USF Libraries
USF Digital Collections

The genesis and hydrogeology of a sandstone karst in Pine County, Minnesota

MISSING IMAGE

Material Information

Title:
The genesis and hydrogeology of a sandstone karst in Pine County, Minnesota
Added title page title:
Karst theses & dissertations
Physical Description:
vii, 171 leaves : ill., maps ; 29 cm.
Language:
English
Creator:
Shade, Beverley Lynn
Publication Date:

Subjects

Subjects / Keywords:
Caves -- Minnesota   ( lcsh )
Hydrogeology -- Minnesota   ( lcsh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (M.S.)--University of Minnesota, 2002.
Bibliography:
Includes bibliographical references.
Additional Physical Form:
Also issued in typescript.
General Note:
Description based on print record.
Statement of Responsibility:
by Beverley Lynn Shade.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001968350
oclc - 269392435
usfldc doi - K21-00004
usfldc handle - k21.4
Classification:
System ID:
SFS0001081:00001


This item has the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader ntm 2200349Ka 4500
controlfield tag 001 001968350
005 20090805144830.0
006 m d
007 cr bn|||||||||
008 081106s2002 mnua sb 000 0 eng d
datafield ind1 8 ind2 024
subfield code a K21-00004
035
(OCoLC)269392435
040
FHM
c FHM
043
n-us-mn
1 100
Shade, Beverley Lynn.
4 245
The genesis and hydrogeology of a sandstone karst in Pine County, Minnesota /
h [electronic resource]
by Beverley Lynn Shade.
260
2002.
300
vii, 171 leaves :
b ill., maps ;
29 cm.
500
Description based on print record.
502
Thesis (M.S.)--University of Minnesota, 2002.
504
Includes bibliographical references.
530
Also issued in typescript.
533
Electronic reproduction.
Tampa, Fla. :
University of South Florida Library,
2008.
n Included in the USF digital collection: Karst theses & dissertations.
Mode of access: World Wide Web.
0 650
Caves
z Minnesota.
Hydrogeology
Minnesota.
740
Karst theses & dissertations.
776
Original
w (OCoLC)62692138
773
t Karst Theses and Dissertations
049
FHMM
090
GB605.M6 (ONLINE)
856
u http://digital.lib.usf.edu/?k21.4



PAGE 1

The Genesis and Hydrogeology of a Sandstone Karst in Pine County, Minnesota t ` A THESIS SUBMITTED TO THE FACULTY OF THE GRADUATE SCHOOL OF THE UNIVERSITY OF MINNESOTA BY Beverly Lynn Shade IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE JULY 2002

PAGE 2

ABSTRACT The landscape of central Pine County, in east-central Minnesota, contains a series of sinkholes, streamsinks, springs and short caves These landforms occur in and over Precambrian Hinckley Sandstone, and the overlying unconsolidated glacial deposits These features serve the same function as in carbonate karst terrains : sinkholes and caves focus recharge, feeding a heterogeneous subterranean flow system This drainage system exploits high permeability zones in the sandstone, and discharges into springs In north-central Pine County, 262 sinkholes, 25 streamsinks and 32 springs have been mapped The Hinckley Sandstone is a quartz arenite No carbonate grains or cements have been found in sandstone samples from the sinkhole area No evidence has been found that calcite solution controls bedrock permeability This is a sandstone karst Three parameters appear to control the occurrence of karst sinkholes : the depth to bedrock, the type of underlying bedrock, and meter-scale heterogeneity in surface sediments The subsurface heterogeneity implied by the presence of active karst sinkholes is supported by the analysis of groundwater chemistry in the area Based on a preliminary groundwater investigation, water in the Hinckley and Fond du Lac sandstones can come from a variety of sources : modern precipitation, old sandstone brines, connate brines in basalt, and mixtures of all three The heterogeneous permeability structure of the Hinckley Sandstone appears to be controlled by outcrop and hand sample scale fractures and sedimentary features i

PAGE 3

ACKNOWLEDGEMENTS Financial support for this project was provided by the Pine County Soil and Water Conservation District (SWCD) and the Geology and Geophysics Department of the University of Minnesota (Minneapolis) Banning State Park granted permits to investigate sinkholes within the park Calvin Alexander served as my advisor for three years He was extensively involved in this project and spent many days out in the field with me, and many weekends in front of the computer with me trying to sort things out Scott Alexander provided so much help in so many ways that it would be impossible to begin a list If I could remember half of what he says, this would have been done at least a year ago Calvin and Scott helped and guided me through every aspect of the fieldwork I had many other great field assistants : Katherine Dalton, Neal Hines, Nick Johnson, Jill Leonard, Stacy Russo, Bill Stone and Hong Truong Terry Boerboom, Al Knaeble, Carrie Patterson and Tony Runkle at the Minnesota Geological Survey provided valuable commentary Dick Noyes and Sam Martin from the Pine County SWCD provided support throughout the project The citizens and landowners of these townships in Pine County shared their knowledge of their land, and gave me permission to look for, look in, and excavate various sinkholes Ultimately, this work was conducted for the people of Pine County I hope that they derive some benefit from my work, both in an increased appreciation of how this system operates and in a better understanding of the groundwater management issues at hand The most important factor in my completion of this degree was the belief from childhood that education is worthwhile My parents and sister have always been supportive of my efforts, which was invaluable And I must acknowledge the unflagging support of caverdom, who gave me the stupid idea to look in holes in the first place ii

PAGE 4

TABLE OF CONTENTS CHAPTER I KARST FEATURES E Interpretations 1 What isn't a sinkhole? t 48 2 Sinkhole Distribution Relative to Bedrock Type t 51 3 Sinkhole Distribution Relative to Depth to Bedrock t 54 4 Sinkhole Distribution Relative to Glacial Features t 54 5 Sinkhole Formation Models t 58 6 Composite Features and Associated Streamsinks t 64 REFERENCES t 65 111 D Observations 1 Sinkholes t 19 2 Excavation Description, Sinkhole D222 t 19 Excavation Description, Sinkhole D144 t 23 Excavation Description, Sinkhole D127 t 26 Excavation Description, Sinkhole D355 t 29 Excavation Description, Closed Depression D326 t 32 Streamsinks t 34 3 Springs t 36 4 Caves t 36 5 Composite glacial/karst features t 48 A Background 1 Sandstone Karst t 1 2 Definition of karst t 1 3 Global distribution of features t 2 4 Previous work t 3 5 Are the sinkholes in Pine County karst features? t 7 B Geological Setting 1 Bedrock Geology t 9 2 Description of Bedrock Units t 12 3 Surficial Geology t 14 4 Description of Surficial Units t 15 C Field Methods 1 Mapping t 16 2 Excavations t 17

PAGE 5

CHAPTER II HYDROGEOLOGY AND WATER CHEMISTRY D Iron Springs t 104 E Samples Excluded from Water Chemistry Discussion t 106 REFERENCES t 108 CHAPTER III PETROLOGY A Introduction t 110 B Outcrop Observations t 111 C Hand Sample Permeability Observations t 112 D SEM Work t 117 E Theoretical Solution of Quartz t 124 REFERENCES t 128 DISCUSSION t 129 iv A Introduction and Methods 1 2 3 4 Introduction t 68 Purpose t 68 Field Methods t 69 Laboratory Methods t 72 B Residence Time 1 Tritium Age Model t 74 2 Chloride/Bromide Age Model t 80 3 Carbon Dating t 85 a Background t 85 b Dated Samples t 87 C Water Chemistry 88 1 Piper Diagrams t 2 Calcium and Magnesium t 93 3 Strontium t 97 a Strontium in Rocks and Sediment t 97 b Strontium in Groundwater t 99 c Major Cations and the Sr Model t 102 d Boron and the Sr Model t 102

PAGE 6

FIGURE LIST Fig 1 Location of Pine County t 8 Fig 2 Bedrock Geology Map t ; t 10 Fig 3 Bedrock Cross Sections t 11 Fig 4 Sinkhole Distribution and Locations of Excavated Sinkholes t 18 Fig 5 Location of Sinkhole D222 t 20 Fig 6 Hells' Gate Block Diagram t 21 Fig 7 Cross Section of Sinkhole D222 t 22 Fig 8 Location of Sinkhole D144 t 24 Fig 9 Cross Section of Sinkhole D144 t 25 Fig 10 Location of Sinkhole D127 t 27 Fig 11 Cross Section of Sinkhole D127 t 28 Fig 12 Location of Sinkhole D355 t 30 Fig 13 Cross Section of Sinkhole D355 t 31 Fig 14 Cross Section of Closed Depression D326 : Trench 1 t 33 Fig 15 Cross Section of Closed Depression D326 : Trench 2 t 35 Fig 16 Robinson's Ice Cave/Bat Cave Map t 37 Fig 17 Hell's Gate Riverview Cave Map t 39 Fig 18 OMR Cave Map t 40 Fig 19 Lonesome Cove Cave Map t 42 Fig 20 Bear Cave Map t 43 Fig 21 Natural Bridge Cave Map t 44 Fig 22 Hovdingbroen Cave Map t 45 Fig 23 Porcupine Cave Map t 46 Fig 24 Wolf Creek Breakdown Cave Map t 47 Fig 25 Tree Fall Rip-up Holes t 49 Fig 26 Sinkhole Distribution over Bedrock Geology t 52 Fig 27 Bedrock cross section through area of mapped sinkholes t 53 Fig 28 Sinkhole Distribution over Depth to Bedrock t 55 Fig 29 Sinkhole Distribution over Superior Lobe Phases t 57 Fig 30 Sinkhole Formation Diagram : Collapse t 59 Fig 31 3D Geometry of Sediments in D127 t 61 Fig 32 Sinkhole Formation Diagram : Subsidence t 62 Fig 33 Sample and Spring Locations t 70 Fig 34 Sampled and Decay-Adjusted Tritium in Precipitation in Ottawa, Canada t 77 Fig 35 Histogram of Tritium concentrations in Pine County t 79 Fig 36 Bromide vs Chloride with Respect to Tritium t 81 Fig 37 Bromide vs Chloride with Respect to Aquifer t 83 Fig 38 Piper Diagrams by Source : Quaternary and Sandstone t 89 Fig 39 Piper Daigrams by Source : Basalt and Springs t 90 Fig 40 Composite Piper Diagram t 92 Fig 41 Calcium vs Magnesium t 94 Fig 42 Alkalinity vs Ca+Mg t 96 Fig 43 Alkalinity vs pH t 96 V

PAGE 7

Fig 44 Calcium vs Strontium : Rock, Sediment and Water, t 98 Fig 45 Sr/Ca Box Plots for Rock and Sediment t 100 Fig 46 Magnesium vs Strontium, Sr vs Ca and Sr vs Mg t 101 Fig 47 Na vs Ca + Mg t 103 Fig 48 Boron vs pH t 105 Fig 49 Excluded Samples t 107 Fig 50 Sketch of Conduit-Bearing Outcrop t 113 Fig 51 Permeability in Samples AH-1 and BH-1 t 115 Fig 52 Permeability in Sample UH-1 and LH-2 t 116 Fig 53 Permeability in Samples LH-1 t 118 Fig 54 Permeability in Samples LH-3 and LH-4 t 119 Fig 55 Micrograph of a grain from UH t 121 Fig 56 Micrograph of a grain from UH t 122 Fig 57 Micrograph of a grain from LH t 123 Fig 58 Micrograph of grains from LH t 125 vi

PAGE 8

TAt i i LIST Table 1 Sinkhole Database t 132 Table 2 Streamsink Database t 138 Table 3 Spring Database t 139 Table 4 Water Sample Locations t 140 Table 5 Major Water Chemistry Data t 143 Table 6 Trace Element Data t 158 Table 7 Groundwater Ages from 14 C t 88 Table 8 Isotopic Data t 167 Table 9 Sr/Ca and Sr/Mg Ratios in Rock, Sediment and Water t 99 Table 10 .Calcite and Silica Equilibrium Solubility t 124 vii

PAGE 9

I A Theoretical Background I A 1 Sandstone Karst Karst on sandstone and quartzites is unfamiliar to many hydrogeologists and totally unknown to most resource managers However, this type of karst is well documented in the scientific literature, and has been recognized in a variety of locations worldwide One aspect of this thesis addresses how sinkholes form on sandstone, and why they exist in central Pine County Interpretation of the role of the Pine County sinkholes in the surface water groundwater systems depends on if the area is a karst system or not I A 2 Definition of karst According to White (1988), karst can be defined as a combination of process and form The process is weathering : the competition between chemical and mechanical forces A balance between the two creates a set of landforms whose origin lies in the direct solution of bedrock The terminology dealing with karst features and landforms is extensive, complicated and varies globally Avoiding the morass of terminology, White's suite of characteristic landforms includes closed depressions, interrupted surface drainages, as well as caves and associated subsurface drainage systems (1988) Almost all of the extensive karst terminology will fit into these three basic categories, with the exception of small-scale textural features such as karren Karst is distinct from other landforms due to "the dominance of solution as a geomorphic agent" (White, 1988) When solution and mechanical erosion compete, solution must be the more important process of the two in order for the resulting landforms to be considered karst Similar features within the set of landforms that owe their development to other causes are termed 'pseudokarst' A fundamental problem with this definition is that it is difficult to quantify the relative importance of mechanical and chemical weathering Under this definition, karst was effectively defined by lithology Limestone was the most important host rock, with a few others (dolomite, gypsum, halite) making up progressively smaller components I Karst Features 1

PAGE 10

Others have, developed broader definitions In his 1983 treatment of some puzzling features on sandstone, Jennings (1983) defines karst as the result of "the process, solution, which is thought to be critical (but not necessarily dominant) in the development of the landforms and drainage characteristic of karst ." He continues to define pseudokarst as "country with resemblances to karst, which are due to other processes ." This distinction previously separated two schools of thought on the classification and analysis of noncarbonate karst Jennings' definitions tacitly acknowledge that a karst system is more than the geomorphologic landscape ; recent research emphasizes the importance of a systems approach to define karst, based on universal characteristics, one that does not call for a specific lithology One of these universal characteristics is the presence of bedrock conduits Conduits are high transmissivity features that connect recharge to discharge areas and are capable of transporting groundwater rapidly in turbulent flow The critical size for transition from laminar to turbulent flow can be as small as 1 millimeter (Ford and Williams, 1989) Karst aquifers are parts of systems that have "a specific type of fluid circulation capable of self-development and self-organization" (Klimchouk and Ford, 2000) These issues are at the heart of what distinguishes karst flow from porous media flow and diminish previous disagreements by focusing on the behavior of an entire system In shallow systems, water carrying out the process of solution is routed through landforms quickly into the subsurface, where it contributes to the function o f a n integrated drainage system Following the spirit of Klitnchouk and Ford's definition, pseudokarst features would mimic the shape of karst, but are the result of different processes or serve a different function Pseudokarst is often composed of small-scale features like rillenkarren, which contribute to surface runoff and more conventional subsurface flow (Frye and Swineford, 1947 ; Alexandrowicz, 1989) I A 3 Global distribution of features Discussion of the traditional definition of karst leads easily into a consideration of karstic development in rocks other than traditional limestone In the words of Wray (1996), "This paradox of karstic landforms on some of the most insoluble of rocks,

PAGE 11

mimicking those on some of the most soluble, is becoming increasingly hard to ignore ." It is exactly this similarity of form that initiated serious study of sandstone karst A compelling example is a picture of sandstone karst in northern Australia that was misidentified in a textbook not only as limestone tower karst, but also as exemplary tower karst Jennings, 1983) This is a far cry from White's statement that "Karst occurs on carbonate rocks, gypsum, and to a minor extent on certain other rocks" (1988) While karst and pseudokarst occur in a variety of lithologies, the emphasis of the research discussed below is on nearly monomineralic silicates primarily low-grade metasediments As sandstones, these rocks would be quartz arenites As metamorphic rocks, they would be orthoquartzites However, after extensive solutional alteration, it is difficult to apply either of these terms uniformly A well-indurated orthoquartizte can be effectively reduced to friable sandstone, either uniformly (Young, 1988 ; Busche and Sponholz, 1992) or in a nonhomogenous fashion related to structure or lithology (Zawidzki, 1976 ; Young, 1987) Spatial variability of residual cementation appears to be an important characteristic of silicate karsts, as will be discussed later To illustrate the widespread nature of this phenomenon, karst features due to solutional processes have been recognized on all continents, across a wide spectrum of climates Wray's review paper notes silicate karst in Venezuela, Brazil, America, Morocco, Chad, Niger, Nigeria, Zimbabwe, Transvaal, South Africa, Thailand, Australia, the UK, Poland, former Czechoslovakia, and scattered sites in Western Europe (1996) Not included in Wray's 1996 list are the sinkhole-topped mesas on northwestern New Mexico (Wright, 1964), solution pans in Mongolia's Gobi Desert (Dzulynski and Kotarba, 1979), and the sandstone/quartzite sinkholes in northeastern Minnesota discussed in this thesis Caves with depths of nearly 400 meters and lengths up to several kilometers have been documented in silicate karst regions, demonstrating a local significance simply in terms of pervasiveness and scale of the phenomenon (Truluck, 1991 ; Wray, 1996) I A 4 Previous work Interest in solutional features in sandstones began as early as 1947 (Frye and Swineford), but was mainly restricted to morphological descriptions of surficial forms This trend has persisted in much of the literature dealing with silicate karst and other 3

