ISSN : 0546 - 500 1

GEOLOGY O F

RENVILLE AND WARD COUNTIES

NORTH DAKOTA

by

John P . Blueml e

BULLETIN 50 - PART 1NORTH DAKOTA INDUSTRIAL COMMISSIO N

GEOLOGICAL SURVEY DIVISION

COUNTY GROUNDWATER STUDIES 11 - PART 1NORTH DAKOTA STATE WATER COMMISSIO N

Prepared by the North Dakota Geological Surve yin cooperation with the North Dakota State Water

Commission, the United States Geological Survey, andRenville and Ward Counties Water Management District s

Printed by Quality Printing Service

1989

CONTENTS

Page

ABSTRACT v i

INTRODUCTION 1Purpose 1Previous Work 1Methods of Study 2Acknowledgments 3Regional Topography and Geology 3

STRATIGRAPHY 6General Statement 6Cretaceous and Tertiary Rocks 8Configuration of the Bedrock Surface

1 0Pleistocene Sediment 1 2

Glacial Stratigraphy 1 7Snow School Formation 20Blue Hill Deposits 20Younger Glacial Deposits 2 2

Holocene Sediment 24

GEOMORPHOLOGY 26General Description 2 6Glacial Landforms 28

Collapsed Glacial Topography 28Waterworn Topography 32Sllopewash-Eroded Topography 33

Glacial Lake Landforms 33Fluvial Landforms 3 5

Spillways 3 5Overridden Fluvial Deposits 39Other Glaciofluvial Deposits 39Eskers and Kames 409

PLEISTOCENE FOSSILS 40

GEOLOGIC HISTORY 41

iii

CONTENTS--continued

Page

ECONOMIC GEOLOGY 52Lignite 52Gravel and Sand 53Hydrocarbons 53Potash 54Halite 54

REFERENCES 56

ILLUSTRATION S

Figure

Page1. Physiographic map of North Dakota

showing the location of Renvill eand Ward Counties 5

2. Stratigraphic column for Renvill eand Ward Counties 7

3. Subglacial geology and topograph yof Renville and Ward Counties 1 1

4. Thickness of the Coleharbor Grou pdeposits in Renville and War dCounties 1 3

5. Ice-movement directions and ou tmargins of glacial advances 1 8

6. Surface till units in Renville an dWard Counties 1 9

7. Generalized map of the landform typesin Renville and Ward Counties 27

iv

I LLUSTRATIONS--continued

Figure

Pag e8. Diagrams showing how circular dis -

integration ridges form in areas o fcollapsed glacial topography 3 1

9. Area affected by catastrophic floods 36

10. Maximum extent of Late Wisconsinanice in Renville and Ward Counties .

. . .

43

11. Readvance of the Late Wisconsina nglacier 44

12. Recession of the glacier 45

13. Recession of the glacier 46

14. Readvance and shearing in the glacier

47

15. Final withdrawal of the glacier 48

16. Flooding and the formation o fglacial Lake Souris 49

Table

Page1 . Very coarse sand percentages 1 5

Plate1 . Geologic map of Renville and Ward

Counties (in pocket)

v

ABSTRAC T

Renville and Ward Counties, located in north-centra lNorth Dakota on the northeast side of the Williston Basin ,are underlain by as much as 11,000 feet of Paleozoic ,Mesozoic, and Cenozoic rocks that dip gently to th esouthwest . The uppermost Paleocene rocks, the Cannon -ball, Bullion Creek, and possibly the Sentinel Butt eFormations, lie directly beneath the Pleistocene glacia ldeposits, which overlie most parts of the two counties .Isolated exposures of the Bullion Creek and Cannonbal lFormations occur in the Des Lacs and Souris Rive rvalleys . The glacial deposits in the two-county area rangefrom 0 to 800 feet thick and average about 165 feet thick .

Renville County and about a third of Ward Countyare located on the Glaciated Plains, an area of undulatingto flat topography . The southwestern two-thirds of WardCounty lies within the Missouri Coteau and Coteau Slop epart of the Great Plains . This area is characterized b yhigh-relief collapsed hummocky topography . Associatedlandforms include numerous ice-contact deposits, lak eplains, and collapsed fluvial plains . Surficial depositsthroughout the two counties are predominantly till an dglaciofluvial deposits, but lake sediments, colluvium, an drecent alluvium and landslide deposits are also present .

The most important economic resources in Renvill eand Ward Counties, other than soil and water, are lignite ,sand and gravel, and petroleum . Although no potash o rhalite are currently being produced, beds of thesematerials do underlie the area .

vi

INTRODUCTIO N

Purpose

This report is published by the North DakotaGeological Survey in cooperation with the North Dakot aState Water Commission, the United States Geologica lSurvey, the Renville County Board of Commissioners, an dthe Ward County Water Management District . It is one ofa series of county reports on the geology and ground -water resources of North Dakota . The main purposes ofthese studies are : (1) to provide geologic maps of theareas ; (2) to locate and define aquifers ; (3) to determinethe location and extent of mineral resources in th ecounties ; and (4) to interpret the geologic history of th eareas . This volume describes the geology of Renville an dWarcl Counties . Readers interested in groundwater shoul drefer to Part II of this bulletin (Pettyjohn, 1968), whichincludes detailed basic data on the groundwater, and Par tIII (Pettyjohn and Hutchinson, 1971), which is a descrip-tion and evaluation of the groundwater resources o fRenville and Ward Counties .

Parts of this report that are primarily descriptiv einclude the discussions of the topography, rock, an dsediment in Renville and Ward Counties . This informationis intended for use by anyone interested in the physica lnature of the materials underlying the area . Such peopl emay be water-well drillers or hydrologists concerned abou tthe distribution of sediments that have potential t oproduce usable groundwater ; state and county watermanagers ; consultants to water users ; water users in thedevelopment of groundwater supplies for municipal ,domestic, livestock, irrigation, industrial, and other uses ;civil engineers and contractors interested in such thing sas the gross characteristics of foundation materials atpossible construction sites, criteria for selection an devaluation of waste disposal sites, and the locations ofpossible sources of borrow material for concrete aggre-gate ; industrial concerns looking for possible sources o feconomic minerals ; residents interested in knowing moreabout the area ; and geologists interested in the physica levidence for the geologic interpretations .

Previous Work

Parts of Renville and Ward Counties were studied b yUpham (1895) during his study of glacial Lake Souris .Wilder and Wood (1902) and Wood (1902) made early

studies of lignite in the area ; Leonard (1916) and Tod d(1923) investigated the drainage systems of northwester nNorth Dakota, an area that included Renville and WardCounties . An early study of the geology and groundwate rresources of the two counties was made by Simpso n(1929) . Alden (1932) described the glacial deposits of th earea ; Andrews (1939) published a generalized geologic ma pof southwestern Ward County and evaluated the lignit eresources of the region . The glacial history of NorthDakota and the Souris River Lobe were described b yLemke (1958) and Lemke and Colton (1958) . An exposureof two drift sheets in Ward County was described byLemke and Kaye (1958) . Lemke (1960) provided a com-prehensive report on the geology of Renville and War dCounties (except for the Missouri Coteau area) . Coltonand others (1963) mapped the general glacial features ofNorth Dakota and, later, Clayton and others (1980 )provided a considerably more detailed map of the geolog yof North Dakota . Pettyjohn (1967a, 1967b) described th estagnant ice features and stratigraphy of multiple-layere dglacial deposits on the Missouri Coteau . Kehew presente devidence for Late Wisconsinan catastrophic flooding in theSouris River area, including parts of Renville and War dCounties (Kehew, 1979, 1982 ; Kehew and Clayton, 1983) .

Akin (1947, 1951), Jensen (1962), Armstrong (1963) ,Randich (1963), Schmid (1963), LaRocque and other s(1963a, 1963b), Froelich (1964), Pettyjohn and Hill s(1965), and Pettyjohn (1967c) described the geology o fsmall areas of the counties or provided basic data relate dto groundwater supplies . Kehew (1983) described th egeology and geotechnical conditions in the Minot area .

Several of the nearby county studies in the presen tseries of county geologic studies have been completed .Bluemle (1971, 1982, 1985) reported on the geology o fMcLean, McHenry, and Bottineau Counties . Clayton (1972)described the geology of Mountrail County, and Freer s(1973) described the geology of Burke County .

Methods of Study

Fieldwork for the geologic study of Renville and War dCounties was accomplished over a long period of time b yseveral people . Wayne Pettyjohn mapped the geology o fthe southwestern part of Ward County during the summe rof 1965 . Alan Kehew mapped the 16 townships in th enortheastern quarter of Ward County in the early 1980s .

2

John Bluemle spent part of the field season in 1987checking the geology of Renville County and parts ofWard County .

Field mapping was done on aerial photographs (scal e1 :20,000) and county road maps (1 :126,720) . Test holesdrilled during the early part of the study were logged b yLarry Froelich and Alain Kahail of the North Dakota Stat eWater Commission . Dan Walker and Randy Rickfordoperated the North Dakota Geological Survey auger durin gthe test drilling . Garth Anderson provided data from hi smapping and drilling in the Minot area (results include din Kehew, 1983) . Unpublished geological and groundwate rdata were provided by the North Dakota State Wate rCommission . Test-hole logs and engineering lab test datawere provided by the North Dakota State Highwa yDepartment, the U .S . Army Corps of Engineers, and Soi lExploration Company . Grain-size analyses were made i nthe Hydrologic Laboratory ; fossil wood was identified b ythe Paleontology and Stratigraphy Branch, and th ecarbon-14 age determinations were made by the Isotop eGeology Laboratory--all of the U .S . Geological Survey .

Acknowledgments

II acknowledge the efforts of Wayne A . Pettyjohn, whoinitiated this study in the 1960s and who mapped aconsiderable part of the two-county area at that time .Special thanks are due to Alan Kehew, who mapped asixteen-township area of Ward County in the early 1980s .

