow cover and i t s l i m i t i n g effects on i c e growth.  Var-  i a b i l i t y i n the depth of the snow cover caused by winds d r i f t i n g the snow and differences i n the snowfall regimes at locations i n the Nelson basin are also complicating factors.  However, summed winter temperatures  should  be more representative of i c e cover thicknesses at locations where snowfall is relatively light.  The seasonal f a l l of snow i s less at Brandon and  Saskatoon than i t i s at the other four locations examined i n t h i s study. Measurements of the accumulated  snow cover on the ground are taken  100. at a number of meteorological stations i n the Nelson basin on the morning of the l a s t day of each month.  Such data are of limited value i n the  determination of i c e thicknesses f o r a number of reasons.  For example,  snow depths measured on the l a s t day of the month are not necessarily c h a r a c t e r i s t i c of the month as a whole.  Freezing i n t e n s i t i e s w i l l vary  during the course of each month and, therefore, a continuous or daily record of snow cover i s necessary to determine with any degree of accuracy the r e l a t i o n s between snow cover and i c e thickness, A second reason i s that snow measurements taken i n comparatively d r i f t - f r e e areas do not necessarily r e f l e c t snow conditions on a r i v e r ' s wind-blown i c e surface. Snow cover data were examined early i n the study and discarded f o r these reasons. As indicated previously, many factors affect the formation and d i s i n t e g r a t i o n of a r i v e r ' s i c e cover.  Such meteorologic elements as  r a d i a t i o n , cloud cover, winds, vapour pressure, evaporation, and prec i p i t a t i o n , as well as the basic physiographic c h a r a c t e r i s t i c s of a drainage area, influence run-off and r i v e r i c e conditions.  However,  there i s i n s u f f i c i e n t hydrometeorologic data to permit a study of the heat budget of the Nelson drainage system.  In t h i s t h e s i s , only r e l a t i o n s  among a i r temperatures, dateB of ice formation and disintegration, and r i v e r discharge are discussed.  Emphasis i s placed on these r e l a t i o n s  for the following reasons: 1. s u f f i c i e n t data are available f o r the examination of these r e l a t i o n s at a number of locations i n the Nelson basin; 2. other studies have shown that various measures of a i r temperatures can be correlated with ice formation and disintegration (e.g.  101. Burbridge and Lauder, 1957; Chemoevanenko, 1954; Mackay, 1961a and 1961b; Shebanov, 1958; Shipman, 1938); and 3. increasing discharge i s an important f a c t o r i n the mechanical d i s i n t e g r a t i o n of a r i v e r ' s i c e cover (e.g. Currie, 1953; McMullen, 1961).  FREEZING DEGREE DAYS PRIOR TO ICE FORMATION A freezing degree day i s considered to be equivalent to one degree Fahrenheit below the freezing temperature of 32 degrees.  Therefore, a mean  daily temperature of 20 degrees i s the equivalent of 12 freezing degree days, and a daily reading of zero degrees i s equal to 32 freezing degree days.  The average number of freezing degree days are cumulated over a  ten-day period p r i o r to f i r s t - i c e dates, 1921 to 1950.  The results are  l i s t e d i n Table XVII f o r s i x locations i n the Nelson Basin. The greatest number of freezing degree days occur at The Pas and at Banff i n the ten-day period p r i o r to f i r s t - i c e a f f e c t i n g discharge. At Banff, i t may be that a steep gradient and consequent high rate of flow require the presence of f a i r l y low l o c a l a i r temperatures before a sufficient  amount of ice to affect discharge i s formed on s i t e or i n the  control section.  At The Pas, i t i s possible that the i n f l u x of warmer  water into the North Saskatchewan-Saskatchewan  River system from the South  Saskatchewan River requires that l o c a l temperatures be below the freezing point f o r some length of time p r i o r to the formation of i c e a f f e c t i n g d i s charge.  At both locations, however, f i r s t - i c e dates have been reported i n  years when l o c a l mean d a i l y temperatures were above the freezing point. Ice  carried downriver from some distance upstream and the formation of  102.  TABLE XVII AVERAGE NUMBER OF CUMULATED FREEZING DEGREE DAYS PRIOR TO FIRST-ICE, 1921 TO 1950 Days P r i o r to F i r s t - i c e Affecting Discharge STATION  0  25.2  36.8  52.9  4.3  12.2  22.4  -  -  0.6  7.0  6  5  4  2.2  4.8  5.1  8.6  15.9  -  -  -  -  -  -  1.0  2.2  8.9  12.6  8  Banff  -  Edmonton  -  Lethbridge  M M  Saskatoon  -  Brandon  -  -  5.5  2.0  1  7  9  The Pas  10  -  3  2  -  -  -  1.0  4.9  12.9  24.2  36.1  3.9  6.1  10.5  16.4  22.8  30.6  39.4  16.9  23.5  32.9  42.7  53.7  64.3  75.8  ice i n sheltered areas under minimum daily temperature conditions may exp l a i n these dates. The least number of freezing days i n the ten-day period p r i o r t o f i r s t - i c e a f f e c t i n g discharge occurs at Lethbridge and at Edmonton.  