PAGE 12

pseudokarst, which has fallen primarily into the realm of physical geography, focused in Europe and Australasia Much of this descriptive literature does not address the specific reactions responsible for the solutional process Nor are controlling factors compared to those driving the same process in limestone (Alexandrowicz, 1989 ; Robinson and Williams, 1992) Instead, form is used to imply process, "The remarkable similarity . [to forms] produced on the flat surfaces of limestones is not coincidental but entails the similarity of formative processes" (Dzulkynski and Kotarba, 1979) However, some work has been done to unravel this process ; both Australia and Venezuela have been the focus of investigations on the genesis of sandstone karst features In many ways, the Roraima region of southeastern Venezuela and northwestern Brazil has been the standard by which all other silicate karsts are measured Despite the difficulty of access, this spectacular area has attracted attention from a wide range of researchers since the late 1950's First noted by pilots, Roraima dissolution features include Angel Falls (Salto el Angel), which has intrigued and amazed many aside from geologists One of the earliest attempts to explain the processes behind the impressive shafts in the Roraima quartzite was headed by White The group felt that features on the large table mountains in southeastern Venezuela indicated large volumes of solution However, given the well-indurated nature and pure composition of this orthoquartzite, they could not ascribe the phenomenon to simple dissolution of Si0 2 by meteoric waters This is understandable, since experimental data suggested that the equilibrium solubility of quartz at surface temperatures and pH was 14 ppm (White et al, 1966) Compared to the solubility of calcite at the same conditions (250ppm at 25€C, P CO2 =10 5 bar), the difference is striking Since the solubility of amorphous silica is significantly higher that that of quartz (100 ppm) and trace amounts of opal were found rock samples, White et al (1966) postulated that the quartz cements still found elsewhere in the Roraima quartzite had been hydrated to opal by some undetermined process, and later removed by meteoric waters An interesting piece of information in White et al (1966) is their estimation of annual rainfall (about 7m/yr) on the mesa tops, extrapolated from nearby low-lying areas (about 1m/yr) due to sparse meteorological data at the time Despite the intervening three decades of continued study and improved technology, this number continues to be citied as reliable data, as recently as 1996 (Young) If this number has been misestimated, it 4

PAGE 13

would contribute to the prevailing idea that sandstone/quartzite karsts require a tropical climate A decade later, a theory by Zawidzski et al (1976) also drew upon the general alteration of the Roraima Quartzite, leaving some quartz grains easier to remove than if they were in a pristine quartzite Instead of hydration to opal, they suggested hydrothermal alteration This theory pointed to nearby granitic intrusions as a source for hydrothermal energy They cited obvious joint control of dissolution features as evidence that the hydrothermally driven fluids would have exploited these zones of higher permeability to move through the quartzite Structural control would therefore constrain hydrothermal alteration to specific bands and zones The aggressive hydrothermal fluids could dissolve the quartz cements The resulting weak areas would then be exploited as preferential flow paths when the rocks were exposed to surficial weathering It would be easier to remove material from those weak areas within the rock mass than to remove material from the rock surface The theory of Zawidzski et al makes use of solution long before the bedrock is exposed at the ground surface During surficial weathering, material is removed by purely mechanical forces This theory leaves unexplained the silica-based precipitates that have been documented in several different locations in the region, which imply that dissolved silica in the groundwater is currently precipitating This indicates that chemical erosion is more active than in their theory The work of Chalcraft and Pye (1984) refutes the original theory of opal hydration, based on a number of observations Primarily, the solubility of geologic opal (mostly composed of christobalite and tridymite) is significantly lower than that of experimental amorphous SiO 2 If geologic opal is roughly four times less soluble than experimental opal, its range of solubility overlaps that of quartz Chalcraft and Pye (1984) use more recent sources than White et al for silica solubilities They note that the solubility of the various forms of silica can be significantly enhanced by alginic and amino acids given off by algae The tops of the Roraima mesas have widespread colonies of algae and accumulations of organic material in bogs, very different from the floral assemblage in the surrounding lowlands The opal argument is further weakened by a lack of mechanism, empirical or theoretical, to provide the in-situ hydration SEM (scanning electron microscope) images show etching on grain surfaces as well as etching and flaking of the silica cements The 5

PAGE 14

authors attribute both groups of microscopic features to dissolution They argue for solutionally controlled process, aided by mechanical removal of grains, driven by the sheer mass of precipitation and time as well as elevated acidity of water originating on the mesa tops due to organic mediation Several sandstone karsts have been studied in northwestern Australia ; the largeand small-scale geomorphic features have been extensively described (Young, 1986 ; 1987), but there has also been a significant effort to correlate the existence and variability of these features with the processes behind them This has mainly been done with petrographic studies, to look for microscopic dissolution features, as well as some signature in lithology of dissolution that reflects the distribution of different landforms across the Bungle-Bungle massif ., In the East Kimberley region of northwest Australia, a variety of karst forms are developed in Paleozoic and Proterozoic sandstones It is difficult to apply either the term sandstone or quartzite to these rocks, due to spatial variability Solution of quartz can change a well-indurated orthoquartzite to friable sandstone, either pervasively, such as in the Bungle Bungle Range of northwest Australia (Young, 1988) and in Niger (Busche and Sponholz, 1992) or in a nonhomogenous fashion relate to structure or lithology, such as in Roraima (Zawidzki, 1976) and in the Cockburn Range of northwest Australia (Young, 1987) On a regional scale, there is variability in karst type by lithology : the Paleozoic Bungle Bungle massif in the southeastern part of the Kimberleys is a tower and ridge terrain, while the Cockburn Range to the northwest has a mesa and scarp morphology (Young, 1987) However, there is also variability in karst forms within these areas (Young, 1986) Young believes that the difference in form between the two ranges is due to differences in primary porosity at the onset of dissolution If both areas were karstified at the same time, the older Cockburn sandstones were understandably more mature than the Paleozoic Bungle Bungle rock, which retained more of its primary porosity Thus in the latter case, aggressive fluids were able to penetrate the rock mass, while in the former case dissolution was dominated by open fractures and joints For this reason, the Cockburn karst (mesa and scarp) is controlled by structural control, while the Bungle Bungle karst 6

PAGE 15

appears to be independent of this control The range of karst features across the Bungle Bungle massif is instead subject to depositional lithologic control (Young, 1986) These rocks weather by granular disintegration, and are easily broken, yet maintain near vertical slopes This behavior is due to two causes The first is that a thin residual crust can protect these friable rocks, so that water will slide off very steep slopes with little effect The other cause is very strong under vertical compression (as needed to maintain such steep slopes), even though it is not well cemented today Young believes that the character of the sandstone is due to its past as a quartzite, where the grains become close packed and cemented In thin sections from the Bungle Bungle area, the grains are still interlocking, but they are not longer connected due to solutional activity The interlocking grains are strong under vertical compression, but very susceptible to shear since they lack effective cements (Young, 1988) Extensive etching of grains and optically continuous overgrowths is seen in SEM images Overgrowths and interstitial cements seem more susceptible to dissolution Two main etching textures are seen in the Bungle Bungle The first are v-shaped notches that show strong crystallographic control These are surface reaction controlled features, indicative of slow fluid movement The other group are embayments ; a single embayment can cross a grain-overgrowth boundary, which does not happen with v-notches Embayments are flow-controlled features, indicative of higher fluid velocities (Young, 1988) I A 5 Are the sinkholes in Pine County karst features? Scientific literature demonstrates that karst exists in sandstones and quartzites Karst features exist in Pine County, but mapping them is complicated because Pine County was completely ice covered during the Wisconsinan glaciation Glaciation created many closed depressions at a variety of sizes Some closed depressions are sinkhole due entirely to karst processes Others are entirely due to glacial processes, while some closed depressions are composite features involving both karst and glacial processes 7

PAGE 17

I B Geological Setting of the Study Area I B 2 Bedrock Geology The study area lies in central Pine County, which is located in east-central Minnesota (Figure 1) The major bedrock divisions and the faults are shown in Figure 2 Moving from east to west across the study area, the first unit is composed of Mesoproterozoic basalts The basalts are separated from the Hinckley Sandstone by the Douglas Fault, a major northeast-southwest trending feature associated with the failed MidContinent Rift The smaller Hinckley Fault, developed west of the Douglas Fault, is also associated with this geologic event (Boerboom, 2001) Between the Douglas and Hinckley faults, Fond du Lac sandstone is covered by a very thin layer of Hinckley Sandstone, which should be the base of the Hinckley section West of the Hinckley Fault, the Hinckley Sandstone is considerably thicker, up to 150 meters thick (Mooney et al, 1970) Further west, the Hinckley/Fond du Lac contact is a stratigraphic contact These structural relationships are shown in Figure 3 The bedrock geology in this area is dominated by the indelible traces of the failed Midcontinent Rift The 1 .1 billion year old rift is over 2000 km long Its western limb extends as far south as Kansas, continues northeast into Lake Superior, and bends south in an eastern limb that terminates in Michigan (Cannon et al, 2001 ; Craddock, 1970 ; Hutchinson et al, 1990) The study area is located in the northern part of the western limb of the Midcontinent Rift System In this area, rifting probably began as early as 1130 m .y .a (Zartman et al, 1997), accompanied by crustal thinning and subsidence Normal growth faults formed parallel to the rift axis Rifting led to an active erosional environment, and eroded sediments were deposited in growing subsidence basins Magnetic studies show that younger basalt flows truncate older flows in several locations, implying unconformities (Cannon et al, 2001) These unconformities may correlate to a postulated volcanic hiatus from 1107 to 1096 m .y .a (Davis and Green, 1997) In the study area, volcanism associated with rifting did ended after 1094 m .y .a (Cannon et al, 2001) In the Lake Superior region, rift volcanism occurred in a shorter time period (1098 to 1091 m .y .a .) (Nicholson and Shirey, 1990) The relatively short period of volcanism and large amount of accompanying 9

PAGE 18

Figure 2 Bedrock units and selected structural features in Pine County, based on Boerboom, 2001 10

PAGE 19

Figure 3 Bedrock cross sections through the sinkhole area Bedrock data based on Boerboom, 2001 Pmhn = Hinckely Sandstone, Pmfl = Fond du Lac Sandstone, Pmcb = Mesoproterozoic Basalt The location of these cross sections is shown on Figure 2 ; they are parallel, both oriented roughly NW-SE

PAGE 20

basalt prompted Hutchinson et al (1900) to propose a model of rifting of a mantle plume coincident with the Midcontinent Rift System Some of the rift grabens were partially inverted to form horsts ; the study area abuts the western edge of the St Croix Horst The St Croix horst consists of a suite of uplifted volcanic rocks that are flanked on both east and west sides by basins of younger and contemporaneous rift-derived sediments The extent of the Midcontinent Rift has been defined by geophysical methods : it creates a magnetic and gravitational high, flanked on either side by magnetic and gravitational lows Basalts produced during rifting are thick, dense and highly magnetic, while, the rift-derived sediments such as sandstones are non-magnetic and less dense The sharp magnetic and gravitational gradients between the high and low anomalies attest to very steep faults bounding the St Croix horst (Mooney et al, 1970) The major graben growth faults found in this part of the rift are within the horst block (Pine Fault to the west and Cottage Grove-Lake Owen Fault to the east) The reverse faults that define the horst are the Douglas Fault on the west and the Hastings-Atkins Lake Fault on the east (Cannon et al, 2001) The basins adjacent to the Midcontinent Rift System contain thick sequences of sedimentary rocks The Fond du Lac is one such sedimentary rock : close to the Douglas Fault it may be as thick as 2 3 km, according to seismic studies, and thins westward, away from the rift zone, as shown in Figure 3 (Mooney et al, 1970) The Hinckley is also thicker towards the axis of rifting (up to 500m thick), and thins westward (Mooney et al, 1970) Morey (1972) interpreted the lithology and sedimentary structures of the Fond du Lac as "a fan-shaped wedge of clastic material that was deposited in a shallow, oxidizing, deltaic environment by a system of streams emerging from a western highland and dispersing material to the east and southeast" The more mature Hinckley Sandstone appears to be primarily composed of reworked Fond du Lac materials (Morey, 1972) I B 2 Description of Bedrock Units The youngest bedrock in Pine County are Paleozoic sedimentary rocks in the southern part of the county These are mostly sandstone, with some shale These rocks 1 2

PAGE 21

are located at the bedrock surface and are very thin ; they probably correlate with rocks from the Mt Simon to the St Peter sandstones (Boerboom, 2001) The youngest Mesoproterozoic rock is the Hinckley Sandstone Tryhorn and Ojakangas (1972) describe the Hinckley as a tan to orange, fineto mediumgrained quartz arenite that is typically about 96% quartz They report that the sandstone as well sorted and well rounded Close to the Douglas Fault the Hinckley contains other lithic elements, such as cobbles of maroon quartzite and pebbles of felsite and agate (Boerboom, 2001) A thin section of Hinckley Sandstone lies between the Douglas and Hinckley Faults : it is bounded to the east by a fault contact with Chengwatana basalt and to the west by a fault contact with a thicker section of Hinckley sandstone Since both faults are high angle reverse faults, Boerboom (2001) interprets this strip of sandstone as the basal section of the Hinckley sandstone Since this section of sandstone contains more than 10% of weathered kaolinitic feldspar and the lithic elements mentioned previously (Boerboom, 2001), it can be classified as feldspathic arenite The Hinckley is not as thick as the underlying Fond du Lac Formation, but occupies more of the bedrock surface than any other rock in Pine County This areal extent makes phenomena specific to the Hinckley (like sinkholes) of general interest to the entire county The Hinckley Sandstone is underlain by the thick Mesoproterozoic Fond du Lac Formation The Fond du Lac is at least 650m deep, established by drilling (Boerboom, 2001), and may be as thick as two or three kilometers near the Douglas Fault, based on seismic studies (Mooney et al, 1970) The Fond du Lac is a "quartzose to subarkosic sandstone, pale-orange to dusky-red, mediumto coarsegrained ; [with] interbeds of darkbrownish-red siltstone and minor shale" (Boerboom, 2001) Tryhorn and Ojakangas (1972) report that the Fond du Lac is subarenite composed of 36-68% quartz, 5-29% feldspar, with 1-10% lithic fragments, including chert, quartzite, basalt, felsite, ironformation, schist and mica The top contact of the Fond du Lac with the overlying Hinckley Sandstone is a sharp erosional unconformity in outcrop Near the contact, the Fond du Lac is a mediumgrained quartz-rich sandstone with weathered detrital material such as feldspar, mica and lithic grains The sandstone here also contains localized conglomerates, mostly composed 1 3

PAGE 22

of basalt Near its basal contact, tie Fond du Lac is a coarser pale-orange sandstone There are interbeds of siltstone and shale ,rid often a basal conglomerate of angular preKeweenan clasts (Boerboom, 2001) A group of Mesoproterozoic sedimentary rocks roughly contemporary with the Fond du Lac occurs in a swath in the southeast portion of the county There sandstone, siltstones, and shales are associated with the Bear Creek Syncline in the underlying older basalts and may be interbedded with volcanic rocks (Boerboom, 2001) The oldest Mesoproterozoic are rift basalts from the failed Mid-Continent Rift, in the St Croix horst These basalts include the Minong Basalt Sequence, which is composed of a poorly exposed ophitic basalt, an unnamed ophitic basalt, and rocks of the Chengwatana Volcanic Group The Chengwatana rocks are the closest basalt to the karst area (Boerboom, 2001) The youngest Paleoproterozoic rocks are a group of pelitic schist and metagraywacke in the northwest corner of the county The base of this unit appears to be transitional into the underlying Denham Formation (Boerboom, 2001) The Denham Formation lies south of the schist/metagraywacke The areal extent of the Denham Formation is small The Denham Formation is composed of a variety of amphibolite grade metamorphosed rocks The formation has three mapped members : an amphibolitic metabasalt, a mica schist, and a group of metasediments (Boerboom, 2001) The oldest rocks at the surface in Pine County are in the McGrath Gneiss, which is Late Archean in age This unit is composed of augen and flaser gneiss, and at the surface is located south of the Denham Formation (Boerboom, 2001) I B 3 Surficial Geology The surficial geology of all of Pine County is dominated by glacial deposits and landforms sculpted by glacial advance and retreat This area has surely been glaciated many times But this area was so pervasively glaciated during the Wisconsinan Glaciation of the Midto Late Pleistocene that the record of previous glaciations was eliminated During the Wisconsinan, there were two major sources of ice : the Superior Lobe moved into the county from the north and at its maximum covered the entire county and extended down in the central part of the state The Superior Lobe retreated and 1 4

PAGE 23

readvanced across Pine County at least twice (Knaeble et al, 2001) The Superior Lobe's Sandstone and Askov-Lookout Tower phases (or ice margins) fall within the study area, as shown in Figure 24 (Patterson and Knaeble, 2001) The second source of ice was the Grantsburg Sublobe of the Des Moines Lobe This body of ice entered Pine County from the southwest It arrived much later than the Superior Lobe and melted first At its maximum extent, the Grantsburg Sublobe covered about the southern quarter of the county, but Glacial Lake Grantsburg covered almost half the county, stopping just south of the sinkhole area (Knaeble et al, 2001) I B 4 Description of Surficial Units The sinkholes occur over five different surficial map units, according to Plate 4 of Atlas C-13 (Patterson and Knaeble, 2001) All of these surficial units are unconsolidated Pleistocene deposits of the Superior Lobe, in the Sandstone and Aksov-Lookout Tower phases The sinkholes near Sandstone along the Kettle River occur over (Qsgs), which is described as "sandy glacial sediment" (Patterson and Knaeble, 2001) All of the sinkholes mapped around Sandstone and in the Hell's Gate portion of the Kettle River were located over this unit In excavation, the till was sandy Furthermore, a layer of loess that may be stranded in the sinkholes was encountered in excavation The deposits in this area with mapped sinkholes had numerous sandstone and erratic boulders on the surface and in the shallow subsurface (in excavation) Further north along the Kettle River, near Log Drive Creek, the main surficial unit (Qsgf is described as "silty and clayey glacial sediment" Qsgf (here in the Askov-Lookout Tower Phase) is described as generally having more clay than the previous unit (Qsgs) (Patterson and Knaeble, 2001) Excavations made in Qsgf (D144) encountered significantly more clay than in the Qsgs (D222) In excavation, there were not many boulders of the sort described in the previous unit Some sinkholes in the Log Drive Creek Area were mapped over Qssi, which is described as "sorted sediment proximate to ice", which is dominantly sand and gravel (Patterson and Knaeble, 2001) The rest of the mapped sinkholes, around Askov and further east and northeast, are located over Qssm Qssm is described as "sandy glacial sediment" (Patterson and Knaeble, 2001) This portion of Qssm is near and on the Askov-Lookout Tower moraine 1 5