Regional Topography and Geolog y

IRenvillle and Ward Counties, in north-central NorthDakota, have a combined area of 2,915 square miles i nTps151 to 164N, and Rs 81 to 89W . The area is locatedbetween 47° 50' 52" North Latitude on the south, 49 °North Latitude (the U .S .-Canadian border) on the north ,100° 58' 12" West Longitude on the east and 102° 14' 2"West Longitude on the west .

The surface deposits in Renville and Ward Countie sconsist almost entirely of Pleistocene sediments of Lat eWisconsinari age overlying deposits of older glaciations an dbedrock of Cretaceous and Tertiary age . The Paleocen eBullion Creek Formation is exposed along the slopes of th eDes Lacs valley and along the Souris River valley ,especially southeast of Burlington . It is also exposed i n

3

the many small tributary valleys bordering the Des Lac sand Souris Rivers, particularly the tributaries that ente rthe larger valleys from the southwest . The Bullion CreekFormation exposures consist of alternating beds of sand ,silt, clay, and lignite . Marine sediments of the Paleocen eCannonball Formation are exposed in road cuts just eas tof Sawyer . The Cretaceous Fox Hills and Hell Cree kFormations, which subcrop beneath the glacial deposits i nthe northeasternmost part of the area, are not exposed i neither county .

The slope on the preglacial bedrock surface i nRenville and Ward Counties is northeastward, generall ydown the bedrock surface beneath the modern Missour iCoteau . The dip of the preglacial bedrock layers i stoward the southwest, reflecting the position of the two -county area on the eastern side of the Williston Basin, acratonic basin, the center of which is in western Nort hDakota .

Renville County is in the Central Lowland Provinc eand Ward County lies in both the Central Lowland an dGreat Plains physiographic provinces . In the report area ,the Central Lowland is subdivided into the Western Lak eDistrict and the Missouri Escarpment District of th eGlaciated Plains . The Great Plains in Ward County consis tof the Missouri Coteau and the Coteau Slope (fig . 1) .

The Western Lake District consists mainly of low -relief plains that slope gently to the northeast . Th eglacial deposits range up to 750 feet thick, possibly 800feet thick in northwestern Renville County . Generally, thethickness of the glacial sediment increases to th enortheast, but the thickest glacial deposits are in th eburied river valleys .

The Missouri Escarpment extends from the Des Lac sand Souris River valleys to the eastern margin of th eMissouri Coteau (fig . 1) . The surface of the escarpment ,which is controlled in part by the underlying bedrock, i sinclined rather steeply to the northeast . The northeast -facing escarpment is an abrupt feature along most of it slength in Ward County, and local relief may exceed 300feet .

The Missouri Coteau is a belt of collapsed glacia ltopography that has nonintegrated drainage and highe relevation than adjacent areas . Locally, bedrock of th eFort Union Group of Paleocene age crops out ; however,the glacial deposits on the Missouri Coteau exceed 250 feetin thickness in places . The southwestern margin of the

4

Figure 1 .

Physiographic map of North Dakota showing the location of

Renville and Ward Counties .

Missouri Coteau separates an area with nonintegrateddrainage (northeast of the margin) from one that ha sintegrated drainage (southwest of the margin) .

Most of the two-county area is drained by the Souri sand Des Lacs Rivers, which nearly bisect the area o f

study . These streams are in the drainage of the Re dRiver of the North, which empties into Hudson Bay . Onl ytwo tributaries of the Missouri River traverse south -westernmost Ward County . The remainder of the area i spoorly drained and is characterized by numerous prairi e

potholes . The northeast-flowing tributaries of the Souri sand Des Lacs Rivers are relatively long and deepl yincised, compared to the southwest-flowing tributaries ,which are generally only a mile or two long . Long ,narrow ice-marginal channels trend parallel to the Souri sRiver in a wide arc and empty into the Souris River i n

Bottineau and McHenry Counties . Most of these channel s

carry only intermittent flow . Local relief northeast of th eMissouri Escarpment is only a few feet and the area i s

poorly drained . An average of nine closed depression s

occur per square mile . These depressions collect runof ffrom small areas and contain water for at least six month s

5

of the year . In addition to the larger depressions ,thousands of small ponds or prairie potholes contain wate rfor only a few days or weeks at a time after a rain or i nthe spring when the snow melts .

Only two waterways on the Missouri Coteau drain intothe Missouri River system . Lakes and intermittent stream slie in an elongate sag that extends from Rice Lakethrough Douglas and into McLean County . During somewet periods, the lakes overflow and may drain into Middl eBranch Douglas Creek .

Several tributaries enter the Hiddenwood Lak edrainage, which is blocked by glacial deposits in McLea nCounty . During the spring and in wet seasons, the lake smay overflow into swampy areas prior to gaining acces sto the Missouri River by way of Deep Water Creek i nMcLean County .

With the exceptions noted above, drainage on th eMissouri Coteau and the Coteau Slope in Ward County i sinterior . Potholes are abundant and collect runoff an dprecipitation throughout the year .

STRATIGRAPH Y

General Statement

Renville and Ward Counties lie within the Willisto nBasin, one of the largest structural troughs in Nort h

'America . The center of the basin is in McKenzie County ,to the southwest of the study area . Rocks near th esurface in the area are flat-lying or dip only a fewdegrees to the southwest into the basin .

Most of the rock and sediment exposed at the surfac ein Renville and Ward Counties was deposited by glacier sduring Pleistocene time . Marine and nonmarine sedimentar yrocks underlie the glacial deposits . Several hundred oi land gas tests drilled in the two-county area indicate tha tthe sedimentary sequence there is between 8,000 an d11,000 feet thick, generally thickest in the western areas .Crystalline rocks of Precambrian age underlie the sedimen-tary rocks ; a total of 12 test holes have penetrated a sdeep as the Precambrian .

The rocks deposited in the Williston Basin during thePaleozoic Era and during Triassic and Jurassic time ,generally consist of evaporites and carbonates interbedde dwith some clastic rocks (fig . 2) . Deposition of clasti c

6

AGE

WUz4.1 UNIT NAME DESCRIPTION THICKNES S

yce

(feet )

Holocene a Oahe Formation Sand, silt, and clay 0-5 0Quaternaryi ole a 'ror

rou Till, sand, gravel, silt, and clay 0-50 0

o Sentinel Butte/ Shale, sandstone, and lignite 0-70 0

TertiaryBullion Cree k

Ludlow/Cannonball Shale, sandstone, marine sand, clay, and 0-36 0

lignit e

Hell Creek Formation Sandstone, shale, and lignite 0-400

ox Hills Formationoxe

Marine sandstone 0-400

Pierre Formation Shale 1,200 3

oNiobrara Formation Calcareous shale

jCarlile Formation Shale 900

Cretaceous . Greenhorn Formation Calcareous shale to

Belle Fourche Fm . Shale 1,500

Mowry Formation Shale 27 0

2 Newcastle Formation Sandstone t o0m Skull Creek Formation Shale 36 0

Inyan Kara Formation Sandstone 2503

Jurassic ndi

erentiated Shale, sandstone, carbonates, and gypsum 700-900—V~

Triassic Spearfish Formation Siltstone and sandstone (redbeds) 180-300

Permian (Absent in Renville-Ward) 0

Pennsylvanian m Amsden and Minneulsa Sandstone, shale, and carbonates0-400

a Fm .

Big Snowy Group Shale, sandstone, and carbonates 0-400

Mississippian Madison Group Limestone, dolomite, and anhydrite 1,200 -1,800

Bakken Formation Siltstone and shale 20-70

N Three Forks Formation Shale, siltstone, and dolomite 100-220

of Birdbear Formation Limestone 80-100

Duperow Formation Dolomite and limestone 300-420

DevonianSouris River Fm . Dolomite and limestone 250-300

Dawson Bay Fm . Dolomite and limestone 100-150

Prairie Formation Anhydrite, halite, and potash 100-500

Winni

gosis Fm . Limestone and dolomite 200-30 0

Interlake Formation Dolomite 500-80 0

Silurian o Stonewall Formation Dolomite and limestone 70-12 0

Stony Mountain Fm . Dolomite, limestone, and shale 120-16 0

Ordovician ,°, Red River Formation Limestone 550-65 0F+

Winni

g Grou Siltstone, sandstone, and shale 200-27 5

Cambrian SAUK Deadwood Formation Limestone, dolomite, shale, and sand 200-55 0

Precambrian basement rocks

Figure 2. Stratigraphic column for Renville and Hard Counties .

7

rocks, however, predominated from the latter part of th ePaleozoic . Throughout the Paleozoic, the Williston Basi nwas at times a cratonic basin or a shelf area tha tbordered a miogeosyncline farther west . At other times ,the area was flooded by seas that followed a troug hextending eastward from the miogeosyncline across th earea of the central Montana uplift and the central Willisto nBasin . At times, the entire Williston Basin area was abov esea level and subjected to subaerial erosion for relativel yshort intervals of time .

The oldest rocks that subcrop directly beneath th eglacial deposits are shales of Cretaceous age . During theCretaceous Period, shale and sandstone were the principa ldeposits throughout the Williston Basin . These sedimentswere deposited in a huge epicontinental seaway tha textended southward through Canada and the wester nconterminous United States . During the Laramide orogeny ,in Late Cretaceous and early in Tertiary time, the areaof the Williston Basin was, for the most part, above sealevel and continental deposits of sandstone, shale, an dlignite were deposited . However, early in the TertiaryPeriod, a narrow arm of a sea did exist in some parts o fthe basin and marine sediments of the Cannonball Forma-tion were deposited .

Cretaceous and Tertiary Rocks

The Cretaceous has traditionally been subdivided intothree main groups : the Dakota Group, the ColoradoGroup, and the Montana Group . The lowermost of thes eis the Dakota, which includes the Inyan Kara, Skul lCreek, Newcastle, and Mowry Formations . The Inyan KaraFormation is an interbedded shale and sandstone unit ,which in Renville and Ward Counties averages about 25 0feet thick . The lower part of the Inyan Kara is nonmarin ein origin, while the upper part of the formation is partlymarine . Drillers commonly report a "first artesian flow"and a "second artesian flow," which are separated by apredominantly shale section .