The  r i v e r s are deeply entrenched at both these locations which suggests that l o c a l a i r temperatures may not be c h a r a c t e r i s t i c of those i n contact with the free water surface.  The S t . Mary River, i n the Lethbridge area, i s  entrenched approximately 250 feet below the r o l l i n g p r a i r i e countryside. I t s sharp-walled valley i s i d e a l f o r the pooling of cool a i r and, therefore, early morning temperatures at the r i v e r surface may f a l l w e l l below those recorded at Lethbridge a i r p o r t .  The same s i t u a t i o n may p r e v a i l at Edmonton  where the North Saskatchewan flows at an elevation of approximately 150 feet below the height at which a i r temperatures are recorded.  103. The years i n which above freezing temperatures occurred on the dates when f i r s t - i c e a f f e c t i n g discharge was recorded are l i s t e d i n Table XVIII.  An examination of Table XVIII suggests the following:  1. above freezing temperatures occurred on f i r s t - i c e dates most frequently at Edmonton and Lethbridge. This suggests that l o c a l a i r temperature observations are less c h a r a c t e r i s t i c of temperatures at the free water surface at these locations than at the  other locations examined;  2. with few exceptions, minimum a i r temperatures are below the freezing point on f i r s t - i c e dates.  In most cases, the ten-day  period p r i o r to f i r s t - i c e had a majority of days with minimum temperatures below 32°F.  Ice formation and accretion i n sheltered  areas under such temperature conditions may be an important factor in the development  of i c e conditions a f f e c t i n g discharge;  3. the extreme v a r i a b i l i t y of a i r temperature conditions i n the tenday period p r i o r to f i r s t - i c e i s i l l u s t r a t e d by the differences between the ten-day mean a i r temperatures.  The degree of v a r i a -  b i l i t y indicates that any given set of a i r temperatures at the observation s i t e does not necessarily give r i s e to ice conditions a f f e c t i n g discharge.  In order to r e l a t e l o c a l a i r temperatures  to the formation of f i r s t - i c e some knowledge of water temperatures seems necessary. The number of days on which mean daily temperatures are below 32°F. p r i o r to f i r s t - i c e dates d i f f e r s markedly from year to year at each location.  The average day on which mean daily temperatures reached the  104. TABLE XVIII YEARS WHEN MEAN DAILY AIR TEMPERATURE WAS ABOVE FREEZING ON DATE OF FIRST-ICE AFFECTING DISCHARGE, 1921 TO 1950  A i r Temp. In 10-day Period P r i o r to F i r s t - i c e on F i r s t - I c e Date No. of Days Max. Min. Mean Mean Air Temp. Minimum op op op op A i r Temp.32°F  Station  Year  BANFF  1941 1939 1938  43 43 39  24 33  1949 1944 1943 1941 1937 1930 1923  55 46 49 45 42 41 46  19 20 24 21 27  LETHBRIDGE  1941 1936 1934 1933 1930 1929 1924  SASKATOON  33.5 38 34.5  19.8 39.4 26.2  10 6 10  29  37 33 36.5 33 34.5 36.5 37.5  35.0 28.3 35.6 37.4 37.2 42.5 32.2  9 10 9 7 9 4 10  52 57 42 56 50 40 51  28 37 30 38 17 29 21  40 47 36 47 33.5 34.5 36  28.6 37.8 34.0 39.4 41.4 41.0 43.4  9 9 8 7 6 7 6  1946 1941 1934 1930  50 64 49 42  21 35 27 30  35.5 49.5 38 36  34.1 47.8 30.0 44.2  10 6 10 3  BRANDON  1941 1940 1936  57 44 60  26 30 31  41.5 37 45.5  30.2 12.8 25.2  9 10 9  THE PAS  1939 1936 1929 1923  50 48 42 48  22 36 30  36 42 36 40  27.3 14.8 27.5 29.7  10 10 10 8  EDMONTON  30  32  32  105. freezing point p r i o r to the onset of i c e conditions (based on the period 1921 to 1950)  at each s i t e i s indicated i n Table XVII.  The  differences  between the average length of the freezing period p r i o r to f i r s t - i c e at the s i x locations makes i t impractical to compute a measure of the vari a b i l i t y over a standard ten-day period. separately.  considered  Standard deviations of degree days f o r the f r e e z i n g period  at each l o c a t i o n are l i s t e d i n Table XIX. standard  Each location must be  In a l l but one instance, the  deviation values exceed the expected number of freezing degree  days p r i o r to f i r s t - i c e .  the extreme v a r i a b i l i t y indicates the impractic-  a b i l i t y of r e l a t i n g l o c a l a i r temperature records to f i r s t - i c e dates without some knowledge of water temperatures.  MELTING DEGREE DAYS PRIOR TO ICE DISINTEGRATION  Above freezing temperatures occur at an observation various periods of time p r i o r to i c e d i s i n t e g r a t i o n . freeze-thaw cycles may  site for  In some years,  begin two months or more i n advance of break-up.  In the l a t t e r stages of the winter phase, the cycles are usually diurnal with thawing periods gradually increasing i n length as the incursions of P a c i f i c maritime weather systems become more frequent during the spring phase.  