PAGE 24

ridge In three excavations (D127, D355, and 'closed depression 326), Qssm contained clean wells sorted sands as well as poorly sorted sands, clay-rich till and small bands of almost pure clay For a detailed description of the sediments encountered during the sinkhole excavations, see section I .D .1 I C Field Methods I C 1 Mapping Sinkhole locations were mapped by GPS (Global Positioning System) unit and comparison to USGS (United States Geological Survey) topographic maps Locations were recorded in UTM units Before the spring of 2000, a Garmin GPS 12CX unit was used with a portable base station With this system, internal precision was generally less than 2 meters Starting in the spring of 2000, the Garmin unit was used alone or with an amplifying antenna This system was more portable for hiking and internal precision was generally less than 5m A geographical information system (GIS) database was built for all the karst features, using Arcview Once the database was established, the location of every sinkhole was checked against field notes and field maps The results of this fieldwork is summarized in a database of sinkholes, streamsinks and springs (see Tables 1, 2, and 3) Karst features were found in a variety of ways by map inspection, aerial photos, ridgewalking, and talking to local residents Since the majority of features are below the resolution of map scale, the first two methods had limited utility When available, information supplied by residents was very useful ; all features that were found in this way were visited to assess whether they were karst features or not Additionally, between the spring of 1999 and 2002, hundreds of hours were spent searching for karst features on foot The sinkhole survey initially focused on Partridge Township The searched area was later extended into the townships of Bruno, Finlayson, and Sandstone The entire area underlain by Hinckley Sandstone has not been searched Every sinkhole in the areas searched has not been located, and sinkholes may be discovered in other areas of Pine County 16

PAGE 25

I C 2 Excavations Five excavations were conducted between September 2000 and November 2001 ; three were dug by hand and two were dug by backhoe The hand-dug excavations were documented at greater detail In these excavations, the trench area was oriented perpendicular to the most likely local bedrock fracture orientations The trench area was measured and defined by strings and corner posts Before excavation, a surface profile was measured along the entire trench length, using one of the leveled strings as a vertical reference Vertical measurements were taken to the nearest centimeter, at 10 cm intervals along the length of the trench During work, the reference string was removed to prevent damaging the string and corner posts The location of the string was measured with respect to the corner posts and could be reinstalled quickly, in order to measure the trench stratigraphy Trenches were excavated inside these measured areas, with vertical sidewalls The stratigraphy of the sidewalls was measured at the same intervals as the surface profile Stratigraphy was mapped by measuring depth to surfaces and prominent rocks in the sidewalls Sidewalls were measured at the end of each day and at the termination of each excavation Excavations lasted between 2 and 4 days The trenches were excavated as deep as possible within the constraints of safety, time and the water table Two features were excavated by backhoe The backhoe trenches were much larger and were necessarily measured in less detail, but the same method of mapping stratigraphy was used The results of these excavations are reported in II D 1 and their implications are discussed in II E 5 I D Observations This project mapped 262 sinkholes, 25 streamsinks and 32 springs in the study area The sharp boundary along the southeast side of the sinkhole array appears to be an actual boundary of the sinkholes' occurrence, while the edge of the sinkhole area in other directions remains undefined Figure 4 shows detail of the study area, and the location of all currently mapped sinkholes Not all karst features were found : more will be discovered 1 7

PAGE 26

Figure 4 Location of sinkholes mapped in Pine County ; the dashed areas show the location of larger-scale maps that accompany the discussion of each excavation Small inset map in lower right shows location of this map area within Pine County Sinkholes locations are shown with black triangles Exact locations of sinkholes are provided in Table '1 18

PAGE 27

I D 1 Sinkholes In order to understand how the sinkholes in this area form and how they are related to local geology, five excavations were conducted (Figure 4) The goal of these excavations was to explore and map the three-dimensional subsurface geometry of sediments Those sediment sequences are a record of the sequence of events that produced the surface features seen today Excavation Description Sinkhole D222 The first excavation was in sinkhole D222, in the southwestern portion of the study area (Figure 4) near Hell's Gate in Banning State Park (section 3 of Sandstone Township) D222 is one of the large cluster of sinkholes shown in Figure 5 The cluster continues to the northeast out of the figure, and contains >55 mapped sinkholes Figure 5 also shows several karst features on the west side of the Kettle River The most impressive of the features is Robinson's Ice Cave (Bat Cave), which is discussed in detail in section ILD .4 There are several sinkholes on top of the cliffs above the cave The karst features on both sides of the river form a linear ENE-WSW trend This trend probably reflects the location of a dominant fracture or group of fractures There are also numerous springs and seeps along this section of the Kettle River, which discharge a variety of water types The short residence time springs are probably fed by sinkholes The geometrical relationship of these elements is shown in a schematic diagram in Figure 6 A trench was oriented ESE-WNW across the depression, perpendicular to the overall trend of the sinkhole cluster The trench was excavated by hand to an average width of 0 .7 meters, and a depth of 1 .5 to 2 meters but did not reach bedrock The excavation revealed three sedimentary strata (an organic-rich soil and leaf litter layer, an aeolian loess, and a dark red till) and another deposit (gray material with a high concentration of roots and sandstone blocks) that crosscut the latter two strata (Figure 7) The top 20cm was composed of a black organic-rich material, which is composed of recent leaf litter and abundant shallow roots, as well as bigger tree roots This layer was up to 30cm thick in the center of the depression Underlying the dark organic layer was a tan layer of loess The loess was silty to very fine sand, without any significantly larger grains or rocks The loess had a few tree roots, but the thick tangle of roots from grass and 1 9

PAGE 28

Figure 5 Sinkhole array in Banning State Park, near Hell's Gate along the Kettle River Sinkholes are shown by black triangles The sinkholes follow a NE-SW trend, and align with sinkholes and a cave across the Kettle River See Figure 2 for location of this map area Sinkhole D222 is shown in cross section in Figure 4 The labeled UTM grid uses the 1983 North American Datum 20

PAGE 29

Figure 6 Schematic of Hell's Gate Area (Not to Scale) 2 1 Figure 6 A large group of sinkholes are located near the end of the Hell's Gate section of the Kettle River Most of the sinkholes in this group are located on the east side of the river Robinson's Ice Cave (or Bat Cave) is located on the west side of the river, and has several sinkholes above it Robinson's Ice Cave, the west side sinkholes and the east side sinkholes form a roughly linear array, NE to SW This linear array probably reflects a dominant fracture or group of fractures There are numerous springs and seeps along this reach of the river that discharge a variety of water types The short-residence time springs are probably fed by the sinkholes There are also springs and groundawter-fed wetlands along the lowest river terrace, as shown in the figure The springs emerge from talus piles at the base of the sandstone cliffs The springs and wetlands along the lowest terrace also produce a variety of waters

PAGE 30

Figure 7 Subsurface Profile ; Sinkhole D222 Meters Figure 7 Subsurface Profile, Sinkhole D222 The stratigraphy uncovered by excavation in Sinkhole D222 shows a crosscutting relationship The position of organic material within the throat of the sinkhole is clear evidence of downward movement of surface materials The expression of this karst feature is more pronounced in the subsurface than on the surface 2 2 Key 1=organic rich soil 2= loess 3= till 4= leached organic material 5=sandstone boulder 6=glacial erratic 7= open drain out of sinkhole Dashed line is limit of excavation

PAGE 31

small plants are concentrated in leaf litter In total, the loess was about 0 .75 meters thick on the east end of the transect, and over 1 meter thick on the west end The loess was underlain by dark red sand-rich till The till was often separated from the loess by a lag of fist-sized well-rounded crystalline rocks The till contained both locally derived sandstone (more angular) and glacial erratics (mostly basalt) The contact between loess and till appears to be unconformable This crosscutting material mentioned earlier was gray and organic rich, apparently an older version of the organic-rich duff ; it was lighter in color than the modern leaf litter, and the organic material was less intact It also contained the highest concentration of roots and small amphibians (frogs and salamanders) Furthermore, this area also contained many large rocks, of which most were large angular blocks of Hinckley Sandstone, although there was a significant portion of well-rounded crystalline glacial erratics Excavation Description Sinkhole D144 Sinkhole D144 was approximately 2 .5 by 2 meters in diameter and was located north of D222, also directly east of the Kettle River (section 13, Partridge Township, Figures 4 and 8) This sinkhole cluster is near the intersection of Log Drive Creek and the Kettle River The north-south trench across Sinkhole D144 was 4m long and had a maximum depth of 1 .75m (Figure 9) In this sinkhole, only two sedimentary layers (an organic rich duff, and dark red till) were mapped The top stratum was a black organic-rich material, composed of recent leaf litter and abundant shallow roots, as well as bigger tree roots This layer was about 20cm thick outside of the sinkhole drain In the sinkhole, the recent organic material formed the entire drain funnel, and extended all the way to the bottom of the excavation In D222, the funnel appeared to contain older, leached organic material The organic duff was directly underlain by dark reddish brown sand-rich, till This till contains more clay than D222 and D355, both of which are south of the AskovLookout Tower ice margin D144 is on the north side of that moraine The till contained both locally derived rocks (angular sandstone pieces) and glacial erratics (rounded basalt cobbles), but there were not many large rocks of any kind All of the Hinckley boulders were found inside or close to the organic funnel Excavation in this sinkhole did not reach 2 3

PAGE 32

. Figure 8 Location of Sinkhole D144 Figure 8 Sinkholes in the vicinity of Log Drive Creek, in Banning State Park near the Kettle River See Figure 2 for the location of this map area Sinkhole D144 is shown in cross section in Figure 6 The labeled UTM grid uses the 1983 North American Datum 2 4

PAGE 33

Figure 9 Subsurface Profile, Sinkhole D144 2 5 Figure 9 Subsurface Profile, Sinkhole D144 The stratigraphy uncovered by excavation in Sinkhole D144 As with D222, the crosscutting relationship of organic material within the throat of D144 illustrates the mechanism by which these sinkholes form The expression of this karst feature is more pronounced in the subsurface than on the surface Key 1 =organic rich soil 2=till 3=sandstone boulder 4=glacial erratic 5=open drain out of sinkhole Dashed line is limit of excavation

PAGE 34

bedrock Excavation Description Sinkhole D127 Sinkhole D127 is located in the northeastern portion of the study area (sec 11 of Partridge Township, Figures 4 and 10) Unlike other mapped sinkholes, this was obviously the site of a recent collapse : a 1 .3 by 1 .4 meter topsoil plug had dropped up to 80 cm below the flat ground surface The grass on the displaced topsoil was still alive and the sides of the hole showed fresh dirt In addition, the topsoil was undercut on the north and west sides, so that the actual sinkhole had a diameter of 2 by 1 .8 meters This recent collapse is located 46 meters east of a well-established stream sink (D126) which drains a swampy closed depression approximately 13 acres (0 .05 km 2 ) in area (Figure 10) Since these sinkholes are close together relative to other karst features, it is assumed that the same fracture controls both A trench was oriented N-S across this sinkhole, roughly perpendicular to the assumed orientation of an underlying joint The 3 meter long trench was excavated by hand to an average width of 1 .7 meters, and a depth of 2 .5 meters (Figure 11) The excavation revealed six matrix sedimentary strata (an organic rich duff, mixed sand and aeolian loess, mature reddish brown sand, dark reddish brown clay, immature gravelly sand and a dark red till) all of which had been locally disturbed by the initial sinkhole collapse The top layer was a very dark grayish brown organic-rich material, which is composed of dead grass, soil, and abundant roots ; it ends at a well-defined A-horizon This layer was uniformly 20 cm thick across the entire excavation The organic-rich layer overlies a 25 cm thick layer of loess and sand There were few roots in this light yellowish brown material The sandy loess has gradational contacts with both overlying and underlying strata The sandy loess was underlain by one meter of dark brown sand This sand is mature, uniform and well sorted There were no rocks and only a few small gravel lenses This sand bears a strong resemblance to Hinckley Sandstone The "Hinckley" sand was underlain by dark red clay The clay layer was an average of 15 cm thick, although its thickness and orientation was variable within the excavation 2 6

PAGE 35

Figure 10 Location of Sinkhole D127 Figure 10 Sinkholes and streamsinks in two composite (glacial and karst) sinks See Figure 2 for the location of this map Sinkhole D127 is shown in cross section in Figure 8 D126 and D327 are streamsinks that appear to drain the two large closed depressions of glacial origin that are lightly shaded Closed depresssion 326 was mapped as a sinkhole, and is the only excavation that was not a sinkhole The labeled UTM grid uses the 1983 North American Datum 2 7

PAGE 36

2=loess and tine sand 3=mature medium grained sand 4=clay, some gravel and sand 5=gravely sand, some clay 6=till 7=old bedrock surface 8=in-situ weathered bedrock 9=bedrock and boulders 10=fracture in bedrock 11=soil borings Dashed line is limit of excavation Figure 11 Subsurface Profile, Sinkhole D127 The stratigraphy uncovered by excavation in Sinkhole D127 Stratigraphic unit 3 has been vertically displaced at least five meters downward into the underlying bedrock fracture Rip-up clasts of units 4, 5, and 6 were present within the collapse funnel defined by 3, attesting to the suddent formation of this sinkhole 28

PAGE 37

The sand-clay contact appears to be unconformable The clay layer was underlain by more sand, but this sand was immature, and poorly sorted, with frequent lenses of gravel or clay The immature sand was underlain by dark red, clay and sand-rich, rocky-to-gravely till The till contained both locally derived rocks and glacial erratics This contact appears to be unconformable The top of the till was very damp the day after a short rainfall, while the overlying sands were dry, because the till is much less permeable than the overlying sands This sinkhole was excavated by hand within safety limits, without reaching bedrock Two months later, the Minnesota Geological Survey's truck-mounted Giddings Probe was used to investigate the relationship of this clear collapse sinkhole to the underlying bedrock Bedrock outside of the sinkhole collapse area is about 5 meters deep (see Figure 11) Within the throat of the sinkhole bedrock was deeper, forming an asymmetrical funnel in the rock surface, which gets to 8 meters depth before narrowing into a much smaller area, ostensibly a fissure Excavation Description, Sinkhole D355 Sinkhole D355 is located south of the town of Askov, near the southeastern end of the municipal sewage lagoons (sec 29, Partridge Twp ., Figures 4 and 12) The natural closure on this feature was 10 by 12 meters, with a main sinking point that measured 1 .5 by 2 meters, half a meter deeper than the bottom of the main depression There were two secondary drains, each less than half a meter across, and less than 1 meter deep With the aid of a backhoe, a trench was excavated across this feature oriented roughly N-S, to a length of 9 meters and a depth of 4 .5 meters (Figure 13) The excavation reached bedrock and encountered five sedimentary layers This feature is located low in the landscape, lying within a shallow stream bottom The water table is fairly close to the surface ; in fact, the seasonal high water level is above the ground surface For this reason the till was wet, and the sidewalls of the trench slumped continuously throughout the excavation, making it difficult to construct a detailed cross-section These slumps also obscured the bedrock at the base of the trench, but I was still able to observe two bedrock fractures The top stratum was a black organic-rich material composed of recent dead grass and leaves, roots, and soil It was about 25cm thick across the sinkhole, and nearly 45cm 29

PAGE 38

Figure 12 Sinkholes near the town of Askov See Figure 2 for the location of this map Sinkhole D355 is shown in cross section in Figure 9 The labeled UTM grid uses the 1983 North American Datum 30

PAGE 39

Figure 13 Subsurface Profile, Sinkhole 355 Figure 13 Subsurface Profile, Sinkhole D355 The stratigraphy uncovered by excavation in sinkhole D355 Water was visibly moving through the soil into underlying bedrock fractures during the excavation The visible depression in the figure was one of several smaller drains in a much bigger sinkholes (>15 m diameter) 3 1 Key 1=organic rich soil 2= leached organic material 3= laminated lake sediments 4= till 5= till with organic material? 6= sandstone boulders 7= joints in sandstone bedrock 8= groundwater path Dashed line is limit of excavation

PAGE 40

thick in the sinkhole drain This was underlain in places by laminated fine silt and sand, which was up to 1 meter thick The rest of the excavation was in till and gravel, until reaching bedrock at about 5 meters' depth There were several large slabs of sandstone lying on the bedrock surface There were several striking features in this excavation The first was the hourglassshaped material extending from below the primary drain This leached organic material cross cuts other strata, like the funnel in the D222 excavation . This hourglass persisted to about 1 .5 meters beneath the drain, where it gained a significant amount of gravel, but remained gray in color, in contrast to the surrounding till, which was dark red The shape of the gray gravel-rich deposit was complex beneath the hourglass organic material it reached downward for another meter, while it formed a `tail' higher to the south, thinned out, then got thicker again about two meters to the south The gray material here reached almost to bedrock, but by this depth, the slump collapses interfered with our observations In this second bulge, the drain gravel was joined by another gravel stringer, this one carrying water The current ground water path is indicated on the figure by a black line with arrows The lowest point in the entire gray gravel deposit is located roughly above ,what appeared to be a joint during excavation Excavation Description, Closed Depression 326 This large closed depression is located in the northwestern portion of the study area in an actively cultivated field (section 11 of Partridge Township, Figures 4 and 10) The closure on this feature is roughly 30 x 45 meters across, and up to three meters deep This feature had an area about 40cm across and 3-4 in long showing bare sandstone This area was about eight meters east-southeast of the center of the depression and appeared to be a bedrock fracture with boulders piled into it (Figure 14) Initially, rocks were removed by hand to a depth of about 1 .5 meters In hopes that this was a sinkhole drain, several trenches were excavated in closed depression 326 These trenches were excavated with a backhoe The excavations revealed that this feature is not a sinkhole It appears to be entirely glacial in origin Two excavations were made in the east and southeast sides On the east slope, sedimentary deposits exposed in excavation suggest that a ridge 3 2