The sandy marginal marine characteristics of th eupper Inyan Kara are overlapped by siltstone and shal eof the Skull Creek Formation . The overlying Mowr yFormation consists of dark-gray to black shale and i ssimilar to the Skull Creek . The Mowry and Skull Cree kgenerally range from 220 to 360 feet thick in Renville an dWard Counties . Locally, a lenticular sandy zone is presen t

8

at the base of the Mowry . Where present, this sandstoneis called the Newcastle Formation .

The Colorado Group, of Late Cretaceous age, consist sof as much as 900 to 1,050 feet of light-gray to dark -gray, limy shale that overlies the Dakota Group . It isdivided into four formations (in ascending order) : th eBelle Fourche, Greenhorn, Carlile, and Niobrara .

The Colorado Group is overlain by the Montan aGroup, which consists of about 1,200 feet of gray, locall ybentonitic, Pierre Formation shale, and more than 400 fee tof fine-grained clastics of the Fox Hills Formation . Th eMontana Group shales, in turn, may be overlain by a smuch as 400 feet of clastic sediments of the Hell Cree kFormation .

The Fort Union Group was deposited during earlyTertiary time . The continental and marine beds of th eFort Union Group are commonly divided into the Cannon -ball, Ludlow, Bullion Creek, and Sentinel Butte Forma-tions . The Cannonball and Ludlow Formations, which for mthe basal part of the Fort Union Group, were deposite dcontemporaneously--the Cannonball as a marine deposi tand the Ludlow as the equivalent continental strata . Th eCannonball Formation crops out along the walls of th eSouris River valley in the vicinity of Sawyer, where th eformation consists of sandy shale . Elsewhere, the Cannon -ball generally consists of alternating beds of sand an dshale . The Ludlow Formation is not exposed in Renvill eand Ward Counties, although it underlies the area . Th eLudlow generally consists of interbedded fine sand an dlignitic shale where it is exposed in other parts of Nort hDakota . The Cannonball and Ludlow Formations probabl yinterfinger in the subsurface of the report area as the ydo in other parts of the Dakotas . Consequently, thes eformations are difficult to differentiate by drill-hol ecuttings . The combined thickness of the Cannonball an dLudlow Formations in Renville and Ward Counties is abou t360 feet .

The Bullion Creek and Sentinel Butte Formations i nthe two-county area consist of about 700 feet of inter -bedded shale, siltstone, sandstone, and lignite in thewestern part of the area, but they thin to almost nothin gwhere the Cannonball Formation crops out near Sawyer .The sediments were deposited in rivers, streams, swamps ,and shallow lakes on an alluvial plain . The Bullion Cree kFormation crops out at many places along the Des Lac sRiver valley and along the Souris River valley belo w

9

Burlington (pl . 1) . On the Missouri Escarpment south ofSawyer, the glacial deposits thin appreciably and ligniteseams interbedded with sandstone and siltstone areexposed .

After the Sentinel Butte Formation was deposited andbefore the first period of glaciation, the Fort Union Grou pwas gently warped . Erosion removed the Sentinel Butt eFormation and much of the Bullion Creek Formation in th ecentral and northern parts of the two counties andproduced a near-badlands topography throughout th earea . This rugged topography was modified considerabl yby subsequent glacial action .

It is generally difficult or impossible to distinguis hthe various formations in the Fort Union by examining th etest-hole cuttings . In fact, the rocks from the base o fthe glacial sediments, downward to the top of the Pierr eFormation, are so similar that even tentative correlation sare difficult . The Ludlow Formation appears to underli ethe Cannonball Formation in the Carpio area, bu telsewhere the Cannonball beds intertongue with th eLudlow and Bullion Creek Formations . The Fort UnionGroup has a total thickness of about 615 feet in th eCarpio area where the Ludlow and Cannonball have a tota lthickness of 360 feet and the Bullion Creek is about 255feet thick .

Configuration of the Bedrock Surfac e

The topography on the bedrock surface underlyin gthe glacial deposits in Renville and Ward Counties isshown on figure 3 . This generalized map is based mainl yon test-hole data, although a few outcrops of bedroc kprovided additional control . In general, the erodedpreglacial surface sloped toward the northeast at ap-proximately 16 feet in a mile . The valleys of the Des Lac sand Souris Rivers are deeply incised on the regional slop eas are the glacial diversion trenches .

Southwest of the Des Lacs River and near the lowe rSouris River, downstream from Burlington, the regiona lslope is considerably greater than to the northeast . Thisslope, which is particularly steep in the vicinity ofKenmare (T160N, R88W), reflects the more abrupt rise ofthe preglacial elevation of the Missouri Escarpment an dthe Missouri Coteau along the Coteau Slope . Th enortheastward slope there is 50 feet or more in a mile .The remainder of the two-county area has a bedroc k

10

R . 87 W7~0

R . 84 W.T. 164 N .

T.16I N .

T.I59N .

R.83 W . R .81 Wb

T. I58 N .

115I N .

00

Figure 3 . Subglacial geology and topography of Renville and WardCounties . The highest bedrock elevations and younges tformations occur in the western part of the area, whichslopes generally northeastward . Kf = Fox Hills Fm . ; Kh = Hel lCreek Fm . ; Tc = Cannonball Fm . ; Tb = Bullion Creek Fm . ; Is =Sentinel Butte Fm .

11

surface that is flat to gently sloping at about 13 feet pe rmile to the northeast . The Des Lacs and lower SourisRivers have eroded narrow, steep walls nearly 200 feetbelow the slope of the regional bedrock surface .

Major bedrock topographic features probably wereformed by Tertiary weathering and erosion . In a fewplaces, particularly on the Missouri Coteau, a weathere dzone on the bedrock was penetrated by test holes .Throughout most of the area, however, glacial erosio nstripped off most or all of the Tertiary soil and weatheredzone, and modified the topographic features, including th eexisting drainage system .

Pleistocene Sedimen t

Renville and Ward Counties are covered by glacia land glaciofluvial sediment, except in places where th eFort Union Group sandstones and shales (mainly th eBullion Creek Formation) crop out in a few small and loca lexposures . The thickness of the glacial deposits generall yranges from absent at the bedrock outcrops to about 400feet (fig . 4) . The greatest thickness of glacial deposit spenetrated in the two-county area was 800 feet in a tes thole in section 27, T162N, R87W, about 12 miles northeas tof Kenmare . This test hole was drilled after the basic -data report was published (Pettyjohn, 1968) . Another testhole in the same section penetrated 655 feet of glacie rsediment . Test holes drilled in sections 4 and 9, T155N ,R83W, penetrated 432, 430, and 422 feet of glacia lsediment . In most places, however, the glacial deposit sare fairly uniform in thickness, except where the majo rstreams and meltwater valleys have removed them . Th eaverage thickness of the glacial deposits penetrated b y178 test holes drilled during the test-drilling program wa s165 feet .

All the sediment related to glacial deposition i nRenville and Ward Counties, that is, all the materials thatwere deposited by the glacial ice as well as by flowin gand ponded water associated with the ice, are collectivel yreferred to as the Coleharbor Group . The ColeharborGroup has been subdivided into a large number ofinformal units and formal formations by various geologists .However, most of the detailed stratigraphic work has bee ndone in eastern North Dakota where the glacial sediment swere deposited by glaciers that advanced southward, u pthe Red River Valley (the Des Moines Lobe) and aroun d

12

Figure 4 . Thickness of the Coleharbor Group deposits in Renville andWard Counties . The thickest glacial deposits are in thenorthwest part of the area, in a buried valley, where morethan 600 feet of sediments occur ; one test hole in T162N ,R87W reported over 800 feet of sediment . The shaded area srepresent discontinuous glacial deposits with numerou sexposures of pre-Coleharbor formations . Thicknesses given infeet .

13

the east side of the Turtle Mountains (the Leeds Lobe) .The glacial sediments in Renville and Ward Counties wer edeposited by glaciers that advanced around the west sideof the Turtle Mountains (the Souris Lobe), and althoug hsome stratigraphic work has been done along Lak eSakakawea in McLean County (Ulmer and Sackreiter ,1973), the formal glacial stratigraphy in this part ofNorth Dakota is not yet well understood . Some of th eformal units that have been described for the easternNorth Dakota glacial sediments may be regionally cor-relatable with units in Renville and Ward Counties .However,, since the present study did not include an yspecific effort to formally describe the discrete units o fglacial sediment that can be attributed to specific units o fdiffering ages, no attempt has been made to compare th eglacial stratigraphy in the two-county area with that i nany other area .

Kehew (1983) studied the geology and geotechnica lconditions in a 16-township area of eastern Ward Countythat includes the Minot area and he later studied a par tof southwest Ward County (Kehew, 1984) . Much of th ediscussion of till stratigraphy included in this repor tcomes from his studies . Kehew ' s are the only studies thathave dealt in any detail with the specific physical charac-teristics of the glacial till units . One of Kehew' s objec-tives was to characterize and differentiate till units usin gquantitative lithologic parameters (table 1) and to relat etills to ice-marginal positions representing specific glacia ladvances . Differences in lithology are to be expected i ntills deposited during glacial advances of different age sand from ice moving in different directions . The methodsused for till characterization by Kehew included hand -specimen appearance, field stratigraphic relationships ,weight percentages of sand, silt, and clay fractions of th etill matrix, and lithology of the very coarse-sand fractio n(1-2 mm) .

Laboratory grain-size analyses were done by standardsieve and hydrometer methods . The specific laborator yprocedure used by the North Dakota Geological Survey fo rtill samples is described by Perkins (1977) . The verycoarse-sand lithology is determined by counting individua lgrains under the binocular microscope . These weredivided into groups including Precambrian igneous an dmetamorphic, carbonate, shale, lignite, secondary precipi-tates, and miscellaneous categories . Percentages of lignite ,secondary precipitates, and miscellaneous groups wer e

14

Table 1 . Very Coarse Sand Percentage s

1g/Met . Carb . Sh . Lig .