The number and i n t e n s i t y of freeze-thaw cycles are an  important factor i n r i v e r i c e disintegration due to t h e i r e f f e c t upon the strength of an i c e cover.  106.  TABLE XIX STANDARD DEVIATIONS OF CUMULATED DEGREE DAYS PRIOR TO FIRST-ICE (Computed f o r average f r e e z i n g periods, 1921 to 1950;including f i r s t - i c e dates) Standard Deviations ( i n degree davs)  Average Freezing Period ( i n davs)  STATION Banff  8  65  Edmonton  4  49  Lethbridge  3  18  Saskatoon  6  69  Brandon  9  65  The Pas  11  75  The melting degree day concept does not deal adequately with the problem of freeze-thaw cycles nor with yearly variations i n snow cover depths on a r i v e r ' s i c e surface.  However, i t does indicate l o c a l  thaw i n t e n s i t i e s p r i o r to i c e disintegration.  The number of melting  degree days i n any one day equals the number of degrees the mean daily temperature i s above the freezing point.  In other words, a mean daily  temperature of 40 degrees Fahrenheit i s equivalent to eight melting degree days; a 50 degree mean daily temperature equals 18 melting degree days, and so on. The average number of melting degree days cumulated over the ten-day period p r i o r to l a s t - i c e dates, 1921 to 1950, i s l i s t e d i n Table XX, f o r s i x locations i n the Nelson Basin.  107.  TABLE XX AVERAGE NUMBER OF CUMULATED MELTING DEGREE DAYS PRIOR TO LAST-ICE, 1921 TO 1950 Number of Days P r i o r to Last-ice Dates 10  Station  -  Banff Edmonton  3.7  -  Lethbridge Saskatoon Brandon  -  The Pas  1.4  9  10.0  -  8  17.2  7  6  5  4  3  2  1  0  -  -  -  -  2.2  2.8  4.1  4.1  23.9  32.8  42.5  53.9  65.2  76.4  87.9  98.6  1.6  2.2  4.9  7.6  8.1  9.5  10.3  -  -  1.2  2.4  4.0  5.5  10.0  17.9  28.8  38.8  -  -  0.4  3.5  6.1  9.8  15.9  22.9  29.9  36.4  3.3  6.8  11.2  15.7  21.1  26.7  34.3  41.3  47.7  55.9  -  Some locations exhibit marked differences i n the average number of melting degree days occurring over the ten-day period p r i o r to l a s t ice dates.  The smallest accumulations  Banff and Lethbridge.  of melting degree days occur at  In the Banff area, a steep r i v e r gradient may  i n h i b i t the formation of a thick i c e cover; at Lethbridge, the  chinook  effect and the resultant high number and i n t e n s i t y of freeze-thaw cycles may  hinder i c e cover development.  Thus, break-up on the St. Mary River  near Lethbridge, and on the Bow River at Banff appear to occur with r e l a t i v e l y small changes i n climatic  conditions.  The largest average number of melting degree days occur at Edmonton.  Although mean winter temperatures  indicate that the i c e cover  on the North Saskatchewan at Edmonton would not be as thick as the i c e cover on the South Saskatchewan at Saskatoon, or that on the Assiniboine  108. at Brandon, break-up i s l a t e r at Edmonton.  Comparisons of temperature  regimes (Figure 37) at these locations shows that the l a t e occurrence of break-up at Edmonton provides an explanation f o r the greater average number of accumulated melting degree days.  However, the reasons f o r  late break-up at Edmonton are less obvious.  The proximity of Edmonton's  l o c a t i o n to the g l a c i e r and snow-fed headwaters of the North  Saskatchewan  suggests that the amelioration of water temperatures by a i r temperatures during the open season i s l i m i t e d by the flow-travel distance.  I f water  temperatures i n the f a l l are lower at Edmonton than at the other two locations, the change to the winter phase would proceed more r e p i d l y . As, i n f a c t , freeze-up i s generally e a r l i e r at Edmonton, the longer closed season would permit greater i c e accretion and, thus, delay i c e d i s i n t e g r a t i o n i n the spring. The years i n which l a s t - i c e dates, 1921 to 1950, were recorded under freezing conditions at s i x locations i n the Nelson basin are l i s t e d i n Table XXI.  In most instances, maximum a i r temperatures are above the  f r e e z i n g point on l a s t - i c e dates.  The ten-day periods p r i o r to l a s t -  ice dates exhibit marked v a r i a b i l i t y i n mean ten-day a i r temperatures and i n the number of days that mean d a i l y a i r temperatures are above 32°F.  The v a r i a b i l i t y of a i r temperature conditions p r i o r to l a s t - i c e  indicates the importance that other factors such as discharge play i n the process of i c e cover disintegration.  Standard deviations of a i r  temperatures i n degrees Fahrenheit and i n degree days f o r the ten-day period preceding l a s t - i c e are l i s t e d i n Table XXII.  109.  -io  Nov.  Dec.  Jan.  Feb  Mar.  F I G U R E W I N T E R  AIR  T E M P E R A T U R E  R E G I M E S  1921  to  Apr.  May  Months  37  A N D  1950  A V E R A G E  D A T E S  O F  L A S T - I C E ,  110.  