PAGE 41

Key 1 =organic rich soil 2=medium sand 3=peat 4=sand, gravel and small rocks 5=sandy till 6=white clay 7=Sandstone and erratic boulders Dashed line is limit of excavation FIGURE 14 SUBSURFACE PROFILE, CLOSED DEPRESSION 326 Figure 14 Subsurface profile through the east side of a large closed depression in the study area This feature did not have any of the characteristics previously excavated sinkholes, such as cross cutting stratigraphy or a collapse funnel It appears to be of entirely glacial origin

PAGE 42

of large boulders forms the high east side M the depression (Figure 14) A layer of till with lenses of clay, sand, gravel and even peat covers the boulders The deposits on and around this pile of boulders were irregular and discontinuous The peat was located in the southeast end of the first trench and appears to be the remains of a small swampy area or post-glacial lake It was overlain by a swath of clean pink-orange sand, which varied from <5 to 50 cm thick over 3 meters' length The soil here was very thick over 50cm thick Below the peat was a mixture of sandstone rocks and sand The middle of the first trench went through a part of the main boulder pile, and the far north end of the first trench was primarily composed of reddish till, with a few sandstone boulders and a layer of pure white clay This till seems to be mostly on the sink-side of the boulder pile, and onlapping it The strip of sandstone exposed on the southeast side of the depression appears to be a local high spot on the boulder ridge, which extends into the depression The arrangement of boulders was chaotic, but the overlying sediments were generally flat lying, with no indication of sinkhole-type cross cutting relationships The west side is lower and does not seem to be related to the boulder ridge The center of the feature is characterized by uniform, flat-lying sand and soil (Figure 15) The dark organic-rich soil was about 20cm thick, and was uniform along the 17-meter long trench There was a visible plowing horizon at 50-80 cm depth The plowing horizon was also fairly horizontal Below the plowing horizon were two glacial sediments : to the south was a reddish brown sandy till with small rocks This till made a gradational contact about halfway along the trench to the north to a tan sandy clay with some glacial cobbles and sandstone rocks I D 2 Streamsinks Twenty-four streamsinks have been mapped in the area (Table 2) Streamsinks are the termination or sinking point of streams The streams may be either ephemeral or continuously flowing During dry periods streamsinks can be difficult to distinguish from sinkholes Additional streamsinks exist in the area The "Big Sink" south of Askov Township is a good example A stream enters this large closed depression (-r 100m diameter), but no streams exit the Big Sink Two distinct sink points were mapped within the sinkhole, and others may exist The stream has some natural reaches, but in other 3 4

PAGE 43

FIGURE 1 5 SUBSURFACE PROFILE, CLOSED DEPRESSION 326 Key 1=organic rich soil 2=tan sandy clay with a few rocks 3=plowing horizon 4=reddish brown till 5=sandstone boulder 6=tan sandy clay, more rocks Dashed line is limit of excavation Figure 15 Subsurface profile through one section of a large closed depression in the study area This feature did not have any of the characteristics previously excavated sinkholes, such as cross cutting stratigraphy or a collapse funnel It appears to be of entirely glacial origin The boundary between units 2 and 3 appears to be a plowing horizon This trench was oriented roughly north-south across the floor of the closed depression

PAGE 44

places has been turned into a county ditch The original stream may have had only intermittent flow, but now has almost continual flow, due to contributions from the Askov municipal sewage lagoons I D 3 Springs Springs are the natural outlets of the groundwater flow systems and return groundwater to the surface 31 springs were mapped in this area (Table 3) Every spring in the area has not been located Many additional springs exist and can be found Most of the springs are located along the Kettle River, which is an important base level control for the area containing karst features There appear to be several types of springs ; one type produces young water and shows impact from human activities This type of spring can be fed by water from either Quaternary or sandstone aquifers Another spring type produces old water that is relatively unaffected by human activities, but has complex flow paths These springs appear to only issue from the sandstone The mapped springs cluster along the Kettle River, with a few along the St Croix River The lack of springs along the St Croix River is an artifact of field activity : much more time was devoted to finding springs along the Kettle, near the karst area Springs are discussed further in Chapter II (Water Chemistry) I D 4 Caves There are several caves in the Hinckley Sandstone in the study area The best known is Robinson's Ice Cave (also known as Bat Cave or Big Cave), located in Banning State Park due east of the City of Sandstone (Figure 16) The plan view in the upper half of Figure 16 shows the horizontal shape of the cave Labeled cross sections show the passage shape at various locations As illustrated by cross sections, the ceiling in this cave is very flat because sandstone blocks have fallen from the ceiling along horizontal bedding surfaces The profile view in the lower half of Figure 16 shows the vertical shape of the main cave passage, looking north The profile also illustrates the very flat ceiling and bedding plane effects The cross sections also demonstrate that the cave walls are vertical and straight The walls of this cave are formed by two near-parallel vertical, joints in the sandstone Near

PAGE 45

Figure 16 This cave is formed in parallel enlarged fractures in Hinckley Sandstone The cave, along with several others, is located in Banning State Park They align with a major sinkhole array located across the Kettle River 37

PAGE 46

38 its entrance the cave is up to seven meters wide, but narrows to an impassible crack 63 meters in, where the fractures converge The cave occupies the space between the two main joints If the orientation of these major fractures is projected across the Kettle River, they line up with a dense linear cluster of sinkholes (the Hells' Gate sinkhole array) It is also located near the top of the valley wall There are a number of sinkholes above the cave, which actively drain several wetlands in between the valley wall and State Highway 23 The floor of the cave is entirely covered by sandstone blocks, which have fallen from the ceiling Three observations can be made about the floor debris! the first is that natural forces have removed much of the rock debris to create the open space The second observation is that a 'solid floor cannot be seen In other words, the depth of the debris pile (and cave) is unknown, possibly as deep as the bedrock incision below the Kettle River Third, there is no dirt covering the sandstone blocks River action and river flood events should have deposited sediments in the cave The lack of sedimentary deposits implies that the top layer of blocks have fallen since the Kettle incised below this level Other caves have been found in the Kettle River Valley Fourteen smaller sandstone caves have been reported in the study area (Matt Kramar, written communication, 2001) Of these reported caves, eight cave maps have been prepared (Figures 17, 18, 19, 20, 21, 22, 23, and 24) All of the caves are developed in the Hinckley Sandstone Most of these caves show a similar structural control on development that the Robinson's Ice Cave Map shows Bedrock fractures provide an initial weakness and flow path The caves form where significant joints intersect nearly horizontal bedding that has high permeability or is mechanically weak Hell's Gate : Riverview Cave (Figure 17) has a pronounced bedrock fracture along the back wall, while the roof and floor show lithologic/bedding plane control The predominantly flat horizontal surfaces shown in cross section exhibit the importance of bedding plane control on cave development There is one less resistant (or more permeable) bed in this area that has been removed to form the cave Old Military Road Cave (also known as Rock Dam Cave and Rustler's Bend Cave) is developed next to and parallel to the Kettle River (Figure 18) Fracture control on cave

PAGE 47

Figure 17 Hell's Gate : River View Cave is developed in the Hinckley Sandstone along the Kettle River The river is shown in the plan view The back wall of the feature is controlled by a bedrock fracture, while the roof and floor show lithologic/ bedding plane control, as shown in the profile and cross sections The cave walls have sinuous shapes reminiscent of solutional passage shapes seen in typical limestone caves 39

PAGE 48

Figure 18 Old Military Road Cave is developed in the Hinckley Sandstone along the Kettle River Fracture control on cave development can be seen around the locations of A and j cross sections The cave walls are generally sinuous, similar to typical limestone cave walls 40

PAGE 49

development can be seen at either end of the cave, near the locations 'of cross sections A and J The rest of the cave is composed of walls that are generally sinuous ; similar to many of the passage walls found in typical phreatic limestone caves Lonesome Cove (or Boy Scout Cave) has a back wall that shows significant structural control on cave development (Figure 19) There are two joint sets that shape this wall : NW-SE and NE-SW The passage walls themselves show both flat joint faces and also more sinuous shapes that have been carved into the joint faces The bedrock column in the center of the cave also has curvilinear walls Bear Cave (Figure 20) shows very strong fracture control on development This structural control is overprinted by the curvilinear forms found in many of the other caves The major fracture in this cave runs N-S, and can be seen in the shape of the entrance drip line The cave axis is parallel to this major fracture A smaller set of fractures oriented NW-SE appears as protrusions and indentations on the passage walls Bedding control on horizontal surfaces is much less apparent in this cave than many others ; this is because the structural and lithologic control has been overprinted by solutional forms Like many of the caves, this feature has undergone significant collapse, but much of the collapsed material is missing Natural Bridge Cave (Figure 21) is a remnant of a previously larger feature The remaining arch has a span of almost six meters and is about 1 1 /2 meters across at its thinnest point The upper and lower surfaces of the arch show bedding plane control on development Hovdingbroen Cave (also known as Chieftan Bridge or the Great Arch) is another remnant of a larger feature (Figure 22) The arch has a span of over 6 meters and is about 3 meters across at its thinnest Horizontal surfaces still show developmental control exerted by bedding planes Porcupine Cave (Figure 23) exhibits fracture control in the narrow bedrock crawlway along the back wall of cave (near cross section E) The roof shows bedding plane control while the cave walls have more, sinuous shapes reminiscent of the solutional passage shapes seem in typical limestone caves Wolf Creek Breakdown Cave (Figure 24) is formed by an erosional block balanced on several bedrock pillars The morphology of the bedrock pillars shows lithologic 41

PAGE 50

Figure 19 Lonesome Cove/Boy Scout Cave is developed in the Hinckley Sandstone along the Kettle River The back wall of the cave shows structural control from two sets of fractures : NW-SE and NE-SW The passage walls themselves show both flat joint faces and also more sinuous shapes commonly found in limestone caves 42

PAGE 51

Figure 20 Bear Cave is formed in the Hinckley Sandstone along the Kettle River Structural control can be seen in the fracture running parallel to the cave axis in the roof ; its continuation in the floor is obscured by debris The cave walls are curvilinear, reminiscent of typical limestone cave walls Another set of fractures (NW SE) exert developmental control to a lesser extent 43

PAGE 52

Figure 21 Natural Bridge Cave is developed in the Hinckley Sandstone along the Kettle River The river is shown in the plan view The upper and lower surfaces of the archof this bridge show lithologic/bedding plane control 44

PAGE 53

Figure 22 Hovdingbroen Cave is developed in the Hinckley Sandstone in the Kettle River Valley The arch is a remnant of a once-larger feature Horizontal surfaces still show developmental control exerted by bedding planes 45

PAGE 54

Figure 17 Porcupine Cave is developed in the Hinckley Sandstone in the Kettle River Valley Fracture control along the back wall of the feature can be seen in the development of a narrow crawlway (cross section E-E') The roof shows lithologic/ bedding plane control while the cave walls have more sinuous shapes reminiscent of the solutional passage shapes seen in typical limestone caves 4 6

PAGE 55

Figure 24 Wolf Creek Breakdown Cave is developed in the Hinckley Sandstone, in the Kettle River Valley The enclosed space of the cave is formed by an erosional block balanced on several bedrock pillars The morphology of the bedrock pillars shows primarily lithologic control, as does the floor of this feature The shape of the ceiling block is probably constrained by bedrock fractures 47

PAGE 56

control, as does the floor of this feature The shape of the ceiling block is probably constrained by bedrock fractures I D 5 Composite Glacial and Karst Features A final class of features have both glacial and karst characteristics If a karst system is developed in a glaciated landscape, it is inevitable that there will be features with characteristics of both systems In Pine County, these composite glacial karst features are typically large closed basins of presumed glacial origin, which drain internally to one or several specific sinkholes Sinkhole D126 is the drain of one of these larger composite features (Figure 10) Several other sinkholes have been mapped in this glacial depression, but D126 appears to take the most water There is a well-defined dry streambed leading into a pile of clean-washed boulders D327 is another drain to a large glacial depression, creating a composite feature I E Interpretations I E 1 What is not a sinkhole? What is a sinkhole? A sinkhole is a hole in the ground formed from below, but the removal of material By the nature of its formation, a sinkholes transmits water and material quickly below ground What is not a sinkhole? A hole formed by the removal of material from above is not a sinkhole While there are many karst sinkholes in central Pine County, there are also many closed depressions that are not sinkholes Like sinkholes, these holes come in a variety of shapes and sizes Common examples are the holes left by fallen trees (Figure 25) In virgin forest . you find whole root structures tipped into the air and looking like radial engines As you will nowhere else, you find the topography of pits and mounds In its random lumpiness, it could be a model of glacial terrain When a tree goes over and its roots come ripping from the ground, they bring with them a considerable mass of soil When the tree has disappeared, the dirt remains as a mound, which turtis Kelly green with moss Beside it is the pit that the roots came from When no other gtrace remains of the tree, you can see by the pit and the mound the direction in which the tree fell, and guess its approximate size (McPhee, 1997) 4 8

PAGE 57

Figure 20 Comparison of sinkholes and tree fall rip-up holes The sinkholes are formed by the removal of material from below and have cross cutting stratigraphy and a drain The rip-up holes are fomred by the removal of material by tree roots from aobve, and develop a layered stratigraphy as the hole fills in

PAGE 58

5 0 The same criteria can be applied to man-made holes Holes that are not formed by collapse or piping (the way sinkholes form), must be formed by moving material up and out of the hole For efficiency, this material is usually put as close as possible to the excavation Thus, man-made holes are characterized by nearby piles of dirt Glacial activity can create a variety of closed depressionsover a wide range of scales that are not sinkholes Glaciers make closed depressions both through deposition and erosion A common type of depositional glacial depression is the "prairie pothole" These depressions form when a retreating glacier leaves blocks of ice mixed with sedimentary deposits When the ice melts, it leaves a hole in the unconsolidated material surrounding it The difference between a purely glacial basin and a composite glacial /karst feature can be hard to determine, especially when both exist in the same area Closed Depression 326 was originally mapped as a sinkhole within a glacial basin It is in an area that has both sinkholes and composite features Several features in this area are unambiguously karst features However, closed depression 326 is not In the portions of this basin that were excavated, the stratigraphy was quite different than in all of the excavated sinkholes There were no obvious drain points, no cross cutting relationships, no downward moved funnel of recent organic debris It appears that the shape of the land surface is primarily controlled by a large pile of boulders, which forms a roughly north/south ridge on the east side of the depression On top of the ridge, soil cover was very thin Below the horizontal sediments was an irregular surface of sandstone boulders Toward the center of the depression, the soil was much thicker and plowing horizons were apparent As stated in the composite features section, karst sinkholes can develop in closed depressions of glacial origin However, the development of sinkholes will depend on having shallow depth to bedrock and well-developed bedrock fractures Sinkhole development also depends on the composition of unconsolidated deposits on top of the bedrock These unconsolidated sediments have a lot of variability, even at meter-scale This variability is demonstrated by the area shown in Figure 9, where there are several closed depressions on glacial origin ; several of these basins are drained by sinkholes or streamsinks (D126, D327), while others nearby are not

PAGE 59

I D 2 Sinkhole Distribution Relative to Bedrock Type The Hinckley Sandstone underlies all of the mapped sinkholes, and all sinkholes occur northwest of the Hinckley fault Most sinkholes lie within five kilometers of the fault, and none have been mapped more than seven kilometers from it (Figure 26) Southeast of the Hinckley fault, the Hinckley Sandstone is very thin, and the underlying Fond du Lac Formation is close to the surface (Figure 27) No sinkholes have been found southeast of the Hinckley Fault Because the Hinckley Sandstone is not as thick in this area and somewhat feldspathic, the joints may not penetrate as deeply as they do to the northwest of the fault, where the sandstone is thicker In addition, the Fond du Lac Formation has a higher content of clay and silt that are more likely to clog a developing fracture system In contrast, the Hinckley Sandstone northwest of the Hinckley Fault is composed almost entirely of quartz, and hence, lacks clay or minerals that alter to clay The Hinckley Fault forms a southeastern boundary to the sinkhole area The same argument may pertain to a northwest boundary to the karst area as well As can be seen in the cross sections in Figure 27, sinkholes have been mapped where the Hinckley Sandstone is thickest Southeast of the fault, the Hinckley is very thin and no sinkholes have been found, despite extensive searching To the northwest, there is not such a well-defined edge to the karst This area has not been well searched, so the absence of sinkholes may simply be an artifact of field activity However, based on the fracture model described above, sinkhole formation would be unlikely near the Hinckley-Fond du Lac contact northwest of the study area Sinkhole formation is unlikely because the same part of the section forms the bedrock surface as between the Hinckley and Douglas Faults to the southeast Karst development in general is usually focused along fractures and bedding planes and at their intersections In outcrops along the Kettle River, the Hinckley Sandstone contains enlarged vertical joints, caves, and horizontal layers with small conduits If these features persist in the subsurface away from the valley, they would provide ideal flow paths for the rapid movement of water from the surface into the subsurface Sinkholes form when shallowly buried bedrock develops openings of sufficient size to allow surface material to collapse or be washed downward into the openings 5 1

PAGE 60

Figure 26 Distribution of sinkholes over bedrock geology in Pine County Sinkholes only occur over the Hinckley Sandstone, northwest of the Hinckley Fault, where the Hinckley sandstone is thickest Location of cross sections in Figure 22 are shown by heavy lines Bedrock geology map based on Boerboom, 2001 52

PAGE 61

Figure 27 Bedrock cross sections through the sinkhole area Pmhn = Hinckely Sandstone, Pmfl = Fond du Lac Sandstone, Pmcb = Mesoproterozoic Basalt The location of these cross sections is shown on Figure 13 ; they are parallel, both oriented roughly NW SE MSL= mean sea level Bedrock data based on Boerboom, 2001