Unit SD x SD x SD x SD

A 63 68 .5 4 .3 28 .3 7 .3 3 .3 1 .4 0 0

B 94 74 .3 4 .1 21 .3 3 .8 4 .3 2 .5 3 .8 6 . 2

Martin 4 49 .9 5 .5 28 .7 4 .3 21 .4 3 .9 4 .1 2 . 2

v Ryder 6 34 .6 3 .6 30 .1 4 .1 35 .5 5 .4 0 .5 1 . 2

Makoti 13 40 .9 8 .2 22 .8 2 .4 36 .3 9 .1 7 .7 5 . 6

Blue Mtn . 22 33 .4 5 .6 24 .1 4 .1 42 .6 7 .4 0 .3 1 . 1

Snow School 6 43 .6 5 .0 35 .0 3 .0 21 .4 3 .2 * *

* - abundant lignite in sampl e

based on the total grain count . Percentages of igneou sand metamorphic, carbonate, and shale grains werecalculated by normalizing the sum of these groups as 10 0percent . Finally, the ratio of igneous and metamorphic t oigneous and metamorphic plus carbonate grains wa scalculated . Of the two laboratory techniques, grain-sizedistribution is most useful for tills of vastly differen tprovenances . This technique was of limited usefulness fo rdifferentiating tills in Ward County . Very coarse-san dlithology has proven to be an effective stratigraphic toolin localized stratigraphic studies in areas such a snortheastern North Dakota (Harris and others, 1974 ;Hobbs, 1975) and has been used as an aid to regiona lstratigraphic correlation (Moran and others, 1976) . I nWard County as well, coarse-sand lithology is an effectiv estratigraphic tool .

The very coarse sand grain counts indicate acombination of local and distant sources . Igneous an dmetamorphic grains include Canadian Shield Precambria nsources as well as local sources derived from Fort Unio nbedrock . Carbonate grains are derived from distan texposures of Paleozoic limestone and dolomite formations .The shale content of the till is probably derived mainl yfrom Cretaceous marine shales exposed east and northeas tof central North Dakota . Some grains derived from For tUnion bedrock may be included in this fraction . Th esource of the lignite is the Paleocene Fort Union Group .The relative proportions of the various very coarse-san dgroups give a relative indication of the direction of glacia ladvance . For example, tills with high shale content wer eprobably derived from glaciers advancing from the east o rnortheast . In contrast, tills low in shale must have bee ndeposited by south- or southeast-moving glaciers .

The most current compilation of ice-marginal correla-tions in North Dakota (Clayton and others, 1980) indicate s13 ice-marginal positions in the state . Although age datin gof most North Dakota glacial deposits is lacking, itappears that most of the ice margins represent advance sand readvances of the Late Wisconsinan Laurentide ic esheet . During the late phases of Late Wisconsinan deglaci-ation, glacial advances and readvances occurred by th emovement of thin ice lobes into low-lying areas . Two suchlobes may have flowed southward around the TurtleMountains into North Dakota, the Leeds Lobe east of theTurtle Mountains, and the Souris Lobe around the wes tside . Renville and Ward Counties lie entirely within the

16

extent of the Souris Lobe . Several ice-marginal position scross the two-county area (Clayton and others, 1980)and, consequently, several different till units occur i nthe area (fig . 5) .

Samples for laboratory analysis were taken fro msurface exposures and auger holes drilled during th estudy . Auger samples included samples brought to th esurface during auger rotation as well as split-barre lpenetration samples . The maximum depth reached wa susually about 75 feet . Therefore, only the upper portio nof the glacial section could be studied by this method .Some samples obtained during drilling for the count ygroundwater study were analyzed . These rotary-drilledsamples were often contaminated by overlying material ,particularly in places where a sand layer was penetrated .

Glacial Stratigraph y

Preliminary studies of the stratigraphy of the glacia ldeposits of the Coleharbor Group have been restricted tothe eastern part of Ward County and to southwester nWard County (Kehew, 1983, 1984) . No studies have bee nconducted in Renville County, but the results of theeastern Ward County area should be generally applicabl eto Renville County because all of Renville County lie swithin--behind--ice margin 13, the most recent margi nrecognized in North Dakota (fig . 5) (Clayton and others ,1980) . In southwestern Ward County, at least threediscrete layers of glacial sediment can be recognized .Each glacial advance resulted in a new layer of sediment ,but during each subsequent advance, much of the olde rglacial material probably was incorporated into th eadvancing glacier and redistributed . The separate layersof glacial deposits are lithologically similar, but they ca nbe distinguished by the presence of buried oxidizedzones, boulder pavements, and widespread buried san dand gravel deposits . Kehew (1984) found that the olde rtills can be differentiated by a study of their verycoarse-sand lithology .

As the glaciers advanced and retreated, the under -lying bedrock surface was eroded and subdued, and i nmost areas buried under thick deposits of glacial sedi-ment . Much of the glacial ice that covered the Missour iCoteau stagnated there and left characteristic topographi cfeatures that will be discussed later .

17

l

` . ♦ `

/

,♦dot `

My {

-1 t

1 J

3

CUE

{

')y

r\y

\\~

`hq~`

33 Z ` •

B' mor

R . 87 W .

R, 84 W .

T.164 N

R . EIS W

T .161 N .

UNIT A

T.158 N .

81 w .

CO .

A

Blu eHill

T.151 N

Figure 6 . Surface till units in Renville and Ward Counties . The oldes ttill unit, the Snow School, occurs only at the southern edg eof the area and is probably early Wisconsinan in age . TheBlue Hill, Makoti, and Ryder units are all probably Lat e

IWisconsinan in age, as are Units A and B .

19

Snow School Formatio n

The oldest glacial deposits exposed in the two-count yarea cover a small area in the south-central part of War dCounty, in the southeast part of T151N, R85W (fig . 6) .This small area of collapsed glacial topography is bounde don the west by the Blue Hill moraine and on the nort hand east by younger glacial deposits and by outwash . Theolder glacial deposits cover wide areas to the south i nMcLean County where they were identified as belonging t othe early Wisconsinan Napoleon Glaciation (Bluemle, 1971) .Clayton and Moran (1982) correlated the Napoleon deposit sin this area with the Snow School Formation (Ulmer an dSackreiter, 1973) . This formally named unit will be usedhere to refer to these deposits . The Snow School deposit swere penetrated beneath the Makoti moraine at a depth o f173 to 190 feet in a test hole in section 15, T151N, R87W .Here, they lie directly on the Sentinel Butte Formationand are overlain by stream gravel and younger glacia ldeposits .

The Snow School glacial deposits consist of silty an dsandy till that is relatively high in carbonates with a nabundance of lignite chips (table 1) . Pebbles in the til lare coated with manganese oxide . Well data indicate tha tthe entire thickness of the till, which may be as much a s30 feet, is oxidized .

The undulating surface of the Snow School depositsprobably reflect the topography of the underlying pregla-cial surface . Drainage on the Snow School Formatio ndeposits is integrated .

Blue Hill Deposits

An area in southwestern Ward County is surfaced b yLate Wisconsinan glacial deposits that have been informall ycalled " Blue Mountain " (Colton and others, 1963 ;Petiyjohn,. 1967a) or "Blue Hill " (Clayton an'd others ,1980) . The deposits will be referred to here as Blue Hil l(fig . 6) . In the extreme southwestern part of War dCounty, where the Blue Hill deposits are at the surface ,the entire thickness is oxidized to shades of red, yellow ,and brown . The till is generally silty to sandy an dslightly cohesive with a relatively high percentage o fshale (table 1) . Associated surficial sand and gravel i scommonly a poorly sorted mixture of sand and angular tosub,angular gravel composed of carbonates and chips of

20

shale and lignite . Lake silts and sands that occur i nassociation with the Blue Hill deposits are found mainly o nelevated lake plains .

Pettyjjohn (1967a) considered the Blue Hill deposits o fsouthwestern Ward County to be of Early Wisconsinan age .This is unlikely in view of the fact that stagnatio nfeatures present on the Blue Hill glacial deposits form a ninterlocking pattern with similar stagnation features all th eway to the Missouri Escarpment . Such an interlockingpattern could result only if a continuous sheet of stagnan tice covered the Missouri Coteau and melted everywher eover that area at about the same time . Pettyjohn state sthat the Blue Hill deposits are oxidized to an averag edepth of 26 feet and that he has traced this oxidized zon enearly 40 miles north and more than 30 miles east of theBlue Hill moraine in Ward County . However, it is morelikely that he was actually drilling into the oxidized zoneon top of the Snow School Formation ; certainly there i sno reason to believe the oxidized zone is necessaril ydeveloped entirely on the Blue Hill deposits . Theunusually thick oxidized zone (44 feet) on the Blue Hil ldeposits in southwestern Ward County is probably acomposite of oxidized Blue Hill deposits lying on oxidizedSnow School deposits .

The Blue Hill deposits crop out along the valley wall sof the Des Lacs River in the Donnybrook area and alon gthe Souris River valley from Minot nearly to Verendrye i nMcHenry County . Presumably, the oxidized material i nthe upper part of the Blue Hill deposits in the area was ,in most places, incorporated into glacial ice during th esubsequent advance of the Souris Lobe .

Lemke and Kaye (1958) described an exposure of tw oglacial drift sheets near Donnybrook . In that area, th eolder layer of glacial sediment, the Blue Hill deposits ,consists of till and associated glaciofluvial deposits thatare at least 30 feet thick . Although the upper and lowe rlayers are lithologically similar, the older material show sa greater degree of oxidation . The two deposits ar eseparated by a boulder pavement that is generally lessthan 3 feet thick and that grades laterally into a lime -rich zone . The younger layer of glacial sediment is a smuch as 50 feet thick . Along both walls of the Souri sRiver valley and along several of its tributaries, the olde rmaterial is overlain by 10 to 80 feet of younger till .

21

Younger Glacial Deposits

The Blue Hill glacial deposits are overlain in most o fRenville and Ward Counties by younger glacial till thattends to be somewhat more pebbly and bouldery than th eunderlying Blue Hill deposits . These materials wer edeposited by several minor pulsations of the Souris Lob eover a relatively short period of Late Wisconsinan time .They include materials that were deposited by glacie rlobes that have been variously referred to as Marti n(Lemke, 1960 ; Colton and others, 1963 ; Clayton, 1966 ;Pettyjohn, 1967a ; Bluemle and others, 1967 ; and Claytonand others, 1980), Makoti and Ryder (Pettyjohn, 1967a) ,and Minot (Clayton and others, 1980) . These LateWisconsinan glacial deposits cover the surface over mos tof the two-county area .