TABLE XXI YEARS WHEN MEAN DAILY AIR TEMPERATURE WAS BELOW FREEZING ON DATE OF LAST-ICE AFFECTING DISCHARGE, 1921 TO 1950  Station  Year  BANFF  1950 1949 1948 1946 1945 1942 1940 1939 1935 1932  EDMONTON LETHBRIDGE  SASKATOON  BRANDON  THE PAS  on Last-ice Date Min. Mean Max. °F °F °F  Mean A i r Temp. °F  No.of Days Mean Daily Air Temp.32°F  17 26  17 8 3 6 12 11 - 4 -14  26.5 30.0 28,0 22.0 19.0 15.0 20.5 22.0 6.5 6.0  25.7 31.4 18.7 29.2 33.2 26.3 34.8 41.8 18.5 11.0  0 6 0 3 6 2 7 7 0 1  1934 1927  38 34  25 6  31.5 20.0  42.4 33.7  10 6  1940 1930 1927 1926 1924 1923 1921  43 35 42 36 38 40 13  20 18 15 16 21  31.5 26.5 28.5 26.0 29,5 31.5 12.5  37.0 26.1 31.6 20.2 28.0 35.0 9.9  7 3 6 0 2 6 2  1947 1945 1939 1933  37 48 33 38  22 14  29.5 31.0 28.0 31.5  26.2 33.1 31.0 32.0  1 6 4 4  1946 1945 1942 1935 1933 1931 1928  36 37 33 22 29 40 25  26 22 29 17 22 21  31.0 29.5 31.0 19,5 25.5 31.5 23.0  39.1 32.5 27.0 35.0 34.1 39.9 27.8  8 6 1 6 7 8 2  1948 1934 1931  48 27 36  25 10 19  31.5 18.5 27.5  36.9 32.4 41.6  8 4 8  38 37 39 36 35 24 29  33  15  23  23  12  23  25  23  111.  TABLE XXII STANDARD DEVIATIONS OF MEAN AIR TEMPERATURES AND CUMULATED DEGREE DAYS (Computed f o r ten-day periods p r i o r to l a s t - i c e , 1921 to 1950) Standard Deviations A i r Temperature (°F)  Station  Degree Days  Banff  6.7  67  Edmonton  4.3  43  Lethbridge  7.1  71  Saskatoon  5.0  50  Brandon  5.3  53  The Pas  4.9  49  SUMMED WINTER TEMPERATURES, ICE THICKNESSES, AND ICE DISINTEGRATION DATES  Maximum i c e thicknesses on the Assiniboine River at Brandon have been measured and recorded f o r the 16-year period, 1943 to 1958 (Meteoro l o g i c a l Branch, Department of Transport, Canada, CIR-3195 ICE-4, Oct. 1961).  A comparison of maximum i c e thickness data and break-up dates at  Brandon (Figure 38) indicates there i s some measure of c o r r e l a t i o n but the r e l a t i o n s h i p i s not s t a t i s t i c a l l y s i g n i f i c a n t .  Similar r e s u l t s were  obtained with comparisons between i c e thicknesses at Brandon and l o c a l winter temperatures summed over various periods.  Break-up dates and  l a s t - i c e dates at Edmonton and at Saskatoon were plotted against l o c a l winter temperatures summed over the periods from freeze-up to break-up, f i r s t - i c e to l a s t - i c e , and November to March i n c l u s i v e .  The r e s u l t s  112  Days Marc i  from 31.  BRANDON  23-  •  • •  19-  / /  •  Break-up dalres  15-  • / /  3 -  /  /  /  /  • 1 26  /  /  /  / /  •  /  / • •  /  -1 -  / /  /  •  / i  i  28  30  1 32  Maximum  1 34  i 36  ice thickness  1 38  40  FIGURE 38 MAXIMUM ICE THICKNESSES AND B R E A K - U P D A T E S ON T H E ASSINIBOINE R I V E R , 1943 to 1958  113. show that summed l o c a l winter temperatures are unreliable guides to dates of ice d i s i n t e g r a t i o n . RIVER DISCHARGE AND  ICE DISINTEGRATION  Each of the major tributary basins of the Nelson system cover vast areas which have manifold physical and c l i m a t i c differences. problem of assessing the influence of i n d i v i d u a l physical and  The  climatic  factors upon r i v e r ice disintegration i n such areas i s an extremely d i f f i c u l t task.  A B stated previously,  i c e disintegration at any  given  point on a r i v e r i s dependent not only upon l o c a l environmental conditions but on those e x i s t i n g i n the upstream portions of the drainage basin as w e l l .  One  element i n the disintegration of a r i v e r ' s ice  cover which measures the aggregate effects of the varying environmental conditions that exist upstream i s r i v e r discharge. Increasing  discharge i s a dynamic factor i n the mechanical d i s -  integration of a r i v e r ' s ice cover.  Although the process of i c e d i s -  integration i s not necessarily dependent upon s i g n i f i c a n t increases discharge, the rates at the seven locations examined generally  in  increased  three to f i v e - f o l d i n the period ranging from seven to f i f t e e n days immediately p r i o r to i c e d i s i n t e g r a t i o n .  In cases where l a s t - i c e dates  have occurred without a rapid increase i n discharge immediately p r i o r to break-up, discharge records usually show e a r l i e r marked f l u c t u a t i o n s . These may  have materially weakened the ice cover to the extent that l o c a l  environmental conditions  can give r i s e to break-up.  Cases i n which no  rapid increase i n discharge rates preceded ice disintegration occur i n approximately ten per cent of the years under examination.  114. Mean discharge rates p r i o r to l a s t - i c e dates, 1921 to 1950, are computed f o r seven gauging stations located i n the Great P l a i n s physiographic region.  Discharge records from two stations i n the headwater'3  area of the South Saskatchewan River, Lethbridge and discarded f o r two reasons.  and Banff, were examined  F i r s t , such locations show minimal s t a b i l i t y  of flow rates from year to year due to the flashy character of headwater streams.  