PAGE 62

5 4 Sinkholes located immediately northwest of theHinckley fault appear to drain to wetlands located on the southeast side of the fault, where the landscape is lower and wetter On the west side of the sinkhole array, numerous springs occur along the Kettle River and are probably fed by flow related to the surrounding sinkholes ; however, the connection between the feeder sinkholes to the springs remains to be demonstrated and the groundwater chemistry indicates the presence of multiple water types I D 3 Sinkhole Distribution Relative to Depth to Bedrock The thickness and composition of glacial drift are also important in determining where sinkholes are likely to form In places where the glacial drift is thin, surface water is able to move relatively quickly into the underlying bedrock Areas of high transmissivity within the glacial drift such as sand lenses, gravel lenses or boulder concentrations also permit the surface water to move downward rapidly Thus, areas with thin and sandy, highly permeable drift underlain by fractured bedrock are most favorable for sinkhole development (Figure 28) In the area where the sinkholes have been mapped, the glacial drift is sandy and contains an abundance of large boulders and slabs of locally derived Hinckley Sandstone In many places these are piled on top of one another in a fashion that produces many large openings through which surface water can rapidly move These conditions could promote periodic flushing of fine-grained sediment and the development and maintenance of sinkholes In this way, piles of boulders near joint openings may act as strainers : allowing fast moving water and its sediment load to pass through at some points and blocking it at others This process may prevent clay and other fine-grained till transported by water from uniformly clogging the joints I D 4 Sinkhole Distribution Relative to Glacial Features The sinkholes generally lie near the Askov-Lookout Tower ice margins from the Superior Lobe of the Wisconsinan glaciation, but the causal relationship between the moraines and sinkhole distribution is uncertain (Figure 29) High volumes of unsaturated water resulting from glacial discharge may have allowed solution of quartz (and therefore bedrock) to proceed faster than at normal rainfall levels Discharge of glacial melt water

PAGE 63

Figure 28 Sinkhole distribution over depth to bedrock The Pine County Sinkholes are associated with relatively shallow depth to bedrock Thin overburden allows water to reach bedrock flow systems more easily Metric scale is converted from original scale in feet 55

PAGE 64

off the front of the moraines may also have cleared fine-grained sediment from joints in the underlying bedrock Such open joints could subsequently to act as subsurface conduits over which the sinkholes could form The movement of surface sediment into the underlying fractures/conduits is an ongoing process as demonstrated by the recent collapse of sinkhole D127 The Superior Lobe's Sandstone, Askov-Lookout Tower, and Kerrick phases (or ice margins) fall within the study area, as shown in Figure 29 Sinkholes occur in several locations relative to the surficial geology : the majority lies in the Askov-Lookout Tower phase of the Superior Lobe This area is characterized by sandy glacial sediments and outwash sand and gravel Another group of sinkholes occurs in similar sandy glacial sediments behind the Sandstone ice margin, just north of the City of Sandstone A third group of sinkholes occurs in the more clay-rich glacial sediment between the AskovLookout Tower and Kerrick ice margins, near the junction of Log Drive Creek and the Kettle River Modern distribution of sinkholes occurs over a variety of sedimentary deposits, in several different Superior Lobe phases There does not appear to be a strong correlation between modern sinkhole locations and any specific surficial feature at map-scale, although there are some general correlations between glacial features and bedrock geology The correlation to bedrock geology and bedrock depth is much stronger As demonstrated by excavation, unconsolidated glacial deposits must play an important role in the development of individual sinkholes, but I think that the important constraint is variation in the glacial deposits at meter scale (in response to the driving force of fracture flow in the bedrock), not map scale A final important point in the relationship of sinkholes and glaciation is that this area has been glaciated many times since the beginning of the Cambrian The age of the conduits is unknown Glacial activity may not only keep the fracture system open by periodic flushing with melt water, it would also add and remove overburden in which the sinkholes are developed, so that the sinkholes are 1) ephemeral at a geologic time scale and 2) possibly shifting position through time 5 6

PAGE 65

Figure 29 Sinkhole Distribution over Superior Lobe Phases The Sandstone and Askov-Lookout Tower Phases are shown by bold lines These ice margins lie close to the mapped sinkholes The rest of the Superior Lobe phases are shown by thin lines The Grantsburg sublobe of the Des Moines lobe did extend into southernmost Pine County, but the Grantsburg sublobe phases are not shown Geological interpretations after Patterson and Knaeble, 2001 Sinkholes are shown by triangles 5 7

PAGE 66

I D 4 Sinkhole Formation Models The sinkholes in Pine County formed when turbulent surface water sank through the overburden and transported fine-grained sediment into an underlying system of open joints in the Hinckley Sandstone This soil piping eventually causes the land surface to subside or collapse Once formed, sinkholes are fed by surface runoff areas and can drain large volumes of water during spring snowmelt and after heavy rain, at which time large whirlpools may form and water can be heard running underground Several of the sinkholes serve as the terminal sinks of perennial or ephemeral surface streams The seasonal high water table is in or above some of the sinkholes, and they are drowned during periods of high water It is unknown whether the drowned sinkholes continue to accept water during these floods or if the flow reverses and they act as estavelles and intensify local floods Our excavations indicate that there are two modes of sinkhole formation : collapse and subsidence The collapse features are illustrated by the excavations of sinkholes D222, D144 and D127, while the subsidence features are illustrated by D355 Sinkhole formation by collapse : infiltrating water moves material into open fractures in the sandstone bedrock, leaving a void in the dirt over the fracture The process continues and the void enlarges until collapses (Figure 30) The collapse may work its way progressively to the surface in a series of failures, or the surface may collapse in one catastrophic failure This sequence can be seen in Figure 11 The area surrounding the sinkhole has relatively flat-lying sedimentary layers, which are consistent across the whole area (verified with other soil borings) The sequence is only different within the sinkhole Furthermore, the sediments in the sinkhole are disrupted and mixed, indicating that they have been physically moved Specifically, higher strata have been moved downward, while lower strata are reduced or are completely missing Collapse is the simplest explanation for this geometry Human activities or natural processes such as tree fall would produce a different stratigraphy Although the excavations of D144 and D222 were not as extensive, they both follow the same pattern : shallow sediments are being moved downward, into the subsurface D144 probably represents a more advanced stage of the process seen in D127 : the initial collapse created a sinkhole, which continues to slowly move material downward 58

PAGE 67

Figure 30 Sinkhole formation by collapse : infiltrating water moves material into open fractures in the sandstone bedrock, leaving a void in the dirt over the fracture The process continues and the void enlarges until collapses The collapse may work its way progressively to the surface in a series of failures, or the surface may collapse in one catastrophic failure As time passes, none of the original collapse features are left, due to the ongoing activity of the sinkhole 59

PAGE 68

None of the original collapse features are left, due to the ongoing activity of the sinkhole The material in the throat of the sinkhole was mostly relatively unweathered organic material, which indicates that the sinkhole is active Sinkhole D222 is an even older version of this type of sinkhole, where the collapse funnel has stabilized and an open hole (up to two inches in diameter) now drains the sinkhole and continues to transport sediment In this case, the organic material deeper in the original collapse funnel has weathered to a paler gray because this part of the sinkhole is no longer moving Figure 31 illustrates several important points about the formation of sinkhole D127 First, the surrounding sedimentary layers are flat lying and relatively uniform in thickness Soil borings outside of the excavation area confirm that these strata occur at similar depths all around the sinkhole, up to 45 meters away For this reason, the sinkhole is not simply a misidentified irregularity within disrupted and jumbled strata A second important point is how well defined the collapse funnel (A) is The edges of the collapse zone are distinct ; while this is obviously a catastrophic collapse (as shown by the rip-up clasts shown in the lower portions of the excavated funnel), the disruption is very localized From these sedimentary observations, I conclude that the driving force of sinkhole formation (fracture flow in the bedrock) must exploit some irregularity or weakness in the overlying sediments (i .e ., a thin spot in the till) to create a sinkhole on the surface Furthermore, when a sinkhole does form, it is a focused collapse Afterward, the open sinkhole continues to move material slowly into the subsurface, in the manner of D222 and D144 Sinkhole formation by subsidence : in these sinkholes, infiltrating water moves material into open fractures in the sandstone bedrock, and the unconsolidated material above is loose enough to continually settle, or only fine grains are removed out of a high porosity deposit, such as very clean coarse gravel or boulder pile (Figure 32) The excavation in sinkhole D355 show the best example of this process This sinkhole is larger (about 10 meters across) The movement of material into these fractures has affected the geometry of the sedimentary deposits directly below the drain there is a cross-cutting sinkhole funnel, and deeper there is a distinct `channel' of gravel bearing organic material Interflow or very shallow groundwater was moving though part of the main gravel deposit, 6 0

PAGE 69

Figure 31 3D Geometry of Sediments, Sinkhole D127 61 Figure 31 Sinkhole D127 : 3D Geometry of Sediments The east face of excavation is shown in Figure 11 Outside of the throat of this sinkhole, excavation and soil borings demonstrate that sedimentary layers are uniform, continuous and flat-lying Within the throat of the sinkhole (A), this stratigraphy is absent Sediments are disrupted, mixed and some are missing altogether The transition between the collapse zone and the surrounding strata is sharp

PAGE 70

Figure 32 Sinkhole formation by subsidence : small-scale variation in the glacial deposits provides a high permeability connection between the ground surface and underlying bedrock fractures Fine grained sediment is moved out of and through the high permeability deposit, such as very clean coarse gravel or boulder pile 6 2

PAGE 71

which has two distinctly low points, both above bedrock fractures From the morphology of the gravel deposit below the sinkhole drain, it is assumed that when the sinkhole is taking water, the surface water follows the gravel connectivity to the same point that interflow water always goes to, as indicated by the arrowed line The higher bulge of gravel directly below the sinkhole drain is interpreted as an older flow path toward the joint that was briefly examined This gravel lens extends out of the plane of the Figure to the east beneath the auxiliary drains It is assumed that the water from these auxiliary drains reaches the bedrock at this joint, to the east of the trench, in a similar fashion The presence of the large boulders near the bedrock surface such as these may serve as filters All of my sinkhole formation models involve water moving into enlarged bedrock fractures as the driving mechanism This is based on our observations in our sinkhole excavations Such fractures were observed in sinkhole D355, and the Giddings Probe borings in sinkhole D127 In D127, bedrock was at a relatively constant depth (about 5 meters) around the collapse feature, both in the soil borings shown on Figure 11, and several more that are out of the plane of the figure Eight borings were made in this area, in which the stratigraphy and depth to bedrock was fairly uniform The only place where these conditions differed was in the throat of the collapse In one boring, the probe drilled to a depth of eight meters without hitting bedrock The boring encountered only smaller rocks and the same mixed sand and till that typified the sinkhole funnel material Drilling stopped at this point because the Giddings Probe bit broke by binding in between rocks, not because it hit bedrock Materials that are consistently seen at 1 to 1 .5 meters in the borings outside of the sinkhole have been moved at least 7 meters further down into the subsurface in the sinkhole throat! Furthermore, this very deep area was small (narrow) Just inches to the side the probe was not able to enter the fissure Since the excavation and set of auger holes were oriented in a line perpendicular to a trend between D127 and the nearby streamsink D126, I expected the fissure to be narrow in this direction, and the elongate dimension to stretch between D126 and D127 There are more sinkholes beyond D127 in the woods to the east, ostensibly driven by the same large fracture 6 3

PAGE 72

I D 6 Composite Features & Associated Streamsinks The large-scale glacial/karst features discussed previously will be controlled by many of the same features as the smaller sinkholes ; bedrock type, depth to bedrock, depth of water table and variation in underlying sediment However, the initial closed basins were created by glacial advance and retreat Many of these large closed drainages can be seen on topographic maps of central Pine County They are usually aligned with moraines or travel direction of the glacier Depending on the type of sediment they are hosted in, they may hold water A feature that holds water is more likely to form catastrophic collapse sinkholes, or initiate soil-piping sinkholes Large basins in very permeable materials are more likely to simply leak into their surroundings fast enough not to initiate collapse features Figure 10 contains several examples of composite glacial/karst features D126 is a sinkhole/streamsink at the edge of one of the large elongate basins that are parallel to other glacial features and presumed to be of glacial origin There is a well-defined creek bed in the lowest part of this basin that leads down into a clean washed boulder pile in the mouth of D126 The closure of the glacial basin includes approximately 13 acres (0 .05 km) The water collected by a feature of this size could trigger sinkhole formation in an area that already has shallow bedrock with enlarged through going fractures The water collected in the watershed of this glacial basin has created the creek bed that runs into D126 Nearby sinkholes form a roughly east/west trend .' This trend is presumably the surface expression of a bedrock fracture draining all of the sinkholes This type of feature is fairly common in the study area The lower half of Figure 10 shows another example Once again, a large closed basin (about 25 acres) of glacial origin is visibly drained by a sinkhole (D327) Several local residents have described large whirlpools forming over these sinkholes during snowmelt or after large rains They also describe loud sounds of rushing water underground near the sinkholes 6 4

PAGE 73

REFERENCES Alexandrowicz, Z ., 1989 Evolution of weathering pits on sandstone tors in the Polish Carpathains Zeitshrift fur Geomorphologie, 33 : 275-289 Boerboom, T ., 2001 Plate 2-Bedrock Geological Map and Sections, County Atlas Series, Atlas C-13, Part A Minnesota Geological Survey, St Paul, Minnesota Busche, D ., and Sponholz, B ., 1992 Morphological and micromorphological aspects of the sandstone karst of eastern Niger Zietshrift fur Geomorphologie, suppl 85 : 1-18 Cannon, W .F ., Daniels, D .L ., Nicholson, S .W ., Phillips, J ., Woodruff, L .G ., Chandler, V .W ., Morey, G .B ., Boerboom, T .J ., Wirth, K .R ., and Mudrey, M .G ., Jr ., 2001 New map reveals origin and geology of North American Mid-Continent Rift Eos, 82 : 97, 100-101 Chalcraft, D ., and Pye, K ., 1984 Humid tropical weathering of Quartzite in southeastern Venezuela Zeitschrift fur Geomorphologie, 28 : 321-332 Craddock, C ., 1972 Regional Geologic Setting In Sims, P .K ., and Morey, G .B ., eds ., Geology of Minnesota : A centennial volume Minnesota Geological Survey, St Paul, Minnesota, 281-291 Davis, D .W ., and Green J .C ., 1997 Geochronology of the North American Midcontinent rift in western Lake Superior and implications for its geodynamic evolution, Canadian Journal of Earth Sciences, 34 : 476-488 Dzulynski, S ., and Kotarba, A ., 1979 Solution pans and their bearing on the development of pediments and toys in granite Zeitshrift fur Geomorphologie, 23 : 172-191 Frye, J .C ., and Swineford, A ., 1947 Solution Features on Cretaceous Sandstone in Central Kansas American Journal of Science, 245 : 366-379 Howard, A .D ., and Groves, C .G ., 1995 Early development of karst systems : 2 Turbulent Flow Water Resources Research, 31 : 19-26 Hutchinson, D .R ., White, R .S ., Cannon, W .F ., and Schulz, K .J ., 1990 Keweenaw Hot Spot : Geophysical Evidence for a 1 .1 Ga Mantle Plume beneath the Midcontinent Rift System Journal of Geophysical Research (B), 95 : 10869-10884 Jennings, J ., 1983 Sandstone Pseudokarst or karst? In R .W Young and G .C Nanson (eds .), Aspects of Australian Sandstone Landscapes Australian and New Zealand Geomorphology Group Special Publication 1, University of Wollagong, Wollagong, 21-30 6 5

PAGE 74

Klimchouk, A ., and Ford, D ., 2000 Types of Karst and Evolution of Hydrogeologic Setting In Speleogenesis : Evolution of Karst Aquifers, Klimchouk, Ford, Palmer and Dreybrodt, eds ., National Speleological Society, Alabama, 45-53 Knaeble, A .R ., Patterson, C .J ., and Mayer, G .N ., 2001 Plate 5-Quaternary Stratigraphy, County Atlas Series, Atlas C-13, Part A Minnesota Geological Survey, St Paul, Minnesota McPhee, J ., 1997 In Virgin Forest In Irons in the Fire Farrar, Straus, and Giroux, New York, 3-56 Mooney, H .M ., Craddock, Campbell, Farnham, P .R ., Johnson, S .H ., and Volz, G ., 1970 Refraction seismic investigation of the northern midcontinent gravity high Journal of Geophysical Research, 75 : 5056-5086 Morey, G .B ., 1972 Petrology of Keweenaw Sandstones in the Subsurface of Southeastern Minnesota In Sims, P .K ., and Morey, G .B ., eds ., Geology of Minnesota : A centennial volume Minnesota Geological Survey, St Paul, Minnesota, 436-449 Mossler, J .H ., 1987 Paleozoic lithostratigraphic nomenclature for Minnesota : Minnesota Geological Survey Report of Investigation 36, 36 pp Nicholson, S .W ., and Shirey, S .B ., 1990 Midcontinent Rift Volcanism in the Lake Superior Region : Sr, Nd and Pb isotopic evidence for a mantle plume origin Journal of Geophysical Research, 95 : 10851-10868 Patterson, C .J and Knaeble, A .R ., 2001 Plate 4-Surficial Geology, County ,Atlas Series, Atlas C-13, Part A Minnesota Geological Survey, St Paul, Minnesota Shade, B .L ., Alexander, E .C ., Jr ., Alexander, S .C ., and Martin, S ., 2001 Plate 6-Sinkhole Distribution, County Atlas Series, Atlas C-13, Part A Minnesota Geological Survey, St Paul, Minnesota Sims, P .K ., and Morey, G .B, 1972 Resume of Geology in Minnesota In Sims, P .K ., and Morey, G .B ., eds ., Geology of Minnesota : A centennial volume Minnesota Geological Survey, St Paul, Minnesota, 3-17 Thiel, E .C ., 1956, Correlation of gravity anamolies with the Kweenawan geology of Wisconsin and Minnesota Bulletin of the Geological Society of America, 67 : 1079 Truluck, T .F ., 1991 Deepest and Longest Caves in Africa and Southern Africa, and the Deepest Sandstone Caves in the World Bulletin of the South American Speleological Association, 32 : 99-101 Tryhorn, A .D ., and Ojakangas, R .W ., 1972 Sedimentation of the Hinckley Sandstone of east-central Minnesota In Sims, P .K ., and Morey, G .B ., eds ., Geology of Minnesota : A centennial volume Minnesota Geological Survey, St Paul, Minnesota, 416-424 6 6