The relatively few samples that have been analyzedfrom the Makoti, Ryder, and Martin tills suggest thatthey can be differentiated by a study of their verycoarse-sand lithology (table 1) . Oxidized zones were notpresent at the tops of the Makoti and Ryder tills whenthese tills were encountered in the subsurface beneat hyounger tills . Contacts between these till units arecommonly marked by boulder concentrations .

Kehew (1983 ; 1984), in a study of eastern War dCounty, was able to recognize two stratigraphic units inthe upper portion of the till section . Laboratory data fo rthese units are shown in table 1 . Unit A, the uppermos tunit, is assumed to be the till deposited by the glacierthat advanced to ice margin 13 . Unit B consists ofsurface till samples beyond (south and west of) margin 1 3and samples of the till below unit A behind (north an deast of) margin 13 . The contact between units A and Bin the subsurface is usually recognized by the presenc eof a sand lens or boulder concentration between the tw ounits . Although units A and B are similar in texture,small but consistent differences were noted in the twotills . Unit B is higher in igneous and metamorphic grains ,lower in carbonate, and slightly higher in shale . Unit Balso contains more lignite . The lignite percentage cannotbe used as a precise indicator because of the tendency o flignite fragments to easily break apart during sieving an dother laboratory procedures . Lignite is useful as ageneral indicator of till lithology and the difference i nlignite content between units A and B is real although th epercentages are not accurate . Although unit A doe s

22

contain

lignite,

the percentages were less

than 0 .5 an dthey are therefore rounded off to zero on table 1 .

The basic similarity between units A and

B in bothtexture and coarse-sand lithology suggests that unit Arepresents a readvance of the same glacier that depositedunit B after only a short retreat . Unit B is correlatedwith ice margin 11 or 12 (Clayton and others, 1980) .These margins represent the maximum advance of th eSouris Lobe in North Dakota . Ice margin 13, correlatedwith unit A, forms a lobate shape behind margins 11 an d12, indicating a readvance of the Souris Lobe .

One explanation for the differences in very coarse -sand lithology between units A and B can be formulate dfrom the stratigraphic relations in the Minot area .Exposures near the bottom of the Souris River valley nea rMinot indicate that unit B commonly rests directly upo nbedrock . This suggests that, as the Souris Lobe wa sadvancing to margins 11 and 12, it incorporated materia lfrom the local bedrock into the till . Sand and lignite fro mthe Fort Union Group increased the crystalline and lignit epercentages in the very coarse-sand fraction and dilute dthe carbonate content to some extent . After retreat of theSouris Lobe, the glacier began to advance to margin 13 .During this advance, the glacier would have travelle dover the thick till of unit B and therefore it would no thave been influenced as much by local bedrock . Conse-quently, unit A is lower in crystalline and lignite grain sand slightly higher in carbonate content .

In several places in the Souris and Des Lacs valleys ,small amounts of older till crop out above bedrock . Theseexposures are common in the northwestern part of War dCounty in the vicinity of Donnybrook (see previou sdiscussion) . Average values of nine samples from thes eolder tills are shown in table 1 for comparison with unit sA and B . While textures again are not significantl ydifferent, a major difference in shale content is apparent .The older tills contain much more shale in the verycoarse-sand fraction . The high variation in the older tillsindicates that more than one till is present . These till shave not been differentiated by the laboratory studies ,but we do have a reasonable idea of the overall regiona lrelationships (see previous discussion of Napoleon an dBlue Hill tills) . The shale percentages do suggest that th eadvances that deposited the older tills (at least some o fthe older tills) came from the east and northeast, wher ethey obtained a high content of shale from the Cretaceou s

23

bedrock formations . During deglaciation, the ice sheetthinned and became segmented into topographicall ycontrolled lobes near the ice margin (Bluemle, 1965) . A sthe lobes flowed around topographic obstacles, theyadvanced over different bedrock formations and thereforedeposited lithologically different tills . The Souris Lobeflowed around the west side of the Turtle Mountains anddeposited the shale-deficient tills of units A and B in th eRenville-Ward County area . These tills contrast with thoseof the Leeds Lobe, which flowed around the east side o fthe Turtle Mountains and deposited shale-rich tills(Hobbs, 1975; Hobbs and Bluemle, 1987) .

Holocene Sedimen t

The Holocene sediment in Renville and Ward Countiesis designated Oahe Formation . The Oahe Formatio nincludes material deposited during Holocene time, thegeologic time period beginning at the end of glaciatio nabout 10,000 years ago, and continuing until the present .The Oahe Formation consists largely of organic clay an dsilt deposited in sloughs and in shallow channels erode dduring deglaciation . This sediment, which may overli esand and gravel of the Coleharbor Group, was deposite dby Holocene streams, intermittent runoff from valle ysides, and wind . The Oahe Formation deposits ar egenerally thin and confined to valley and slough bottoms .

The most extensive Oahe Formation deposits includ eHolocene alluvium deposited by the Souris and Des Lac sRivers and their major tributaries (map unit Qor) . Thi sunit is restricted to valleys that contain a recognizabl eflood plain . The sediment is dark, fossiliferous, obscurelybedded material of clay to sand size representing bot hchannel and overbank deposition . The valleys of theSouris and Des Lacs Rivers contain a maximum of 150 fee tof this type of sediment in combination with Coleharbo rGroup fluvial sediment . As is typical with fluvial environ -ments, the surface and subsurface sediment is highl yvariable both laterally and vertically . Some test holes i nthe valleys indicate the presence of more uniform fine -grained sediment suggesting that shallow lakes occupie dportions of the valleys from time to time . An example i sthe Souris River valley upstream from its confluence wit hthe Des Lacs River . Test holes drilled for the Burlingto nDam project and for Kehew's study of the geology an dgeotechnical conditions in the Minot area (Kehew, 1983 )

24

penetrated thick silty clay deposits of possible lacustrin eorigin .

Some sediment was deposited in the Souris and De sLacs valleys by intermittent streams flowing in gullies o rcoulees tributary to the main valleys . These fan-likedeposits formed where the tributary coulee meets the mai nvalley . The sediment is generally coarser in grain siz ethan that found beneath the flood plains . This type ofdeposit constitutes a significant part of the thick alluvia lvalley fill in the Souris and Des Lacs valleys . As themodern rivers meander laterally across their flood plains ,they can erode and rework alluvial fan sediment . Alluvia lfan deposition, on the other hand, from large coulees ,can force the river toward the opposite side of the valleyand occasionally dam the main valley . The Souris Rivermay have been dammed intermittently during Holocene tim ein this manner .

Fine-grained, organic-rich sediment in sloughs i smapped as Qos (slough sediments) . Closed depression s(sloughs) occur in collapsed glacial topography as a resul tof the uneven subsidence of supraglacial debris as th estagnant glacial ice beneath slowly melted . Runoff fromsurrounding higher ground transports sediment to theslough basin . Additional sediment is contributed by windand by the decomposition of the abundant vegetation ,which grows in the wet environment .

The last Oahe Formation unit, Qol, consists oflandslide deposits involving material from the Coleharbo rand Fort Union Groups . Landslides have occurred alon gthe valley sides, especially where the bedrock-glacia lcontact is close to the surface . Most of the landslidesoccurred during or soon after the rapid erosion of th eSouris and Des Lacs valleys as the bedrock materials wer esubjected to a loss of lateral support . Because thegeologic history of the Fort Union Group sediments ha sresulted in a condition known as overconsolidation ,exposure during erosion or excavation causes rebound orvolume expansion of the sediments with accompanying los sof strength . Other factors, such as bedding composed o falternating coarse- and fine-grained sediments and aspect sof groundwater flow, influence the stability of slopes i nthe Fort Union Group . In the Des Lacs valley, in th ewestern part of the area, the bedrock-till contact occurswell above the present valley floor . Abundant landslidesand slumps, ranging from old to presently active ,characterize the valley and tributary slopes . Smalle r

25

landslides in till are rare, but they do occur in places .The landslides mapped commonly consist of paralle l

slump ridges and slump blocks . Material within the slum pblocks remained mostly intact and relatively undisturbed .With several exceptions, landslides in the area appear tobe inactive . Cuts or excavations in these deposits coul dcause reactivation of old planes of failure .

GEOMORPHOLOG Y

General Descriptio n

The modern landscape in Renville and Ward Countie swas formed by the Wisconsinan glacier that covered th earea, by meltwater that flooded the land when hugeglacial lakes to the north periodically burst, and bymeltwater that carved the valleys of the Souris and De sLacs River valleys . The northeastern two-thirds of th etwo-county area--the part of the area northeast of th eMissouri Escarpment--is primarily low-relief collapsedglacial topography (fig . 7) . Over much of the norther npart of this area large numbers of low-relief, ring-shapedhummocks ( " doughnuts") can be seen on aerial photo -graphs . The Des Lacs and Souris Rivers have carved100- to 200-foot-deep valleys across the two counties ,through the glacial deposits, and into the underlyin gTertiary rocks . As a result, the Paleocene Bullion Cree kFormation is extensively exposed along the valley walls ,especially along the Des Lacs River valley . To the east ofthe Souris River, repeated floods of glacial meltwate rscoured the surface extensively (Kehew, 1982) whe nglacial Lake Regina, to the northwest of the two-count yarea, drained repeatedly, releasing huge volumes of wate rin short periods of time .

The Missouri Escarpment, a 200- to 300-foot-high ris eto the southwest, separates the Glaciated Plains from th eMissouri Coteau, an area of hummocky, irregular plain sthat resulted primarily from collapse of superglacia lsediment . Areas of collapsed fluvial sediment are found i nplaces on the Missouri Coteau, along with collapsed lak eplains . Relief is highest in the areas of high-reliefcollapsed glacial topography . It is less pronounced , i nareas of rolling collapsed glacial topography, but i ngeneral, the glacial topography southwest of the Missour iEscarpment is considerably more rugged than is the relie f

26

R .87W

R. 84 W .

erode d

lapsed' tLY

Abundant

Lo wglac

Abund ashaped

°

O

T.I64 N .