Second, t h i s c h a r a c t e r i s t i c suggests that the representation  of these discharge rates by averages would be meaningless. Mean discharge rates i n the immediate period p r i o r t o l a s t - i c e dates are i l l u s t r a t e d i n Figure 39 f o r two locations on the North Saskatchewan, two on the South Saskatchewan, two on the Assiniboine, and one on the Saskatchewan River.  The hydrographs show that the rates of i n -  crease i n the six-day period p r i o r to l a s t - i c e are approximately the same at a l l stations except The Pas i n spite of the fact that mean d i s charge d i f f e r s markedly on the various t r i b u t a r i e s .  The time of the  i n i t i a l upswing varies from seven to eleven days p r i o r to dates of l a s t ice at stations on the North Saskatchewan, South Saskatchewan, and Assiniboine Rivers.  A l e s s e r rate of increase spread over a longer  period i s indicated at The Pas on the Saskatchewan River. Following i c e d i s i n t e g r a t i o n , discharge  continues  to r i s e f o r a  period of one day at Saskatoon and two days at Edmonton before a drop i s noticeable.  Mean discharge rates at the other locations show a tendency  to drop or l e v e l o f f immediately following the date of l a s t - i c e .  Whether  dates of i c e conditions i n the gauging records from which l a s t - i c e dates are extracted are i n c l u s i v e at a l l locations i s d i f f i c u l t to ascertain. Thus, differences i n the rates of discharge following l a s t - i c e dates may  FIGURE 39 HYDROGRAPH OF MEAN DISCHARGE RATES PRECEDING AND FOLLOWING LAST-ICE DATES, 1921 to 1950  116. be n e g l i g i b l e or, possibly greater than they appear to be i n Figure 39. The examination of relations among r i v e r discharge rates, a i r temperatures, and i c e dates points up some of the effects of varying environmental conditions on freeze-up and break-up.  For example, freeze-  up and break-up dates at locations i n entrenched r i v e r valleys are i n fluenced by the pooling of cool a i r .  Other physical factors such as  vegetative, cover, rock types and strata tend to regulate run-off and affect the time of i c e formation and d i s i n t e g r a t i o n .  Again, differences  i n r i v e r gradients and, consequently, flow v e l o c i t i e s are important i n the determination of i c e dates.  Two of the more important hydrometeor-  o l o g i c a l factors a f f e c t i n g break-up are i c e cover thicknesses and r i v e r discharge rates.  Unfortunately, data of the former has not been recorded  i n s u f f i c i e n t quantities to ascertain the r e l a t i o n between break-up dates and i c e cover thicknesses. However, discharge data i s available f o r analysis and results show that increasing rates of discharge greatly influence dates of break-up.  As f a r as freeze-up i s concerned, i t  appears that water temperature data are required before estimates of freeze-up dates i n the Nelson system can be made.  When more hydro-  meteorological data are available, a greater understanding of the processes involved i n r i v e r i c e formation and disintegration may be gained.  CHAPTER VI  SUMMARY AND CONCLUSIONS  This study i s based c h i e f l y on s t a t i s t i c a l comparisons of two kinds of data.  Hydrologie records of f i r s t - i c e and l a s t - i c e are compared  with h i s t o r i c a l records of freeze-up and break-up. desirable because records overlap at some locations.  Such comparisons are The sections which  overlap were tested f o r s i m i l a r i t i e s (Pearson's product-moment c o r r e l a t i o n c o e f f i c i e n t ) and f o r differences (the Wilcoxon matched-pairs test). 1.  signed-ranks  The p r i n c i p a l conclusions are as follows: Dates of (a) f i r s t - i c e with freeze-up, and (b) l a s t - i c e with break-up can be correlated s i g n i f i c a n t l y .  2.  S i g n i f i c a n t differences exist between paired series (see 1.) of observations at some locations.  However, the degree of  c o r r e l a t i o n suggests some extrapolation from one kind of data to the other i s possible. 3.  Hydrologie records of f i r s t - i c e and l a s t - i c e may be used to extend h i s t o r i c a l records of freeze-up and break-up, and to increase the coverage i n c e r t a i n areas.  A considerable amount of data related to i c e conditions a f f e c t i n g discharge i s available f o r the Nelson drainage system.  These data were  used to describe and to analyse (a) the areal progress of i c e formation and ice d i s i n t e g r a t i o n i n the Nelson drainage system, and (b) the progress  118.  of i c e formation and i c e disintegration along i n d i v i d u a l t r i b u t a r i e s of the Nelson River. 1.  The p r i n c i p a l conclusions are as follows:  In general, isopleths of mean f i r s t - i c e dates and tests of d i f ferences (the Wilcoxon matched-pairs signed-ranks test) between f i r s t - i c e records from l a t i t u d i n a l l y - s e p a r a t e d stations support the expected north-to-south progress of i c e formation.  