PAGE 75

White, W .B ., 1988, Geomorphology and Hydrology of Karst Terrains Oxford University Press, New York, 464pp White, W .B ., Jefferson, G .L ., and Haman, J .F ., 1966 Quartzite Karst in Southeastern Venezuela International Journal of Speleology, 2 : 309-314 Wray, R .A .L ., 1997, A global review of solutional weathering forms on quartz sandstones Earth-Science Reviews, 42 : 137-160 Wright, H .E ., Jr ., 1964 Origin of the Lakes in the Chuska Mountains, Northwestern New Mexico Geological Society of America Bulletin, 75 : 589-598 Young, R .W ., 1986 Tower Karst in Sandstone : Bungle Bungle massif, northwestern Australia Zeitshrift fur Geomorpholie, 30 : 189-202 Young, R .W ., 1987 Sandstone landforms of the tropical East Kimberley region, northwestern Australia Journal of Geology, 95 : 205-218 Young, R .W ., 1988 Quartz etching and sandstone Karst : Examples from the East Kimberleys, Northwestern Australia Zeitshrift fur Geomorphology, 32 : 409-423 Zartman, R .E ., Nicholson, S .W ., Cannon, W .F ., Morey, G .B ., 1997 U-Th-Pb zircon ages of some Keweenawan Supergroup rocks from the south shore of Lake Superior Canadian Journal of Earth Sciences, 34 : 549-561 Zawidzski, P ., Urbani, F ., and Koisar, B ., 1976 Preliminary notes on the geology of the Sarisarinama Plateau, Venezuela, and the origin of its caves Boletain de la Sociedad Venezolana de Espeleogaia, 7 : 29-37 6 7

PAGE 76

II Hydrogeology and Groundwater Chemistry II A Introduction and Methods II A 1 Introduction Chapter II is concerned with groundwater in the part of Pine County that contains karst features These karst features are described in Chapter I of this thesis Recharge to karst aquifers often bypasses soil-zone processes Karst aquifers also exhibit strong heterogeneity in flow systems Due to their complex flow systems and short residence times, karst systems are often vulnerable to contamination (Smart and Hobbs, 1986) Groundwater contamination is a major concern to resource management Classification of the Pine County sinkhole area as a karst prompted an investigation into the associated hydrogeologic systems A groundwater study was conducted in the project area A total of eighty-eight samples were collected and analyzed Thirty-four of these samples were collected from waters in the Hinckley/Fond du Lac and Quaternary aquifers The samples were taken from municipal, public supply and domestic wells, and springs The wells are in the Hinckley/Fond du Lac sandstones of the study area, and the overlying Quaternary deposits Eighteen wells in the rift basalts (and overlying Quaternary deposits) east of the study area were also sampled Surface water was sampled at fourteen points along surface streams and in sinkholes Twenty-two of the thirty-two mapped springs were sampled This groundwater study was carried out in parallel with other geologic investigations, prior to the publication of the Pine County Geologic Atlas (County Atlas Series, C-13) The study evolved as new information became available from other atlas studies The analytical results of the groundwater study are summarized in Tables 5-8 II A 2 Purpose There are two points of societal interest in the study of groundwater : where sufficient water can be obtained for human needs and whether that water is safe The first issue is not a pressing one for Pine County, but all groundwater users must be concerned 6 8

PAGE 77

with the second issue Hydrogeology approaches the question of water quality by means of a group of questions that address the source of groundwater, its flow paths, residence times, what chemical reactions have occurred in transit, and its destination This groundwater study was designed as an initial reconnaissance of water quality in the study area, which is located in east-central Minnesota (Figure 1) In the study area, the hydrogeologic system includes three water bearing strata : the Hinckley/Fond du Lac sandstones in the west, Mesoproterozoic basalts east of the sandstones, and unconsolidated deposits from the Wisconsinan glaciation In the study area, the glacial deposits range from 0 60 meters thick, and overlie both sandstones and basalts All three of these strata contain significant volumes of water and have producing wells drilled into them The locations of the water samples are shown in Figure 33 Spring samples cluster along the Kettle River, with a few located along the St Croix River Only one spring was mapped at any distance from these two rivers : Partridge Creek Spring is located near the Hinckley Fault, east of Askov The Quaternary water samples are located in glacial deposits over both basalt and sandstone The Quaternary samples were taken over a relatively large area in order to sample a broad spectrum of the water being recharged to bedrock sources through drift of potentially different chemical composition No sinkholes have been mapped outside of the Hinckley Sandstone However, the water in wells and springs in the sinkhole zone is not so spatially constrained : these waters vary widely in their sources, from recent precipitation to deep connate brinest in the sandstone, to complex waters that appear to have moved into the sandstone from the basalts Samples were taken in a variety of areas in order to interpret the complicated hydrogeologic system at work below the Pine County karst features III A 3 Field Methods The specific field methods used for collecting water samples are described in detail in the Field and Laboratory Methods of the Hydrogeochemistry Laboratory, Department of Geology and Geophysics, University of Minnesota (Alexander and Alexander, 2002) t The conventional definition of brine is >100,000 ppm TDS No samples in this data set have > 500ppm TDS However, some waters appear to have mixed with very old waters, possibly connate brines This component of old water, however small, is geochemically important In this chapter, "brine" will refer to water that shows the influence of brines or connate waters 69

PAGE 78

Figure 33 Location of water samples for part III Spring samples are clustered along the Kettle River ; several more springs were mapped along the St Croix river Very little searching for springs was done along the St Croix The Kettle and St Croix both lie within deep glacially carved river valleys that control local base level Only one spring (Spring Al) is not located near either river : it is adjacent to the sinkhole array Water samples from sandstone and basalt bedrock were selected to sample across the study area at a variety of depths Samples were subject to the location of existing wells and premissions 70

PAGE 79

Temperature, conductivity and dissolved oxygen were measured in the field, at all sites Samples were collected for cation, anion and alkalinity analysis Cation samples were acidified and placed in 15 ml polypropylene bottles Opaque high-density polyethylene (HDPE) bottles were used for the 15 ml anion samples Alkalinity samples were collected in 500 ml plastic bottles Samples were transported back to the lab in coolers A GPS location was recorded from the sample site, and as much information as possible was recorded about the well or spring Several samples collected by other agencies did not have GPS locations These locations were estimated from field notes, and are noted as estimated locations in the attached data tables Temperature was measured with an ASTM 63C mercury thermometer, with 0 .1€ C divisions Conductance was measured with a Hanna DiST WP3 Conductivity/Total Dissolved Solids meter The pH of water samples was measured by an Orion SA210 pH meter and Orion 91-06 electrode The pH meter was calibrated in the field with two buffers, which had been equilibrated to water temperature Dissolved oxygen was measured by a modified Winkler titration A significant number of samples yielded dissolved oxygen contents below the detection limit of 0 .01 ppm A variety of careful sampling techniques were necessary to avoid contamination of the samples with atmospheric oxygen The reduction-oxidation (redox) potential was measured at most sites There are theoretical difficulties with redox potential These problems include the presence of mixed potentials in natural waters, disequilibrium between the redox couples in slow and irreversible reactions, and the formation of compounds on the surface of the platinum electrode (Faure, 1986) The practical difficulty is that many samples do not yield stable readings during a reasonable time period The redox measurements cannot be used quantitatively However, the measurements can provide a general estimate of the redox states of the waters, so redox measurements were recorded whenever stable values were obtained Budget constraints limited the number of residence time samples that could be analyzed Fifty-four samples were selected for enriched tritium analysis Tritium was collected in 1000ml HDPE bottles Twelve samples were selected for carbon dating Samples from which a sufficient amount of dissolved inorganic carbon (DIC) could be obtained were analyzed by liquid scintillation counting (LSC) The DIC from these samples 7 1

PAGE 80

was extracted from about 200 liters of water When is was not practical to collect the DIC from such a large volume of water, a few milligrams of DIC were extracted from liter samples and analyzed by accelerator mass spectrometry (AMS) Most carbon dating samples were collected from batches of water of about 50 gallons for analysis The carbon from bicarbonate and carbon dioxide in the water is precipitated as several liters of barium carbonate slurry, which must be settled and prepared in the lab II A 4 Laboratory Methods The specific laboratory methods used for analyzing water samples are described in detail in the Field and Laboratory Methods of the Hydrogeochemistry Laboratory, Department of Geology and Geophysics, University of Minnesota (Alexander and Alexander, 2002) Alkalinity titrations were performed by acid titration in the Hydrogeochemistry Laboratory Titrations were performed in triplicate in polymethylpentene flasks using a Hach digital titrator, using color reference standards of pH 4 .5 and 4 .8 Alkalinity was reported in ppm CaC0 3 equivalents Cations, anions and trace metals were analyzed in the Geochemistry Laboratory of the Department of Geology and Geophysics, University of Minnesota Cations and trace elements were analyzed separately by inductively coupled plasma mass spectrometry (ICPMS), on a VG Elemental PQExCell mass spectrometer Cation concentrations were reported as the mean of five replications Anions were analyzed by ion chromatography, on a Dionex DX500 chromatography system, using a GP40 gradient pump, a CD20 conductivity detector, and two AS4A columns with carbonate/bicarbonate elements Anion concentrations were calculated using the integrated peak areas All standards used in the Geochemistry Laboratory are National Institute of Standards and Technology (NIST) traceable The cation-anion charge balances for many samples were more than 5%, which is an indication of incomplete analysis The calculated charge imbalances are presumably due to the presence of humic substances Also, these waters are very dilute, which exaggerates the proportion of charge imbalances The average total dissolved solids (TDS) was 188 ppm Concentrations rarely exceeded 7 milliequivalents (meq) and the average was 5 .4 meq The average ionic strength was 0 .003 7 2

PAGE 81

Tritium samples were analyzed by the Environmental Isotope Laboratory (EIL) of the University of Waterloo, Ontario, Canada, using enriched counting techniques Tritium values are reported in tritium units (TU) Tritium units are defined as one atom of 3 H/10 18 atoms of hydrogen, or 7 .1 decays of 3 H/minute/liter of water (Faure, 1986) Since the detection limit in natural water is about 6 TU, samples with relatively small amounts of tritium (all of the samples in this study) are enriched by electrolysis about fifteen times, to yield a detection limit of 0 .8 TU +/0 .6 A detailed description of the analytical techniques used by the EIL can be found online (Heemskerk, 1998) This type of tritium analysis is also described in Taylor (1977) Carbon-14 samples were analyzed by Beta Analytic in Miami, Florida Nine samples were analyzed with liquid scintillation counting (LSC) LSC uses benzene synthesis, as described in Tammers, 1975 Three samples were collected for accelerator mass spectrometer (AMS) counting, which does not require such large samples The modern reference used is the NBS Oxalic Acid Carbon-14 Standard One of the AMS samples was unanalyzable A detailed description of the analytical techniques and interpretations used by Beta Analytic can be found online (Beta Analytic, 2002) II B Residence Time An important question in understanding any hydrogeological system is how long water stays underground, or its residence time Residence time is the time that passes between when surface water enters the ground until it is returned to the surface as, a spring or through a well Therefore, the residence time measured for any given sample is equivalent to its travel time to move from its point of origin as groundwater to the place where it was sampled Residence times vary widely, depending on the geological material hosting the groundwater Residence times in Minnesota can be as short as a few hours or as long as tens of thousands of years (Alexander and Alexander, 1989) My data set from central Pine County has examples throughout this spectrum There are a number of ways to measure residence times These methods include adding physical tracers to water, dating the water with naturally occurring radiogenic isotopes such as 3 H, 14 C, 'He, and 36 C1, measuring the Cl/Br ratio or salinity, and measuring anthropogenic contaminants 7 3

PAGE 82

II B 1 The Tritium Age Model Background One radiogenic tracer commonly used to measure the residence time of groundwater is the radioactive isotope of hydrogen ( 3 H, or tritium) Tritium is generated cosmogenically in the stratosphere as the result of bombardment of 14 N by cosmic ray induced neutrons Tritium is produced concurrently with radiocarbon and 'He Radiocarbon is the major product of reactions with 14 N in the upper atmosphere Tritium is produced by the following reactions (Libby, 1946) : 14N + n = 12 C + 3 H + Q 1 (where Q 1 =-4 .3 Mev) 14 N + n = 3 4 He + 3 H + Q 2 t (where Q2=-11 .5 Mev) The cosmogenic production rate of Tritium has been reported between 0 .12 0 .5 3 H/cm 2 sec (Kaufman and Libby, 1954 ;, von Buttlar and Libby, 1955 ; Craig and Lal, 1961 ; Nir et al, 1966) Tritium in precipitation was not well studied prior to the early 1950's, so pre-bomb concentrations are not well known Several observations and reconstructed values are available for the concentration of 3 H in natural waters before the onset of atmospheric nuclear testing Grosse et al (1951) found a concentration of about 1 TU in Norwegian surface waters collected between 1935 and 1950 Kaufman and Libby recalculated the tritium content for vintage wines from New York and France and found original tritium concentrations of 3 .0 6 .6 TU (1954) Possible fractionation of 3 H by the grape vines is not considered The recalculated value for Ottawa River water from unused reagents varied from 8 20 TU (Brown, 1961) However, the cosmogenic background is generally accepted to have been (and continue to be) a few TU in most temperate locations (Alexander and Alexander, 1989) Tritium directly dates groundwater, because 3 H is incorporated into water molecules as 1 H 3 HO The half-life of 3 H is 12 .43 years, which makes is a useful tracer on a decadal scale (Unterweger et al, 1980) Although it is a direct age determination, it is not a quantitative one This is primarily because the supply of 3 H in precipitation has not been constant through time The supply has not been constant for three main reasons : 1) there is a strong seasonal fluctuation everywhere due to stratosphere-troposphere interactions, 2) 7 4

PAGE 83

there are geographic difference driven by both atmospheric circulation patterns and oceanic exchange, and 3) there was a relatively short-lived human signal (1950's-1970's) up to two orders of magnitude greater than the cosmogenic background There is a strong seasonal fluctuation that results from atmospheric turnover from the stratosphere to troposphere This stratospheric injection happens in the spring in the northern hemisphere, and in late fall in the southern hemisphere (Clark and Fritz, 1997) The residence time of 3 H in the stratosphere is up to 10 years, while it is less than a month in the troposphere (Clark and Fritz, 1997 ; Ericksson, 1965) Since the troposphere residence time is so short, the timing of the stratosphere injection shows up quickly in precipitation (Clark and Fritz, 1997) The concentration of 3 H also varies with location The location of exchange between stratosphere and troposphere is focused in the mid and high latitudes (Craig and Lal, 1961) Atmospheric circulation patterns in the short residence-time moist layer therefore control where the higher concentrations of 3 H are deposited : temperate latitudes have high concentrations and low latitudes generally have very low concentrations (Ericksson, 1965) Another geographical control on 3 H in precipitation is driven by exchange between the air and ocean The ocean readily accepts 3 H ; air masses above the ocean have low concentrations of 3 H, while in the interior of continents concentrations of 3 H are higher In coastal areas steep gradients of 3 H (from low to high) have been measured from the coast inland (Eriksson, 1965) Anthropogenic generation of 3 H has varied widely through time Beginning with the first atmospheric explosion of a nuclear bomb in 1945, radionuclides from nuclear explosions have entered the atmosphere The advent of the atmospheric testing of fusion bombs in 1953 greatly increased the production of 3 H and injected material into the stratosphere, where is contributes to the atmospheric patterns described previously Bombpulse nuclides have affected the whole planet since the early to mid 1950's, although the timing and magnitude of the initial rainout varied with location During the 1950's and 1960's, 3 H concentrations in precipitation were up to three orders of magnitude higher than the cosmogenic background (IAEA) In Ottawa, Canada, a maximum concentration of 5817 TU was measured in 1963 (IAEA) (Figure 34) 7 5

PAGE 84

Due to various political agreements, most atmospheric testing of nuclear weapons was curtailed in 1963, although France and China continued small scale testing until 1980 (Clark and Fritz, 1997) Power plants and other human activities continue to produce 3 H, at much lower concentrations than from nuclear weapons As a result, atmospheric levels of 3 H declined through the 1980's and 1990's, although they are still elevated relative to cosmogenic levels Figure 34 shows two curves : the lighter line is the concentration of 3 H measured directly from precipitation in Ottawa, Canada This record has been used because it is the longest continuous record of the fluctuation of 3 H in one location Samples were taken at one-month intervals starting in August of 1953 and continuing to the present (IAEA/WMO) Data is currently available through the end of 1999 This curve has been smoothed with a 12-month moving average to suppress seasonal variation and better reflect the mixing that occurs in groundwater The large 3H spikes in precipitation is the 1950's and 1960's reflect the large amounts of tritium pumped into the atmosphere by human activities during this time Data prior to 1953 (8 TU) is approximated from Brown's 1961 recalculation of Ottawa river water (15 TU) and an attempt to accommodate the strong seasonal fluctuations found everywhere, as well as the admitted possibility that Brown's reagents has been contaminated by modern tritium (1961) Tritium Age Model The atmospheric precipitation curve can be adjusted for the decay of 3 H (Figure 34) The darker solid line in Figure 34 shows decay-adjusted concentrations If precipitation had immediately entered the ground and become isolated from the atmosphere, its 3 H would begin decaying and would not be replenished The dark solid line shows what the Ottawa precipitation would look like today as groundwater, with a twelve-month moving average to accommodate seasonal variation and to simulate mixing during recharge Figure 34 also shows the data divided into several sections The sections separate the decay-adjusted precipitation curve into distinguishable portions (Alexander and Alexander, 1989) 7 6