T .161 N .

0

T. 159 N .

R .81 WR .83 W%

o e

o

,

?r o

Pam ' RENVIL E CO .`

9io

WARD CO .

111''Ol~ \P~ .

% x P IWL\71M

x x•

~•x

x

x

~ 9

x x x x

,90 #kti!fx

X

x

% ,x

x

x

x

xx

x

x

x

Rolling collapse

VIP'

Figure 7 .

Generalized map of the landform types in Renville and Hard

Counties .

shaped mmocks ° 0

XX

X

rfac escoure d

T.158 N .

glacialT .151 N

R . 88 W

27

over the areas of collapsed topography northeast of th eescarpment .

Glacial Landforms

Collapsed Glacial Topography

Hilly and hummocky glacial topography results fro mthe lateral movement of supraglacial sediment as i tsubsides (collapses, is let down, or slides to lowe relevations in the form of mudflows) when the underlyin gice melts out from under it (Clayton, 1967 ; Deal, 1971 ;Clayton and Moran, 1974 ; Clayton, Moran, and Bluemle ,1980) . Although this is the generally accepted explanationfor the origin of hummocky glacial topography, twoalternatives have been suggested . Stalker (1960) sug-gested that hummocks resulted from the squeezing ofsubglacial sediment into irregularities in the base of astagnant glacier . However, hummocks composed of glacia lsediment are essentially identical to hummocks compose dof collapsed supraglacial fluvial and lacustrine sedimen tthat lacks evidence of ever having been under a glacier .Bik (1967) suggested that hummocks resulted from th emovement of sediment during the growth and decay o fpermafrost ; he considered them to be relict pingos . I nNorth Dakota, hummocks were generally formed at a tim ewhen paleoecologic evidence indicates a climate too war mfor permafrost, and hummocks are generally absent i nNorth Dakota in areas known to have had permafros t(Clayton, Moran, and Bluemle, 1980) .

In Renville and Ward Counties, collapsed glacia ltopography occurs in two different settings : on th eglaciated plains northeast of the Missouri Escarpment, an don the Missouri Coteau . Much of the area of the GlaciatedPlains northeast of the Missouri Escarpment to th einternational boundary is mainly nearly flat-surfacedcollapsed glacial topography that has been referred to a sground moraine (Qccu on pl . 1) . It is the most extensiv etopographic unit in the two-county area . This area ha slow relief, generally less than 10 feet locally, poorly t omoderately well integrated drainage, and numerous small ,shallow undrained depressions . Maximum slope angles ar egenerally less than 4° . Many of the shallow, undraineddepressions are rimmed by low, indistinct, circula rdisintegration ridges that are best seen on aerial photo-graphs . South of Sawyer several transverse ridges

28

("washboards " or "washboard moraines " ) occur . Thes eare especially well developed south and southeast of Velvain McHenry County (Bluemle, 1982) . The surface has beenwashed in places, especially in the eastern part of th earea, by running water--by periodic floods from rapidl ydraining glacier-dammed lakes (Kehew, 1982) . These areasare designated Qces on plate 1 . The numerous, smal lchannel scars indicate that the water flowed mainl yeastward or southeastward over the surface in this area .In other places, conspicuous, although low-relief, ring-shaped hummocks can be seen on airphotos ; much ofRenville County is marked by abundant hummocks tha tare, however, not apparent to the observer on th eground . Areas with abundant hummocks are shown wit ha ring pattern on plate 1 (Qccu) .

The boundary between the hilly collapsed glacialtopography of the Missouri Coteau and the undulatin gcollapsed glacial topography of the Glaciated Plains i sabrupt in many places . In other places, though, th echange is gradational and is most easily mapped on th ebasis of the presence or absence of integrated drainage ;drainage on the Missouri Coteau is unintegrated wherea soff the Coteau, it is integrated . The abrupt nature of th eboundary probably corresponds to bedrock escarpments ,knobs, or hills, and where the boundary is indistinct th ebedrock surface tends to rise more gently .

The thickness of glacial sediment in areas of undulat-ing collapsed glacial topography ranges from 0 to (possi-bly) more than 800 feet . It is thin near the Missour iEscarpment in the southeastern part of Ward County an dthin or absent along the walls of the Souris and Des Lac sRivers .

Hilly collapsed glacial topography (Qcch on pl . 1) i swell developed on the Missouri Coteau where local reliefmay exceed 100 feet and slope angles greater than 7° ar ecommon . Through drainage is essentially lacking on th eMissouri Coteau . The landforms in much of the south -western third of the two-county area are the result oflarge-scale glacial stagnation, most of them ultimately th eresult of mass movement, mainly flowage and sliding . Afew disintegration ridges may have formed by upwardflowage of saturated material into openings in the base ofthe ice . Although most of the disintegration ridges on th eMissouri Coteau in Ward County are either relativelystraight or roughly circular "doughnuts," in severa lareas they are braided or meandering . These probabl y

29

originated in stream channels in the stagnant ice . Mostdisintegration ridges consist almost entirely of till wherea sothers, located mainly in extensive outwash regions, aredraped by gravel . It is difficult to distinguish this typ eof disintegration ridge from an esker . Figure 8 illustrate stwo ways circular disintegration ridges can form .

As the stagnant glacier melted, topography on th esurface of the ice was continually inverted . When sink -holes in the stagnant glacier finally melted through to th esolid ground beneath, circular holes formed in th eglacier . Material flowing down the sides of these hole scompletely filled many of the holes, resulting in hills o fmaterial occupying the positions of the former sinkholeswhen all the ice finally melted . If the amount of materia lflowing into a hole was not enough to completely fill it ,the material formed a doughnut-shaped ridge at the baseof the sides of the hole; ridges such as these arecommonly called " circular disintegration ridges" o r"doughnuts . " Doughnuts are best developed over areas ofthe Makoti and Ryder glacial units (fig . 6), which havelow to medium relief .

If, in the final stages of topographic inversion, thic kdeposits of material in the bottom of sinkholes cause dthem to invert into ice-cored cones, the material may hav eflowed down the sides of the cones, producing, when al lthe ice had melted, doughnut-shaped ridges, also calledcircular disintegration ridges . Ridges formed by materia lmoving down ice slopes and collecting at the base o fslopes are called "disintegration ridges ." The ridgesgenerally form random patterns and they may be an yshape, from circular to straight, depending on the shap eof the former ice slope and the fluid content of th esediment as it slid into place .

Dump ridges are another type of disintegration ridg eon the Missouri Coteau in Ward County . They are relatedto widespread deposits of collapsed outwash and sub-icechannels at the edge of the Martin Glacier . Dump ridge sconsist of esker-like ridges of coarse gravel and boulder sthat were deposited at the edges of stagnant ice fields .The ridges were formed by dumping of coarse outwash a tthe margin of the ice after meltwater carried the fine rmaterial downslope forming adjacent outwash deposits . Atypical dump ridge is 20 to 45 feet wide at the base, 5to 20 feet high, and it may be as long as a mile .

Clayton (1967) presented theoretical reasons why mos tof the circular disintegration ridges are about 500 to 600

30

I .

a .

b .

C.

Figure 8 . Diagrams showing how circular disintegration ridges form i nareas of collapsed glacial topography . Small "doughnuts" (500to 600 feet across) might form in three stages, as shown inthe uppermost series of diagrams (la, Ib, and lc) . Large"doughnuts" (1,000 feet or more in diameter) might hav eformed as shown in the lower two diagrams (IIa and IIb) .Diagram I is adapted from Clayton, 1967 . Figure A-2 ; diagra mII is adapted from Deal, 1971, Figure 13 .

31

feet: across . He pointed out that, beneath a depth o fabout 150 feet, ice of temperate glaciers tends to becom eplastic ; ice depressions deeper than about 150 to 200 fee tare unlikely . If maximum probable ice slopes were abou t40 degrees, and maximum depression depths were abou t200 feet, the average ice depression would have bee nabout 550 feet in diameter . These values are close to th eequivalent values for depressions in modern stagnan tglaciers .

In two areas along the Missouri Escarpment in War dCounty, areas of glacial topography with internal linearityoccur (areas designated Qccl on pl . 1) . These areas hav eup to 100 feet of local relief and numerous chains o fpotholes . The southern area (Tps151-153N, R82W) appearsto represent a lateral or terminal moraine that wa sdeposited as a lobe as the glacier receded from the re-entrant in the Missouri Coteau in that area .

Waterworn Topograph y

Areas shown as Qces, Qcep, and Qcet are waterwor nsurfaces with relatively low relief (as opposed to areas o fQcer, described below) that may be cut from till o rpreglacial materials and may or may not be veneered o rcovered by fluvial deposits . In places east of Minot ,where water spilled out of the Souris spillway and flowedeast to glacial Lake Souris, an intricate system of braide dand anastomosing channels was cut (Qces on pl . 1) .Similar broad areas of washed till that may be veneeredby discontinuous patches of sand and gravel, occur in th enortheastern part of the area near Kenmare . The topog-raphy in these areas is characterized by numerous shallowchannels and by a lack of typical collapsed topographyadjacent to the channels . This eroded or washed topog-raphy is distinct and unmistakable in places, but in othe rareas, the evidence of erosion is less evident as washedtopography grades into unwashed topography .

Along the edge of the Missouri Escarpment, washedareas of till, along with some preglacial bedrock, slop etoward the northeast (Qces on pl . 1 ) . I n places, how-ever, especially flat, apron-like surfaces occur that arecovered by angular gravel deposits up to 5 feet thick .These are considered to be pediments (Qcep on pl . 1) .Pediments are especially well-developed in the southeast -ern part of Ward County in Tps 152 and 153N, Rs 81 ,82, and 83W .