However,  v a r i a b i l i t y i n the rates of i c e formation a f f e c t i n g discharge has s i g n i f i c a n t l y altered the d i r e c t i o n of progress i n some areas.  For example, the map  of mean f i r s t - i c e dates (Figure 15)  suggests that i c e conditions a f f e c t i n g discharge spread r a d i a l l y , west, south and east, from the area of e a r l i e s t occurrence i n the north central part of the Nelson drainage basin. 2.  In general, isopleths of mean l a s t - i c e dates and tests of d i f ferences (the Wilcoxon matched-pairs signed-ranks test) between l a s t - i c e records from l a t i t u d i n a l l y - s e p a r a t e d stations support the expected south-to-north progress of ice d i s i n t e g r a t i o n . However, v a r i a b i l i t y i n the rates of i c e disintegration cause a s h i f t i n the d i r e c t i o n of l a s t - i c e progress i n some areas. In the central section of the Nelson drainage system, the progress of l a s t - i c e trends from southwest to northeast. clockwise swing i n trend i s even more pronounced  The  i n the Rocky  Mountain f o o t h i l l s physiographic region. 3.  Isopleths of l a s t - i c e , 1921 to 1950, and isotherms of mean monthly temperatures f o r March, & p r i l and May show marked s i m i l a r i t i e s i n trends which suggest they can be correlated.  119  4.  Ice formation along those i n d i v i d u a l t r i b u t a r i e s of the Nelson River tested does not seem to respond to controls other than l a t i t u d e and continentality.  5.  Ice disintegration on those i n d i v i d u a l t r i b u t a r i e s of the Nelson River tested, begins i n the upper reaches and progresses i n a downstream d i r e c t i o n .  6.  Chinooks may contribute to early i c e disintegration and to lengthening the mean ice-free season i n the southwestern f o o t h i l l s of Alberta.  7.  Standard deviations of f i r s t - i c e and l a s t - i c e , 1921 to 1950, indicate that the magnitude of 'within v a r i a b i l i t y * i s r e l a t i v e l y homogeneous f o r a l l station records examined from the Great Plains physiographic region. A marked increase i n v a r i a b i l i t y occurs i n records from stations situated i n the f o o t h i l l s and Rocky Mountain regions.  The basic patterns of isopleth orientation f o r mean dates of l a s t - i c e and mean length of ice-free season covering the period of the climatic norm, 1921 to 1950, are i n general agreement with those derived by Burbridge and Lauder (1957) i n t h e i r analyses of freeze-up and breakup data.  However, various differences i n the time element do e x i s t ;  these differences are partly explained by the exclusion from t h i s study of data concerning ice formation and i c e disintegration on lakes and by certain inconsistencies i n the d e f i n i t i o n s of terms used i n recording basic data. On occasion, maps of i c e conditions may contradict the progress  120  of i c e formation and i c e d i s i n t e g r a t i o n suggested by s t a t i s t i c a l tests of the differences. With due respect to the f i v e per cent s i g n i f i c a n c e l e v e l s , i t i s assumed that the s e n s i t i v i t y of means to extreme values, coupled with the differences i n the base periods are responsible f o r any apparent contradictions. The study of trends i n r i v e r i c e condition at a number of locations i n the Nelson drainage system i s based predominantly on filtered  series of i c e formation and d i s i n t e g r a t i o n dates, and on  cumulative percentual deviations from the mean ice-free season, 1921 to 1950. V i s u a l inspection of the graphs leads to the following conclusions ; 1.  The majority of l a s t - i c e records indicate a trend towards e a r l i e r dates of l a s t - i c e from 1920 to the mid-1940's.  A  swing towards l a t e r dates i s indicated i n the l a t t e r years of record.  Series from the Assiniboine-Red basin are more  stationary i n character than series from the other major t r i b utary basins. 2.  F i r s t - i c e series do not indicate any marked secular movement. Most records show a swing to l a t e r f i r s t - i c e dates i n the l a t e r years of record.  The Assiniboine-Red records swing to e a r l i e r  dates from the beginning of record to the late 1930's. 3.  The c o v a r i a b i l i t y between f i r s t - i c e records appears to be greater than the c o v a r i a b i l i t y between l a s t - i c e records.  This  implies that fundamental climatic controls and weather systems operating throughout the Nelson basin exert a greater measurable influence upon dates of f i r s t - i c e than on dates of l a s t - i c e .  121.  The weather conditions that induce break-up on the headwaters of the North and South Saskatchewan Rivers are extremely variable. As i c e d i s i n t e g r a t i o n dates at downstream locations are i n some measure dependent upon dates of i c e disintegration at upstream s i t e s , the v a r i a b i l i t y of break-up on the headwaters  could be  a f a c t o r i n l i m i t i n g the c o v a r i a b i l i t y between l a s t - i c e records throughout the basin. 