PAGE 85

Figure 34 Tritium in meteoric precipitation has been adjusted for radioactive decay The pale line shows the measured value of tritium in precipitation The dark line shows the amount of tritium groundwater of different ages would have today, if it had been recharged from precipitation at any given time This adjustment only accounts for radioactive decay, not fractionation by evapotranspiration during recharge Both series have been smoothed on a 12-month moving average The plot also shows the designation of tritium age groups discussed in the text 7 7

PAGE 86

7 8 Cosmogenic generation of tritium provides an atmospheric background of about 1 15 TU (Kaufman and Libby, 1954 ; Brown, 1961 ; Grosse et al, 1951 ; Craig and Lal, 1961) Thus, water entering the ground before the 1950's would have had a cosmogenic background of at most 15 TU, probably significantly less By 2000 AD, pre-bomb precipitation of 5 TU would have decayed to about 0 .3 TU, which is below our current detection limits Even 1950 water with an initial concentration of 10 TU would now be below detection limit Groundwater with no detectable 3 H ("tritium dead") is considered to have entered the ground before the early 1950's Surface waters in the study area were sampled for the tritium content of modern recharge, and are included in a histogram of tritium concentrations in ground waters for central Pine County (Figure 35) The histogram includes samples from Pine County that were not part of the thesis data set These samples were collected for other reasons, but do reflect the variety of tritium concentrations found in groundwater in Pine County This histogram has twenty-six samples with non-detectable levels of tritium These waters entered the ground prior to 1953 (>0 .8 TU) and are referred to in this text as Vintage Any water containing detectable tritium has some component of water that entered the ground since the early 1950's Surface waters in the study area had measured values between 10 and 14 TU Water samples with concentrations near those of current precipitation and surface waters are probably very recent and have very short residence times However, these waters could also be physical mixtures of Vintage waters and waters with much higher levels of 3 H Unmixed water from the 1950's and 1960's would have decay-adjusted tritium values around 30 TU or ever higher (Clark and Fritz, 1997) Such high values are rarely encountered in ground water because of mixing in groundwater is very effective on a decadal scale Thirty-five samples in this data set fall in the Recent age group, one of which may have been recharged during the period of peak fallout (1955-1967) Another group of physical mixtures have 3 H concentrations between Vintage and contemporary precipitation The lower bound of the mixed group is the presence of any detectable tritium, but the upper bound is hard to define, since there is a significant seasonal fluctuation The upper bound to this mixed group is probably about 8 TU Fifteen samples from this data set were in the Physical Mixtures group

PAGE 87

Figure 35 Concentrations of Tritium in the atmosphere varies through time, due to human activities and atmospheric circulation patterns Decay-adjusted precipitation values of tritium since the 1950's can be used to roughly define determine when water entered the ground Since the supply of tritium has not been constant, absolute ages cannot be calculated Three tritium classifications are used for Pine County : Vintage, Recent, and Physical Mixtures Vintage waters have no detectible tritum (< 0 .8 TU), and have been underground since before 1953 Recent waters have some detectable tritium, generally over 8 TU Physical mixtures between waters from the last 50 years with vintage waters yields detectable tritium concentrations less than about 8 TU 7 9

PAGE 88

II B 2 The Chloride/Bromide Age Model Another group of useful residence time indicators are elements and molecules that result from current human activities The absence of these molecules does not guarantee that waters are old, but may simply imply that the water was not exposed to those human activities However, their presence indicates a groundwater age corresponding to the cultivation and industrialization of an area Phosphates and nitrates are common in urban and agricultural runoff and rarely found in unimpacted water (Alexander and Alexander, 1989) In these samples, neither nitrate nor phosphate makes a significant systematic contribution The study area is neither urban nor heavily cultivated Similarly, high levels of chloride can either be the result of human activity {from road salt, sewage, etc .), or the evolution of old connate brines in bedrock These two sources of high chloride can be distinguished by the Cl/Br ratio of water, because bromide is a significant component of brines but effectively absent in the chloride used for human purposes In these samples, chloride and bromide are useful parameters If initial formation waters are derived from seawater, the initial Cl/Br ratio should be close to that of seawater The Cl/Br ratio of seawater is 288 when calculated from ppm concentrations orris and Riley, 1966) Bromide is concentrated in groundwater when chloride is preferentially sequestered in mineral phases Very old connate brines should have elevated amounts of bromide These evolved waters should have a Cl/Br ratio less than that of seawater Correspondingly, the mineral precipitates contain very little bromide The chloride for many human applications comes from mined sources of evaporite minerals An exceedingly high Cl/Br trend is interpreted as a sign of human activities Even at high concentrations of chloride (over 100ppm), anthropogenic chloride has so little bromide that the Cl/Br ratio is in the thousands Figures 36 and 37 are plots of chloride and bromide All Cl/Br ratios were calculated for weight concentrations Ratios for concentrations in molarity or equivalents have different absolute values, but the relationships are the same The figures show at least two trends : a Cl/Br ratio that is lower than the ratio of seawater, and another that is higher (Figure 36) As well, a few samples have a Cl/Br ratio close to that of sweater These five 8 0

PAGE 89

Figure 36 The ratio of chloride to bromide shows at least two trends The upper trend has a Cl/Br ratio greater than that of seawater In these waters, chloride has been added by human activites, such as the application of road salts or fertilizers, or the chlorination of water supplies or sewage The lower trend has a Cl/Br ratio lower than that of seawater In these waters, bromide has been concentrated due to the differential sequestering of chloride in mineral phases 8 1

PAGE 90

samples are ambiguously connected to each other by plumbing ; all five wells have some element of mixing, and should probably plot much closer to the connate brines trend This mixing is better explained by Figure 37 Figure 36 considers the Cl/Br ratio with respect to the tritium age groups The most evolved samples in this data set (which have a Cl/Br ratio of about 140, are tritium-dead and have very old 14 C ages These samples probably contain a component of very old connate brines Another group of samples has very high Cl/Br ratios The samples with high Cl/Br ratios have some component of modern tritium, indicating that they have been recharged within the past 60 years The highest calculatable Cl/Br ratio is over 11,000 This ratio cannot be calculated if bromide is below its detection limit Chloride was present in measurable amounts in all samples However, a significant number of samples had bromide concentrations below the detection limit For these samples, only a minimum Cl/Br ratio can be calculated The trends on Figure 36 defined by Cl/Br ratios can be used to distinguish water that has been significantly impacted by human activities, which suggests short residence times and quick travel times This tool is more useful for higher values of chloride (over 2030 ppm), because at low values the two trends converge and it is difficult to distinguish them In support of our chloride-bromide model, the samples that plot most decisively in our "human impacted" trend have modern tritium levels (recent water), while samples that plot most clearly in the "deep brine" trend are tritium-dead (older water) (Figure 36) The tritium and Cl/Br age models are derived from completely different sources Their agreement strengthens arguments for both models When the Cl/Br plot is labeled according to water source, three trends appear (Figure 37) The first trend is nearly vertical, and is composed of short residence time waters that have been affected by human sources of chloride These samples have modern tritium and Cl/Br ratios up to 11,000 All of these samples came from sandstone or Quaternary wells These wells and springs probably represent shallow flow systems in the Hinckley sandstone as well as in Quaternary glacial deposits The springs that are clearly part of this trend are A07 and A28 (abbreviated spring numbers, as used in Table 4) Other springs are producing short-residence time waters without high chloride These springs have not yet 8 2

PAGE 91

Figure 37 In this data set, wells in basalt define an old water/unimpacted trend, while wells in the Hinkcley and Fond du Lac Sandstones define a modern trend Springs plot in both trends becuase they issue from a variety of sources Some of the Hinkcley springs define the lowest Cl/Br trend, and appear to be from old, deep sources Other springs plot in the modern trend defined by wells in sandstone These springs are fed by Quaternary or shallow sandstone sources There are no springs flowing out of the basalts in this data set The springs mapped over basalt bedrock were flowing out of Quaternary deposits Most samples do not have enough chloride to plot definitively in either trend While the basalt water samples are tightly clustered, the sandstone and spring samples are not, even though the sandstone bedrock itself has a more uniform composition The wide range of Cl/Br ratios over a small geographic area in the sandstone illustrates the role of fracture and conduit flow, and the simultaneous prescence of different waters in the sandstone 83

PAGE 92

84 been effected by human activities, probably because they are in Banning State Park, but any of the short residence time springs could easily be impacted by human activities in the future The second trend is labeled "basalt brines" This trend line has a lower (more evolved) Cl/Br ratio than seawater The Cl/Br ratio is about 190 ; the Cl/Br ratio of seawater is 288 The wells that are clearly in the basalt brine trend are samples 20, 21, and 40 These wells are tritium-dead and have 14 C ages from 4,000 to 6,000 radiocarbon years b .p (Table 7 and 8) These wells appear to sample long residence time waters with slow travel times A group of five wells have Cl/Br ratios closer to 250, which is lower than seawater but significantly higher than the basalt brine trend These five wells are in a housing development and are connected to each other before they reach sampling points Four wells are finished in basalt, while the fifth is probably producing water out of both Quaternary sediments and basalt bedrock The result is that the bedrock groundwater dominates the chemistry of all five samples, but the small component of water from the Quaternary sediments dilutes the bromide The more concentrated wells (>20ppm Cl) in the basalt brine trend are all tritium dead (except the well mentioned above that is producing more than one water) and have old 14 C ages The third trend on Figure 37 is labeled "sandstone brines" These are all springs that issue directly from the Hinckley Sandstone along the Kettle River The Kettle River acts as a local control on base level These springs ostensibly sample an old, deep flow system in the Hinckley, with long residence times and slow travel times The Cl/Br ratio of these samples is about 140 They are mostly tritium-dead, have low Cl/Br ratios and have old 14 C ages when datable The springs that are clearly in this trend are A05, A08, A33, and A31 These springs have Cl/Br ratios of 139-153 Under the Cl/Br model, they represent the most evolved waters in this data set One of the springs (A05) produces vintage water with a Cl/Br ratio of about 140, while another (A31) contained modern levels of tritium (18 .1 TU), with a Cl/Br ratio of 149 Both of these waters appear to be strongly influenced by some very old brines, but the second was sampled in a surface stream where some mixing with' surface water occurred In this case Br is tracing the spring water and tritium is tracing the surface water Mixing with

PAGE 93

surface waters would not only contribute modern 3 H to the spring, it would also dilute the concentrations of Br Knowing that such surface contamination has occurred, the original Cl/Br ratio of this second spring is probably even lower than 140 The same argument can be made for spring A14 (Cl/Br = 238), which was sampled at the outflow of a pond about 5 m in diameter Springs A12, A15, A21 and A22 are very dilute (Cl = 1 .2 to 6 .2 ppm), but still may be part of one of the brine trends These springs have very low Cl/Br ratios (60-190), but these ratios may simply be the result of machine inaccuracy at the quantification limit The two different types of Hinckley springs (low Cl/Br and very high Cl/Br) can occur within meters of each other, which means that these waters exist in the sandstone simultaneously, without mixing The chemistries of these groups are distinct, and if samples can be taken without including surface water, they do not show subsurface mixing The implication of these different age waters simultaneously existing in the sandstone in close proximity is that the Hinckley sandstone is hydrologically complex Different flow paths exist with radically different residence times, different sources, and correspondingly different chemistries These flow paths must exist very close together, even be intertwined and yet do no mix until they reach the surface or are penetrated by a well bore II B 3 Carbon Dating Background Carbon has two stable isotopes ( 12 C and 13 C) and one radioactive isotope ( 14 C) The natural abundances of the stable isotopes are 98 .89% ( 12 C) and 1 .11% ( 13 C) The natural abundance of 14 C is about 10 -10 Along with tritium, radiocarbon is produced cosmogenically in the stratosphere as the result of reactions between nitrogen and cosmic ray induced neutrons : 14 N+n= 14 C+ 1 H In addition to the cosmogenic background, which has fluctuated through time, there have been significant anthropogenic contributions since the early 1950's Radiocarbon is removed from the atmosphere as 14 C0 2 and is incorporated throughout biospheric organic matter The 14 C fraction decays continuously with a half-life 8 5

PAGE 94

8 6 of 5730 years, which corresponds to a decay constant of A .=1 .210x10' 4 /yr The decaying 14 C is constantly replaced by living organisms After death, the 14 C decays and is not replenished Radiocarbon is incorporated into groundwater from atmospheric CO 2i where it reacts with carbonate rocks to form HCO3 If atmospheric CO 2 were the only carbon source and ground waters were closed systems, 14 C dating of groundwater would be straightforward The messy reality is that ground waters are by no means closed systems and have many carbon sources Other common carbon sources are organic material in the soil, and dissolving carbonate rocks or minerals All of 'these sources have different amounts of 14 C and 8 13 C values Carbon dating is a useful dating tool on materials up to about 35,000 years old, so most carbonate rocks and minerals contribute carbon without any 14 C, which dilutes the 14 C from the atmosphere or soil There are two main ways to dealing with the complexities of carbon additions and loss along groundwater paths : mass balance and inverse mass balance The mass balance approach uses both the measured 14 C content and stable isotope ratio to determine the contribution of biosphere carbon (Zhu, 2000) The inverse mass balance approach divides the groundwater flow path into multiple parts, and considers the carbon interactions along each part (Zhu, 2000) For the purposes of this reconnaissance hydrologic study, the mass balance approach is sufficient, and is calculated applying a Lever Law correction factor (Fmodel) to the decay equation : t = 1/X ln( 14C measured/F mode) Fmodel is the fraction of dissolved inorganic carbon (DIG) from biospheric carbon, instead of from old (carbon dead) rocks This age is model dependent, because the 8 13 C must be estimated for the biospheric DIC This value varies by location, and through time, and has a significant effect on calculated ages The two end members of the biospheric 8 13 C are -18%o and -25%o, depending on the local floral assemblage For this area, where the main rocks are silicates, the lever law calculation is : Fmodel (C measured -C rock )/ ( C biogenic C rock / A significant number of waters sampled in this project were reduced Reduced waters that have undergone methanogenesis require different equations to calculate a

PAGE 95

reasonable 14 C age In the production of methane (CH4), carbon is fractionated Specifically, the carbon given off as CH4 is isotopically light The 8 13 C values of this methane can be as light as -70%0 The carbon remaining in the water will therefore be correspondingly heavier, regardless of the fraction of modern carbon Fsoilmethane = ( 1 + 2b)/2, where b ( 813Cgroundwater 813 CbiogeniJ/ ( S13C biogenic 8 13 CCH4) These very reduced waters were probably recharged through wetlands, where S 13 C ranges in value from -25 to -31%o For carbon ages in waters that have undergone significant reduction, 8 13 C is assumed to be -30%o As previously mentioned, 8 13 C044 i s assumed to be -70%o in these age calculations Dated Samples Twelve samples in the data set were carbon dated A brief summary of the calculated ages is given in Table 7 A more detailed account of these calculations is given in Table 8 Ages are given in years before present (ybp, and are neither radiocarbon ages before present nor calendar ages), where the present is 1950 Carbon dates of water are at best approximate ; they represent complex mixing of different waters and reactions with many poorly constrained sources of carbon Groundwater carbon dates should be considered in a relative sense, such as very recent, moderately old, and very old waters Due to extreme model dependence, the calculated ages are reported to one significant figure Since this analysis is more expensive than most water chemistry tests, the samples were selected on basis of interest, in order to better constrain groundwater ages and residence times The vintage age provided by tritium can be relatively recent (as little as 50 or 60 years old) Since the Cl/Br ratios pointed to several flow systems that might represent connate brines, there was reason to suspect that some of the vintage waters would be much older than 50 or 60 years Four samples from basalt sources were 14 C dated The youngest age was about 200 years before present, while the other three samples produced ages between about 4,000 and 6,000 years before present Seven samples from the Hinckley Sandstone were 14 C dated, along with one sample from a Quaternary deposit near the City of Sandstone These samples had ages from about 40 to 6000 years before the present The 8 7

PAGE 96

final sample (from spring A6) could not be analyzed, because there was not sufficient carbon in the water sample In this data set, there appear to be two groups of dates one group ranges from fairly recent to about 200 500 years before present The other group clusters around 4000 years before present Further groundwater dating in the area may support these two trends If so, the two age groups may represent several different recharge events Table 7 Groundwater Ages from 14 C II C Water Chemistry II C 1 Piper Diagrams A general overview of the groundwater sampled in this project is provided by a set of piper diagrams (Figures 38, 39) Although samples from the Quaternary glacial sediments are located across a large area (see Figure 38), they are tightly grouped for the calciummagnesium-sodium and bicarbonate-sulfate-chloride systems (Figure 38A) This tight chemical grouping suggests a relatively uniform composition of major cations and anions in the glacial deposits Water samples from sandstone wells have similar values to the Quaternary samples (Figure 38B) The spread of this distribution is a little larger, due to increased amounts of chloride and slightly increased amounts of sodium for some samples In general the sandstone aquifers are surficially recharged through glacial sediment The sandstones are 88 Sample # Aquifer Age (rcybp) 3 Sandstone 40 7 Sandstone 500 16 Sandstone 2000 18 Quaternary 4000 19 Basalt 4000 20 Basalt 6000 21 Basalt 200 28 Sandstone 6000 40 Basalt 6000 54 Sandstone 100 55 Sandstone 4000 56 Sandstone Not Analyzable

PAGE 97

Figure 38 Piper Diagrams : Sandstone and Quaternary Wells Figure 38 Piper diagrams of water samples from Quaternary sediments and Hinckley/Fond du Lac sandstones These two lithologies yield waters that are tightly grouped for Ca-Mg-Na-K and HC03-S04-Cl-N03 For this group of parameters, water in the sandstone (B) is very similar to that in the unconsolidated glacial deposits (A) This indicates that water aquires these values early during recharge and no significant reactions are taking place in the sandstone to change them 8 9

PAGE 98

Figure 39 Piper Diagrams : Springs and Basalt Wells Figure 39 Piper diagrams of water samples from basalts and springs Basalt well waters (A) are characterized by high sodium, and some elevated chloride, although these two parameters are not necessarily related Springs discharge waters with a range of compositions (B), some of which have fairly high proportions of chloride 9 0