32

Slopewash-Eroded Topograph y

Slopewash-eroded topography occurs along the side sof the Souris and Des Lacs valleys and along the valleysof their major tributaries (Qcer on pl . 1) . These steeplysloping valley sides have been dissected by water flowin gin rills, gullies, and coulees, which have developed sinc ethe spillways formed during deglaciation . The erosion ofthese slopes is a discontinuous process most evidentduring heavy showers or thunderstorms when sediment i seroded and transported through the system of gullies an dcoullees by intermittently flowing streams . With time, thegullies become deepened and lengthened by headwar derosion . This process has been more active southwest ofthe Souris and Des Lacs valleys because the land slope snortheastward from the higher elevations of the Missour iCoteau to the river valleys, thus forming a natura ltopographic gradient for the intermittent streams .

Glacial Lake Landforms

Glacial lake deposits cover wide areas on the MissouriCoteau . Sediment deposited in lakes that existed o nstagnant glacial ice is designated Qcle and Qcic on plate1 . Two major types of glacial lake deposits are evident :ice-walled and supraglacial . Ice-walled lakes were bottomedon solid ground and surrounded by stagnant ice that i nmost places was no more than a few tens of feet thick .The materials deposited on the former floors of thes elakes are designated Qcle on plate 1 . Superglacial lakeswere bottomed on stagnant ice . The surrounding oradjacent ice melted slowly due to relatively thickaccumulations of sediment on the ice . The sedimen tunderwent collapse at least once, possibly several times ,as the underlying ice melted . The collapsed material sdeposited in these lakes are designated Qcic .

Several unstable ice-walled lakes (Qcic) formed on th esurface of the Makoti deposits . These lake plains are o flimited extent and encompass an area smaller than asquare mile . Modern evidence of the lakes consists of thi ndeposits of laminated silt and clay about 2 to 10 fee tthick; they are barely distinguishable on aerial photo -graphs . The sediment in unstable-environment lakes tendsto be siltier and sandier than that in the stable-environ-ment lakes . Most of the water in the stable-environmentlakes was derived from rainfall, not from melting of th e

33

underlying ice, and, as a result, they probably containedless sand and more clay than the unstable-environmen tlakes . One example of an unstable-environment lak edeposit, in sections 12, 13, and 14, T151N, R87W, i srimmed by till and is probably a supraglacial type .

Evidence for five large ice-walled lake deposits occur sat the distal margin of the Ryder deposits (Qcle on pl .1) . Locally, these deposits are collapsed . Probably onlytwo lakes are represented, the deposits having beenmodified by glacial erosion . The three distinct elevateddeposits at Ryder are extensively collapsed over wid eareas and are dissected by moats or valleys . The lakedeposits at Ryder consist of sand and gravel and th elargest segment is distinctly rimmed with sand . Thisdeposit was formed by stagnant ice damming the meltwate rflowing south and west from the melting Ryder ice mass .The other ice-walled lake deposit, about six miles north -west of Makoti, is collapsed over a wide area, although alarge segment of it is nearly intact . Both large segmentsare surrounded by moats and locally by sand or grave lrims . This lake was fed by meltwater from the Martin icemass only a few miles to the northeast .

The three ice-walled lake deposits mapped on th eRyder deposits are represented by elevated surfaces an dtwo are partly rimmed by sand and gravel . The rims ofthese lake deposits represent disintegration ridges .

The most prominent glacial lake deposits occu rbetween the Martin ice-margin border and the Missour iCoteau Escarpment (pl . 1) . These supraglacial lak edeposits, which exceed three square miles in area, ar eextensively collapsed around the margins . They consist o ffinely laminated clay . Fifty-four feet of lake clay wa spenetrated in a test hole in section 3, T153N, R85W . Th elake deposits have smooth, boulder-free surfaces andnearly all of the deposits are cultivated . Many of th elakes illustrated on plate 1 apparently were combined t oform larger lakes at one time, but now they appear a sseparate deposits owing to subsequent collapse . Apparent-ly, those segments of the lakes that were bottomed on ic ecollapsed as the debris covered Martin ice melted .

A 12-square-mile lake plain (Qcln on pl . 1) in theTolley area (Tps160-162N, Rs86-87W) represents th eformer extent of a lake that existed for a time after theice melted . This boulder-free area is underlain by les sthan 10 feet of fine sand and laminated silt .

34

Fluvial Landform s

Landforms resulting from the action of running wate rinclude deposits of meltwater rivers on and near th eglacier and deposits of non-meltwater rivers . They wereleft undifferentiated because no consistent way of distin-gui :shing them is known . Much of the material called"outwash" on previous maps was deposited by river sconsisting largely of runoff from precipitation rather tha nfrorn meltwater . For example, the youngest collapse dfluvial deposits of the Missouri Coteau (Qcrs on pl . 1 )was deposited thousands of years after the glacier therestagnated, when less than a tenth of the runoff wa sderived from melting ice (Clayton, 1967) . Even the fluvia lmaterials deposited by some of the meltwater rivers is notreally outwash . For example, much of the sand and grave lin the eastern part of the area (Qcrf) was deposited b yfloods of water flowing from glacial Lake Regina to th enorthwest of the two-county area .

Spilllway s

Alan Kehew has described the landforms and fluvia ldeposits associated with catastrophic drainage of glacia lLake Regina through the Souris and Des Lacs valley s(refer to fig . 9) . The next several paragraphs are take nlargely from his descriptions of these spillways (Kehew ,1982) . The Souris spillway begins as a broad, shallo wfive-mile wide channel that merges upvalley with the floorof glacial Lake Regina near Weyburn, Saskatchewan .Immediately downvalley from this outlet, the spillwa yconsists of an inner trench that is about a half mile wid eand as much as 150 feet deep . Both sides of the innertrench are flanked by erosional, terrace-like surfaces tha tare up to four miles wide (Aces) . These upper surfaces ,which are continuous upvalley with the floor of the Lak eRegina outlet, contain conspicuous longitudinal groove sparallel to the inner channel . The grooves have beenmodified in some places by Holocene erosion and they nowcomprise segments of intermittent tributaries to the inne rchannel of the spillway . The lower part of the uppe rerosional surface is mantled by a discontinuous lag deposi tof coarse sand and gravel (Christiansen, 1956) . Thi sbouldery, lag-covered area is less suitable for farmin gthan the adjacent land, and the resulting difference i nland use makes a striking contrast on aerial photographs .

35

.C.iA CiVATv Hy , iAw I IVIAI'{i I J UU Al

MOOSQ

LakendiHI

Mountai nCreek

S .'114CANADA

Laiv ler a

105°

WAR D

102°

101 °

,.;T SPILLWAY-ERODED BY FLOO D,-1-1-r PRE-EXISTING SPILLWAY- MODIFIED BY FLOO DTO; AREAS OF FLOOD EROSION OUTSIDE OF SPILLWAY S

UNDERFLOW FAN DEPOSITS - COARSE DEPOSITS ONL YIN LAKE SOURIS AND LAKE HIN D

Figure 9 . Area affected by catastrophic floods . Floods of water from Lake Regina and lakes north of ther ecarved large channels including those of the Des Lacs, Souris, and Sheyenne Rivers . The erosion ofthe channels resulted in great amounts of find-grained sediment, much of which settled into th evarious lakes along the path of the floods . Diagram is adapted from Kehew and Clayton, p . 189 (1983) .

The inner channel generally maintains a regula rchannel shape along its course, with nearly constan twidth and depth . In places, the inner channel bifurcate sinto two parallel trenches separated by a narrow ridge .About midway between the Lake Regina outlet and theinternational border, the inner channel becomes broad an dshallow for a short distance . In this area, glacial sedimentis thin, and the inner channel is cut entirely into thePaleocene bedrock . Drill holes on the floor of the inne rchannel near Minot indicate a thickness of 135 to 165 fee tof mostly fine-grained, dark-colored, fossiliferou salluvium, colluvium, and lacustrine sediment . Coarse san dand gravel at the bottom of the Holocene valley fill, whic hcould have been deposited by the glacial •spillway flood ,is either thin or totally lacking in some places . Theoriginal erosional depth of the inner channel, prior todeposition of Holocene sediment, was as much as 350 feet .

In the vicinity of the international border, th espillway splits into two branches . One branch consists ofa continuation of the broad upper surface across a lowdrainage divide to the south . For several miles, no inne rchannel exists in the broad, grooved surface; however,over a distance of about 25 miles, an inner channel a mil ewide and 135 feet deep develops . This spillway, the De sLacs spillway, rejoins the Souris spillway north of Minot .The Des Lacs spillway has two topographically differentsegments . The valley sides of the lower segment are muchmore deeply dissected by gully erosion than those in th eupper segment . In addition, the lower segment has noscoured upper surface . Non-eroded, hummocky topograph yextends to the edge of the spillway . Therefore, it i sapparent that floodwater from the Lake Regina discharg ewas channeled into a pre-existing drainageway and thatbankfull flow in the lower Des Lacs spillway was neve rachiieved .

Downstream from the point where the Souris spillwaybifurcates at the international border, the main branch ofthe spillway makes several relatively tight meanders an dthen trends southeastward to the point where it i srejoined by the Des Lacs spillway near Minot . Thi ssegment of the spillway consists of a trench one to tw omiles wide and 150 feet deep and, like the Des Lac sspillway, it has an upper level . Because both the lowe rreaches of the Souris and Des Lacs spillways lac kerosional surfaces and are larger than the inner channe lnear Lake Regina, it is likely that the lower segments of

37

both spillways were pre-existing drainageways that wer escoured and deepened by discharge from glacial Lak eRegina .

Along the path of the Souris spillway downvalley fromthe main bifurcation, water spilled out of the channel an dflowed eastward in several locations . The most spectacula rexample of such overflow can be seen downvalley from th epoint where the Souris and Des Lacs spillways join . Justwest of Minot, the spillway makes a sharp bend to th eeast, and then, east of Minot, it makes another shar pbend to the southeast . At this second bend, a large partof the flow from the main channel breached the east wal lof the channel and continued eastward along a more directpath to the glacial Lake Souris basin . While flowing ou tof the channel, the water eroded the channel wall t oabout a third of its normal height in that area . Justbeyond the breakout point on the channel, the wate reroded a plexus of anastomosing channels in the glacia lsediment adjacent to the spillway (area of Qces on pl .1) . Streamlined fluvial erosional forms are present in thi sarea, but the anastomosing channels contain no fluvia lchannel deposits near their heads . Farther along the floodpath, the channels diverge around obstacles that musthave been higher than the flood-water surface . Channe ldeposits consisting of coarse gravel in point-bar position sin channel meanders are more common with increasin gdistance from the Souris spillway . The water continued toflow eastward to the glacial Lake Souris basin .