4.  Cumulative percentual deviations from the mean length of i c e free season, 1921 to 1950, show that the ice-free season was shorter than average from 1921 to the early 1930's, and longer than average i n the l a t t e r h a l f of record at most locations i n the Nelson basin.  The f i l t e r e d series of open season at Prince  Albert exhibits s i m i l a r movements. 5.  Variations i n the length of open or ice-free season cannot be d i r e c t l y related to variations i n mean annual a i r temperatures.  An examination of climatic and hydrologic factors a f f e c t i n g r i v e r i c e formation and d i s i n t e g r a t i o n suggests the following conclusions:. 1.  Local a i r temperatures i n the ten-day period p r i o r to the onset of i c e conditions are extremely variable.  On occasion, mean  temperatures indicate thawing conditions when f i r s t - i c e dates are reported. (a)  Some possible explanations are:  i c e formed under l o c a l minimum a i r temperature conditions may be s u f f i c i e n t to affect the normal stage-discharge r e l a t i o n curve, and cause f i r s t - i c e dates to be reported;  (b)  i c e formed under freezing i n t e n s i t i e s upstream may f l o a t  122 down i n s u f f i c i e n t quantities to cause f i r s t - i c e dates to be reported; and (c)  data inaccuracies  due to misinterpretation  of the stage-  discharge r e l a t i o n curve or t r a n s c r i p t i o n a l errors may influence 2.  results.  Cumulated melting degree days p r i o r to ice disintegration provide l i t t l e or no i n d i c a t i o n of the onset of ice disintegration at the s i t e s examined. conditions.  Last-ice may occur under thawing or freezing  Due to the averaging process, the computation of  melting degree days from mean daily temperatures does not consider the effect of diurnal freeze-thaw cycles on the s t r u c t u r a l strength of an ice cover.  The number and i n t e n s i t y of freeze-  thaw cycles i s an important factor i n the determination of ice disintegration. 3.  There i s some i n d i c a t i o n of a stochastic relationship between maximum i c e thickness data and break-up dates recorded at Brandon, on the Assiniboine  River.  Comparisons of summed winter  temperatures and maximum ice thicknesses at Brandon y i e l d s i m i l a r results.  However, neither of these relationships i s s t a t i s t i c -  ally significant. 4.  River discharge appears to be the most important single measurable factor involved i n the process of ice disintegration. Records from seven Great Plains locations show that marked, i n creases i n discharge precedes i c e disintegration dates i n nine out of ten cases.  Rates of discharge increase three to four  times i n the ten-day period prior to l a s t ice dates.  However,  123. headwater stations i n the Rocky Mountains and F o o t h i l l s regions exhibit considerable v a r i a b i l i t y i n discharge rates p r i o r to l a s t - i c e dates and statements made regarding increases i n d i s charge do not apply i n these regions,  A knowledge of water temperatures i n conjunction with flowt r a v e l times would be of great value i n explaining i c e formation dates. Although a i r temperatures a f f e c t i n g east-flowing streams crossing the Great Plains are r e l a t i v e l y uniform due to minimal l a t i t u d i n a l effect and the dominant westerly c i r c u l a t i o n , environmental conditions a f f e c t ing water temperatures i n mountainous headwater areas are extremely variable.  Thus, water temperature data from streams crossing the area  of t r a n s i t i o n between the Rocky Mountain F o o t h i l l s and Great Plains Region would seem a prerequisite to the accurate estimation of downstream i c e formation dates from a i r temperatures and flow-travel times. River discharge i s a r e l i a b l e guide to environmental conditions e x i s t i n g upstream p r i o r to i c e disintegration.  However, break-up pre-  d i c t i o n i s also dependent upon such l o c a l environmental conditions as the strength and thickness of the r i v e r ' s i c e cover.  Accurate estim-  ation of these factors without more information related to snow cover conditions i s extremely d i f f i c u l t .  When adequate snow cover and i c e  thickness data are a v a i l a b l e , reasonable estimates of i c e disintegration dates would appear to be possible.  REFERENCES Brown, R.J.E. 1957s "Observations on break-up i n the Mackenzie River and i t s delta i n 1951*". Journal of Glaciology, V o l . 3, No. 22,  pp.133-11*2.  .  Burbridge, F.E. and Lauder, J.R. 1957 s "A r^eliminarv i n v e s t i g a t i o n i n t o break-up and freeze-up conditions i n Canada". Meteorological Branch, Department of Transport, Canada. GTR-2939, TEC-252. Toronto, 13 p. Callaway. E l l i o t t B. 1951*: "An analysis of environmental f a c t o r s a f f e c t i n g i c e growth". U.S. Navy Hydrographic O f f i c e , Tech. Report TR-7, 31 p. Chernoevanenko, E.M. 1953J "Concerning the prognoses of times of break-up and c l e a r i n g of i c e of the r i v e r s of the Don basin". Private t r a n s l a t i o n from Works, No. 37 (57), Central I n s t i t u t e of Prognoses, Hydrometeorological Press, Leningrad, 10 p. Crowe,  R.B.  1958s "Recent temperature f l u c t u a t i o n s and trends f o r the B r i t i s h Columbia coast". Meteorological Branch, Department of Transport, Canada. CIR-3137, TEC-20B. Toronto, 11 p.  Currie, B.W. 1951*s " P r a i r i e Provinces and Northwest T e r r i t o r i e s i c e s o i l temperatures". Physics Department, U n i v e r s i t y of Saskatchewan,  28 p.  Devik, Olaf  191*1*: "Ice formation i n lakes and r i v e r s " . No. 5, pp.193-203.  Geographical Journal 103,  Fraser River Board 1958: Preliminary report on f l o o d control and hydro-electric power i n the Fraser River basin, V i c t o r i a , June 1958, 171 p. Henoch, W.E.S.  1961: "Fort McPherson, N.W.T." Geographical B u l l e t i n No. 16, pp. 86-103. Hoyt,  W.G.  1913:  "The e f f e c t s of i c e on stream flow". U.S.Q.S. Water Supply Paper 337, 77 p. Kendrew, W.G. and C u r r i e , B.W. 1955: The climate of c e n t r a l Canada. Queen's P r i n t e r , Ottawa, 19l* p.  125.  Kennedy, R.E. 1939: "Formation and dissipation of i c e on lakes computed by the - -energy equation". Commission de Limnologie, Association Internationale d'Hydroiogie Scientifique, Reunion de Washington, Tome 1, Question 2, Rapport 2, 15 p.  Kindle, E.M. 1920: "Arrival and departure of winter conditions i n the lyiackenzie River basin". The Geographical Review, Vol. 10, Ko. 6, pp.388-399. Kraus, E.B. 1955: "Secular changes of t r o p i c a l r a i n f a l l regimes". Quarterly Journal of the Royal Meteorological Society, Vol. 81, No. 32*8, pp.198-210. Lloyd, T. 191+3: "The Mackenzie waterway: a northern supply route". Geogrsphical Review, Vol. 33, pp.l4l5-U3l*. Mackay, J . Ross 1961a: "A study of freeze-up and break-up at Fort Good Hope, N.W.T." Thought, Ed. W.G.Dean, W.J. Gage Limited, Toronto, 250 p. 196lb: "Freeze-up and break-up of the lower Mackenzie River, N.W.T." Geology of the A r c t i c , V o l . 2, pp. H19-113U. University of Toronto Press, Toronto, 1196 p. McMullen, D.N. 1961: "Hydrometeorological aspects of the I960 spring break-up i n Southwestern Ontario". Meteorological Branch, Department of Transport. CIR-3l4i;0, TEC-3Wi. Toronto, 11 p. Meteorological Branch, Department of Transport, Canada 1959: "Break-up and freeze-up dates of rivers and lakes i n Canada". CIR-3156, ICE-2. Toronto, 90 p. 1959: "Maximum winter ice thicknesses i n rivers and lakes i n Canada". CIR-3195, ICE-lu Toronto, 20 p. 19ii9: "Temperature and precipitation normals f o r Canadian weather stations based on the period 1921-50". CIR-3208, CLI-19, Toronto, 33 p. Murakami, Masatsugu 1955: "Freezing of the Sunghali River, Manchuria". Snow, Ice and Permafrost Research Establishment, Corps of Engineers, U.ST Army, Translation No. 3h 12 p. 3  126. Parsons, W.J. 1940: "Ice i n the northern streams of the United States". Transactions of the American Geophysical Union, pp.885-893* Shebanov, V.N. 1958: "Long%term prognoses of f a l l and spring ice phenomena i n the estuaries of the Duna, Dvina, Southern Bug, and Dneiper Rivers". Private translation from Works, Ho. 65, Central Institute of Prognoses, Hydrometeorological Press, Moscow, 13 p. Shipman, T.C. 1938: "Ice conditions on the Mississippi River at Davenport, Iowa". Transactions of the American Geophysical Union, Vol. 19, pp. 590-591*. Siegel, Sydney 1956: "Honparametric s t a t i s t i c s f o r the behavioral sciences. McGraw-Hill Book Co. Inc., Toronto, 312 p. Sokolov, S.S. 1955: "Decreasing durations of freeze-up as related to the warming of the climate". Priroda, 1955, u. 96-98, Translated by E.R. Hope, Defence S c i e n t i f i c Information Services, T197B, 3 p. Thomas, J.E.J. 1956: "Saskatchewan River Drainage Basin, 1951-52". Mines Branch, Department of Mines and Technical Surveys, Ottawa, Water Survey Report No. 7* 151+ p. 1958:  "Churchill River and Mississippi River Drainage Basins i n Canada, 1952-51+". Mines Branch, Department of Mines and Technical Surveys, Ottawa, Water Survey Report No. 9, 53 p.  Water Power Branch, Department of the I n t e r i o r , Ottawa. "Reports of hydrometric surveys i n Alberta and Saskatchewan, 1908-19". Water supply b u l l e t i n s , Nos. 1 - 1 1 . Water Resources Branch, Department of Northern A f f a i r s and National Resources, Ottawa. "Arctic and Western Hudson Bay drainage and Mississippi drainage i n Canada". Water resources papers f o r climatic years 1912-55. Williams, J.R. 1955: "Observations of freeze-up and break-up of the Yukon River at Beaver, Alaska". Journal of Glaciology, V o l . 11, pp.2+88-l+95«  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0105663/manifest

Comment

Related Items