Depositional features of the massive discharge fro mglacial Lake Regina and associated spillway erosion includ ehuge point bars located at the inside of each spillwa ymeander . The flat to gently undulating upper surfaces o fthe bars are mantled with boulders up to ten feet i ndiameter . The boulders are probably a lag deposit tha tformed as discharge and velocity dropped, causing thei rdeposition . The bars project as much as 80 feet above th epresent valley floor . Test-hole lithologic logs indicate thatcoarse sand and gravel extend into the subsurface alon gthe valley wall .

One of the largest bars in either spillway occurs jus twest of Minot . This bar, which has been extensivel yquarried for sand and gravel, consists of an upper leve lof moderately well-sorted and cross-bedded sand an dgravel that looks much like most normal outwash deposit sin the area . The lower section of the bar, however, i scomposed of poorly sorted, nonbedded gravel containin g

38

very large boulders . Some of the boulders are nonresis-tant lithologic types such as glacial till and Tertiar ysandstone, shale, and lignite, which must have bee ndeposited quite near their source . One chunk of lignit eweighing two tons was removed from this deposit an dburned for fuel (Laird and Hansen, 1958) . This lowe rpart of the bar suggests a rapid decrease in curren tvelocity, resulting in dumping of material eroded justupstream . Such deposition could be the result of a floo dterminating with a rapid drop in discharge soon afte rpeak discharge had been achieved .

Overridden Fluvial Deposit s

Stratified deposits of sand and gravel covered byglacial till are locally exposed along the walls of th eSouris River between Minot and Verendrye, in McHenryCounty . These fluvial deposits crop out about a third ofthe way up the valley walls and locally they form indis-tinct benches, the result of recent differential erosion .The deposits range up to about 30 feet thick . Bedding i sindistinct in most exposures . The finest material generall yshows the greatest degree of bedding . Where exposed ,these deposits are dirty with iron stains, although bedde dlayers of sand are generally clean and contain little or n oiron oxide . These rather extensive deposits of sand an dgravel were formed prior to the last glacial advance . I nareas where the deposits are exposed, they are commonl yoverlain by several tens of feet of younger glacia ldeposits .

Other Glaciofluvial Deposit s

Stratified deposits of sand and gravel of glacia lorigin are scattered at the surface throughout the mappe darea (Qcrf and Qcrs on pl . 1) . Generally, these deposit sare only a few hundred feet across and they rise 15 feetor less above the surrounding till plain . Most of th edeposits are only a few feet thick and the lithologi ccomposition is similar to outwash and stratified ice-contac tdeposits in the area . Most of these deposits are probablyice-contact in origin .

One rather extensive deposit is in a low topographicsag on the till plain half a mile southwest of Tolley i nnorthern Renville County . It consists predominantly of

39

sand and fine gravel, and in most places it is less tha nten feet thick .

Eskers and Kame s

Eskers and kames are found throughout the area o fthe glaciated plains northeast of the Missouri Escarpmen tin Renville and Ward Counties (red lines on pl . 1) .Typically, they are mounds or sinuous ridges of variou ssize, and shape . They consist predominantly of poorl ysorted gravel, sand, and silt . Their bedding is commonl ydeformed by collapse along the margins of the deposits .

Kames range in height from less than 5 feet to large ,well-defined hills as much as 115 feet high . They aregenerally irregular in shape, although most are nearl yround or oval . Black Butte, which may be a kame, i slocated about a mile and a half northeast of Sawyer . I tconsists of sand and gravel with smaller amounts o fcobbles and boulders . The bedding is considerabl ydeformed in all exposures, possibly because of subsequen tcollapse following melting of the supporting ice . It is alsopossible that Black Butte is actually a block of ice-thrustmaterial, but if it is, it is not a typical example of suc ha feature . There is no apparent source depressio nassociated with the butte, and although sand and grave lare sometimes the dominant material in ice-thrust masses ,other, more competent materials are more common .

Most eskers are sinuous ridges of ice-contact materia lconsisting mainly of sand and gravel ; glacial sediment i slocally common . Eskers are locally abundant on th eglaciated plains, but they are much smaller than the onesfound on the Missouri Coteau . Eskers on the glaciate dplains are generally less than 5 feet high, whereas thos eon the Missouri Coteau stand 10 to 25 feet above th esurrounding terrain .

PLEISTOCENE FOSSILS

Only a few fossils have been collected from Pleis-tocene deposits in Renville and Ward Counties . Those thathave been recovered consist mainly of grasses, woo dfragments, cones, gastropod and pelecypod shells ,ostracods, and a few small bone fragments .

Wood fragments have been recovered from a numberof test holes or wells, but in all but two cases they werevery small . However, a sample of wood fragments, cones ,

40

and grasses was collected from a test hole located in th eNErNEiSE:1sec 3, T153N, R86W from a depth of 137 feet(Pettyjohn, 1967b, p . 128) . The wood fragments wereused for a radiocarbon 'age determination and dated a t10,350±300 B .P . The grasses were not studied, but th econes were identified as white spruce or Picea glauca(Moench) Voss . A white spruce log, nearly 2 feet long ,was collected from a depth of 12 feet from collapsedglacial topography 8 miles west of Berthold (NW*SE*se c181, T157N, R83W) . This wood (W-1818) was dated a t10,330±300 B .P .

Several fossil gastropods were collected from collapsedsuperglacial lake deposits in the NE*NEiNW*sec 32 ,T151N, R83W . Dr . Aurele LaRocque, Ohio State Univer-sity, studied the forms and found one or more species ofostracods in addition to the following snails : Pisidiu msp ., Helisoma anceps striatum (F . C . Baker), Valvatatricarinata (Say), and Gyraulus parvus (Say) .

Dr . LaRocque suggested that these forms indicate awarm, shallow, freshwater environment that contained a nabundance of vegetation . This implies that the ice-contactlake in which the forms lived was well insulated from th eunderlying ice . Although meltwater fed the lake andcarried in large amounts of sand, the water was relativelyfree of silt and suspended clay--as indicated by th epresence of the prosobranch or gill-breathing snai lValvata . Consequently, it can be assumed that the fossil scollected from the collapsed superglacial lake deposit sinhabited a nonglacial environment .

Two genera of pelecypods were collected from sample sof coarse sand and fine gravel penetrated at depths of 5to 40 feet in many places in the Souris River valley .These fossils were not identified . A few bone fragment swere collected from test-hole cuttings in the Souris Rive rvalley, but these fragments were too small for identifica-tion .

GEOLOGIC HISTORY

The oldest glacial deposits exposed in the two-countyarea consist of a thin layer of till on the distal (south -east:) side of the Blue Hill surface till (fig . 5) (secs . 33 ,34, 35, and 36, T151N, R85W) near Douglas . This, theSnow School till, was deposited by a glacial advance tha tprobably occurred during early Wisconsinan time . The

41

thin, deeply oxidized till sheet probably extends at leastas far south as the Missouri River . Following depositionof the Snow School till sheet, the glacier retreated to thenorth and the upper part of the till sheet was thoroughl yweathered .

The series of diagrams (figs . 10-16) show how th eLate Wisconsinan glacier shaped the landforms in th eRenville and Ward County area . All but a small part ofsouthern Ward County was covered by the Late Wiscon-sinan glacier, which at its maximum extent sometime prio rto 13,000 years ago, deposited the Blue Hill moraine an dBlue Hill deposits in the southwest corner of the countyand in northern McLean County (fig . 10) . The pre-LateWisconsinan glaciers had covered the entire two-countyarea and broad areas to the south . It is possible that, by13,000 years ago, the glacier had already stagnated ove rthe Turtle Mountains upland to the northeast, and by thattime the flow directions of the glacier in the region werebeing influenced by the upland in the Renville and WardCounty area .

The recession of the glacier from the area was acomplex and irregular process with numerous readvancesinterrupting the overall deglaciation . About 12,300 year sago, a readvance of the Late Wisconsinan glacier from th enortheast, over the Missouri Coteau, truncated the Blu eHill moraine (fig . 11) . This advance from the northeas tdeposited the Makoti till . As the glacier advanced up thenortheastern margin of the Missouri Coteau, extensiveimbricate thrusting occurred and great quantities ofglacial sediment were incorporated in, and on top of theice . When the glacier ceased to advance, the debris-lade nice on the Coteau separated from the major glacier an dslowly melted during the next 2,000 to 4,000 years . Th eMakoti moraine is well developed four and a half mile ssouth of the village of Makoti where it blocks the Hidden -wood Lake valley in T151N, R87W . The distal margin ofthe Makoti moraine trends northwest-southeast across aten-mile stretch in Ward County . It continues southeastinto McLean County, but becomes indistinct near the Blu eHill moraine . The Makoti moraine has not been tracednorthwest into Mountrail County ; it is indistinct o nai rphotos .

The Makoti moraine consists of discontinuous ,generally elongate, isolated ridge-like masses of glacia lsediment . Most of the gaps between the ridges appear t orepresent areas in which no real moraine was built

42

LI Unglaciate dEarly Wisconsinan Till Plain

n End Morain e

Drift Covered Stagnant ice

®Active Ic e

High Level Erosion Surfac e

Direction of Ice Flow

'~► Direction of Water Flow--_ Missouri Escarpmen t

River Valley

Figure 10 . Maximum extent of Late Wisconsinan ice in Renville and War dCounties . All but the southernmost part of Ward County wa scovered by ice in Late Wisconsinan time, probably prior t o13,000 years ago . This diagram illustrates the deposition o fthe Blue Hill moraine in southern Ward and northern Mc