PAGE 99

relatively nonreactive, so that most of this groundwater has a similar composition to the Quaternary sediment ground waters Several of the samples fall outside of the basic Quaternary distribution These samples have a little more sodium and additional chloride The chloride could come from a variety of sources, such as brines or human activities Like the sandstones, the basalts are also recharged through the glacial materials Water samples in the basalts form a linear trend between quaternary composition and very elevated levels of sodium and somewhat elevated levels of chloride (Figure 39A) Sodium is available from reactions with minerals in the basalt such as feldspars, or the weathering products of these minerals in the fractured basalt flow tops where water wells are often completed The high chloride in some of these wells is probably from connate waters, because the wells are generally tritium-dead, have low Cl/Br ratios, and old 14 C ages Most springs mapped in this project discharge from the Hinckley/Fond du Lac sandstones or quaternary deposits overlying the sandstones Three springs were sampled that discharge from quaternary deposits over the basalt, as labeled on Figure 33 (Springs A20, A24, and A28) Most springs have water very similar to that found in the quaternary deposits or in the sandstone wells (Figure 39B) However, a significant group of springs has water with much more sodium and chloride These springs are producing water that has been impacted by human activities, has interacted with connate brines deep in the sandstones, or has interacted with basalt brines The composite Piper diagram illustrates that the sandstone, basalt and spring samples define compositional ranges that each have quaternary-type water as one end member (Figure 40) The composite Piper diagram is a compilation of the ranges defined in Figures 38 and 39 The sandstone distribution is very close to the Quaternary group, with a handful of samples having more chloride Basalt water samples range from near-Quaternary composition to having almost 100% sodium cations The moderate values of chloride yield an overall distribution dominated by Na The springs are more complex : on the cation ternary plot, springs appear to have quaternary compositions, basaltic compositions, or intermediate (mixed between quaternary and basalt) compositions On the anion ternary diagram, springs appear to have quaternary compositions, basalt compositions, and intermediate compositions Furthermore, there are several springs that have much more chloride than any of the sandstone or Quaternary well 9 1

PAGE 100

9 2 Figure 40 Composite Piper Diagram Figure 40 This piper diagram combines the data shown in the four individual piper diagrams in Figures 33 and 34 Sandstone, Basalt and spring samples define trends that have an end member very similar to the composition of water in quaternary sediments Some basalt water samples are influenced by brines and bedrock reactions Springs produce a mixture of recent recharge, sandstone groundwater, and water that has interacted with sandstone brines and basalt groundwaters

PAGE 101

water samples The source of this chloride is debatable : this could either be chloride from recent human activities or a more concentrated brine than was sampled in any of the bedrock wells Three of the samples have high Cl/Br ratios (from over 300 to 1,277) The other three samples have Cl/Br ratios close to 140 (139 -153) Other springs in Banning State Park are producing short-residence time waters without high chloride These springs have not yet been affected by human activities, probably because they are in a park, but any of the short residence time springs could easily be impacted by human activities in the future Springs are not exactly an average between Quaternary (and sandstone) waters and groundwater from the basalts This is evident in the square plot in the center of the composite Piper Diagram (Figure 40) The high chloride basalt ground waters have more sodium than the high chloride springs The lower proportion of sodium may point to sandstone brine sources It is reasonable that a brine in basalt would have relatively high concentrations Na, while a brine that originated in sandstone would have lower concentrations of Na However, sodium is nonconservative, so it is hard to make a definite division on the source of these springs based entirely on relative concentrations of sodium To approach this question, a model based on the relationship of two of the major cations (calcium and magnesium) and a more conservative minor cation (strontium) is more useful This model is discussed in section II .C .3 II C 2 Calcium and Magnesium The majority of water samples in this data set have compositions dominated by calcium, magnesium and inorganic carbonate species even though investigations have found no evidence of carbonate minerals in the bedrock (Tyrhorn and Ojakangas, 1972) or overlying drift (see Knaeble et al 2001 analysis of glacial sediments) The geochemistry is therefore discussed in terms of carbonate equilibria, even though there are not initial carbonate minerals in the system Figure 41 shows a linear relationship between calcium and magnesium While there are several outliers to this linear relationship, the overall trend is very strong (r 2 = 0 .93) The linear relationship between calcium and magnesium implies that the main source of Ca and Mg is fairly uniform across the study area There are no distinctive patterns within the trend 9 3

PAGE 102

Figure 41 The concentrations of calcium and magnesium in water are clearly related (A) In linear regression, Rsq = 0 .93 The regression is shown with a 95% confidence interval The Ca/Mg ratio from this regression is 1 .83 (molar concentrations) or 64% Ca and 36% Mg Neither the amount nor ratio of Ca/Mg is a function of residence time or aquifer The process controlling this ratio happens very early during recharge and remains relatively stable after that time This is clearly not a bedrock relationship (B) Analysis of basalt whole rock samples shows no systematic relationship between calcium and magnesium 94

PAGE 103

with respect to water source or age These observations suggest that neither the amount nor ratio of Ca/Mg is a function of residence time or aquifer In this case, the Ca/Mg ratio (about 1 .8 for molar concentrations, 2 .4 for weight units) is reached very early during recharge and remains relatively stable after that time Concentrations of Ca and Mg are probably controlled by P CO2 which controls the concentration of HCO 3 ; in the soil zone and unconfined waters P CO2 is increased by organic contributions, but in confined aquifers CO 2 is consumed by reactions without additional input sources The important role of reactions with HC0 3 during recharge is also highlighted by the relationship between calcium, magnesium (the major cations) and alkalinity (the major anion) The dissolution of minerals in the soil zone by H 2 CO 3 yields HC03 and cations at a ratio of 2 :1 Carbon dioxide in water forms carbonic acid : C0 2(aq) + H 2 O H H 2 CO 3 Hydrolosis of Anorthite to Kaolinite : CaA1 2 Si 2 0 $ + H20+ H 2 CO 3 H A1 2 Si 2 O 5 (OH) 4 +Ca t+ + 2HC03 Solution of Olivine (Forsterite) : Mg 2 SiO 4 + 4H 2 CO 3 H H 4 SiO 4 + 2Mg 2+ + 4HC0 3 Most of the water samples agree with the calculated 2 :1 molarity ratio (Figure 42) Excluding the known high sodium and high chloride samples, linear regression of alkalinity vs Ca + Mg with alkalinity as the dependent variable gives a slope of 1 .9, with an R 2 value of 0 .91 The pH of unconfined quaternary aquifers is controlled by carbonate buffering In the sandstones, the carbonate buffering continues to be important when the water has been isolated from additional carbon sources Figure 43 shows that as pH increases in sandstone samples, alkalinity increases also These waters are buffered by carbonate equilibria Carbonate buffering is less effective in basalt waters The reactive minerals in basalt quickly consume HCO 3 once water is cut off from atmospheric and soil-zone carbon sources The Olivine solution reaction shows how effective these minerals can be : one mole of Olivine can consume four moles of alkalinity At high pH, carbon is present primarily as C0 3 2, instead of HCO3 Mg 2 SiO 4 + 4HC03 H H 4 SiO 4 + 2Mg 2+ + 4C0 3 9 5

PAGE 104

Figure 42 Calcium and Magnesium concentrations are primarily controlled by carbonate buffering, as demonstrated by the good fit to the theoretical alkalinity/(Ca+Mg) ratio of 2 Linear regression on the samples gives a ratio of 1 .9, with an Rsq of 0 .91 The regression line is shown with a 95% cofidence interval Samples that plot above the 2 :1 line contain very high concentrations of sodium (>96% of total cations), while samples tht plot below the 2 :1 line contain very high concentrations centrations of chloride (>66% of total anions) Figure 43 Groundwaters in the basalts can have much higher pH than in sandstone and Quaternary aquifers, becuase these waters are not subject to carbonate buffering When waters in the basalts are isolated from the atmosphere, solution of basic minerals consumes hydrogen until all HCO3is sequestered as CO32. The cations liberated by solution of the basalt minerals are precipitated as carbonate minerals 96

PAGE 105

Cations liberated by the dissolution of primary basalt minerals will precipitate with the carbonate, so that the carbon in this system is stored in solid form Consumption of HC03 does not alter the Ca/Mg ratio, as there is no correlation between unbuffered waters and Ca/Mg outliers II C 4 The Strontium Model Strontium in Rocks and Sediment Different water sources in this study can be distinguished by Sr/Ca and Sr/Mg ratios These distinctive Sr/Ca and Sr/Mg ratios reflect weathering reactions before and during groundwater-rock (or groundwater-till) interactions Figure 44 shows a comparison of calcium and strontium in water samples and different source materials Whole rock analyses of twenty-six basalt samples from Pine County were performed by the Minnesota Geological Survey (MGS) (Boerboom, written communication, 2002) These are the same basalts from which the basalt water samples were collected These rocks have high Sr/Ca ratios, with a maximum of 1 .54x10 -2 (Table 9) Fifty basalt samples from the Midcontinent Rift System further north of the study area were analyzed by Schmidt (1990) These basalts have very similar Sr/Ca ratios, with a maximum of 1 .82x10 -2 The top line in Figure 44 plots the latter ratio, to include all of the basalt rock samples Eight whole rock analyses of several southeast Minnesota limestones and dolomites (Cummingsville, Oneota, Prosser, and Shakopee) yielded a tightly clustered set of Sr/Ca ratios, with a minimum ratio of 2 .9x10 4 (Alexander, 2002) The smallest ratio is plotted as the bottom line on Figure 44 These two lines bracket the water sample ratios Several samples of till from the Superior Lobe and Grantsburg Sublobe were analyzed in an experiment on cation exchange capacity (Alexander, written communication, 2002) Samples from a variety of depths below ground surface were treated with both deionized water and ammonium chlorite acetate The resulting solutions were analyzed While this experiment does not provide a bulk composition to compare with whole rock samples, it may give a better measure of the composition of water that recharges though till The water-processed samples are used in the graphical and statistical comparisons Till samples processed with DI water had an average Sr/Ca ratio of 3 .65x10 -3 This till ratio plots very near the lower values of the basalt whole-rock analyses The glacial sediments in 9 7

PAGE 106

Figure 44 The Sr/Ca ratios of rocks and sediments supports the strontium model Basalt rock samples from Pine County have higher Sr/Ca ratios than iether SE Minnesota carbonate rocks or till from the two lobes that occupied Pine County The magnitude of the till samples is lower becuase this measurement is based on cation exchange capacity with water instead of bulk analysis The carbonate and basalt values are from whole rock analyses 98

PAGE 107

Pine County are primarily a mixture of sandstone and basalt, so it is not surprising to see this basalt influence in the till A statistical comparison of these materials is shown in Figure 45, and a list of summary ratios is given below in Table 9 The box plots in Figure 40 visually compare the Sr/Ca ratios by summary statistics Only the three groups used to define lines in Figure 44 are shown The boxes span the first and third quartiles, and are divided by the mean The vertical whiskers one standard deviation, and the outlying dots display 2and 3-sigma outliers The box plots demonstrate what Figure 44 strongly suggests : that Sr/Ca ratios are highest in basalt bedrock, lower in the overlying tills, and very low in sedimentary rocks Table 9 Rock and Sediment Strontium Ratios Strontium in Groundwater In this data set, basalt well waters have higher Sr/Ca and Sr/Mg ratios than the water in Quaternary wells, most sandstone wells, and most springs (Figure 46) The Quaternary wells have intermediate Sr ratios They are producing water that has mainly been stored in glacial materials Like the sediment Sr/Ca ratios discussed above, this glacial material contributes more strontium than carbonate rocks (or rainwater), but less than basalt bedrock, as the glacial material is more weathered Sandstone wells have low to moderate Sr ratios, reflecting the contribution of surface water and groundwater in glacial deposits to recharge in the sandstone aquifers, and an absence of strontium in the sandstone Most springs, both sandstone and Quaternary, have relatively low Sr ratios These two trends (high Sr/Ca and moderate-low Sr/Ca) are well defined for Sr/Ca and Sr/Mg However, there are several springs and sandstone wells that have significantly higher ratios of Sr/Mg and Sr/Ca Because Mg and Ca are linearly related, the same samples stand out on both plots The springs and sandstone wells that have the highest Sr ratios are 9 9 Sam le T s Sr/Ca Sr/M : Basalt-MGS 5 .35E-0 2 .59E-03 1 .54E-0 5 .90E-0 1 .23E-0 1 .49E-0 Basalt-Schmidt 4 .95E-0 6 .66E-0 1 .82E-0 7 .58E-0 1 .26E-0 1 .90E-0 Limestone & Dolomite 4 .50E-0 2 .90E-0 6 .10E-0 2 .21E-0 3 .83E-0 5 .25E-0 Till-water 3 .65E-0 1 .63E-03 1 .01E-0 5 .25E-0 6 .97E-0 1 .16E-0 Tillammonia 1 .41E-0 2 .12E-0 3 .80E-0 9 .98E-03 1 .17E-0 1 .35E-0

PAGE 108

Figure 45 Different material have significantly different Ca/Sr ratios Basalt bedrock in Pine County has the highest median Ca/Sr, and is skewed toward even higher values Carbonate rocks from southeast Minnesota have very low Ca/Sr ratios Superior Lobe and Grantsburg Sublobe Tills treated with deionized water have intermediate Sr/Ca ratios, as the detrital basalt minerals in the tills have undergone more weathering than basalt bedrock The table provides summary statistics for the box plots 1 0 0 Sample 'Basalt (MGS) Till (water treated) Limestone&Dolomite Average 26 8 8 Mean 5 .35E-03 3 .65E-03 4 .51E-04 Standard Deviation 3 .63E-03 3 .01E-03 1 .12E-04 Minimum 2 .59E-03 1 .64E-03 2 .88E-04 1st Quartile 3 .18E-03 1 .70E-03 3 .36E-04 Median 3 .75E-03 2 .29E-03 4 .87E-04 3rd Quartile 5 .89E-03 5 .52E-03 5 .27E-04 Maximum 0 .0154 0 .0101 6 .05E-04

PAGE 109

Figure 46 Calcium and magnesium have similar relationships with strontium Basalt wells have higher Sr/Ca and Sr/Mg ratios than most Quaternary and sandstone wells Most springs produce water with low Sr ratios, but a few springs and wells (all sanstone and Quaternary) produce water that appears to have mixed with basalt water at some point These springs are 65, 70, 86, and 87 101

PAGE 110

1 0 2 probably producing water with a component of brine that originated in the basalts and has moved into the sandstones Sandstone brines should not stand out in this plot In support of this model, all of the springs that were previously described as brines (A5, A8, A33, and A31) in the Cl/Br model also have the high Sr ratios These springs have the same range of Sr ratios as the basalt well waters The sandstone wells that appear to have a component of basalt brine are #3, 6, 13, 16, 36, and 39 In Support of the Strontium Model : Major Cations Calcium, magnesium, and sodium are the major cations in the water samples, accounting for 98-100% of total cations While these cations are not completely conservative, they do show trends that agree with the strontium model (Figure 46) The plot of Ca+Mg vs Na shows several relationships in these waters (Figure 47) Waters in the basalts are characterized by high sodium, not correlated to Ca+Mg concentration Waters in Quaternary and sandstone aquifers are characterized by low values of Na, positively correlated to moderate concentrations of calcium and magnesium A group of samples continues this trend with higher concentrations of Na Most of the samples have high Cl/Br ratios, and the higher proportions of sodium accompany the anthropogenic chloride, However, some of the samples may be concentrated sandstone/Quaternary waters (i .e ., sandstone brines) A final group of samples (mostly springs) has a higher Na/(Ca+Mg) ratio than the sandstone/Quaternary trend This group includes springs issuing from sandstone, and several shallow wells producing a mixture of basalt and Quaternary groundwater These are the same springs that appeared to have interacted with basalt groundwater in the Sr model (A8, A31, and A32) Spring A5 lies completely in the basalt trend, as it did in the Sr model In Support of the Strontium Model : Boron Like strontium, boron is another relatively conservative element Boron is seen in high concentrations in basalt water wells in other parts of Minnesota (Allen, written communication, 2002) Boron is available to groundwater through reactions with basalt ; it is not a significant component of surface recharge, till or sandstone Boron is more soluble at high pH As seen in Figure 48, basalt groundwater samples have both high pH (7 .92 to

PAGE 111

Figure 47 Calcium, magnesium, sodium and are the major cations in the water samples, accounting for 97-100% of total cations Ca+Mg vs Na shows several relationships in these waters Waters in the basalts are characterized by high sodium, not correlated to Ca+Mg concentration Waters in Quaternary and sandstone aquifers are characterized by low values of Na, positively correlated to moderate concentrations of calcium and magnesium A small group of samples has higher concentrations of Na Most of the samples have high Cl/Br ratios, and the higher proportions of sodium accompany the anthropogenic chloride A final group of samples (mostly springs) has a higher Na/(Ca+Mg) ratio than the sandstone/Quaternary trend This group includes springs issuing unambiguously from sandstone, and several shallow wells producing a mixture of basalt and Quaternary groundwater These are some of the same springs that appeared to have interacted with basalt groundwater in the Sr model (A8, A31, A32) SpringA5 lies completely in the basalt trend, as it did in the Sr model 103

PAGE 112

10 .07) and high boron concentrations (50 to about 27,000 ppb) Ten of the twelve water samples from the basalts have boron concentrations higher than the drinking water standard of 600 ppb (Minnesota Department of Health, 2002) The highest concentration of boron is 45 times higher than the drinking water standard The sandstone and Quaternary wells have low concentrations of boron The two nominally basalt wells that lie in the "mixing" region have the lowest boron concentration, moderate pH (about 8), and appear to have a significant component of surface or Quaternary-derived water in the Ca-Mg-Na-K plot (Figure 47) Not all samples were analyzed for boron, so it is not possible to compare the boron and strontium "mixing" wells and springs for all samples For the samples that were analyzed for boron, all of the springs identified from the Cl/Br and Sr/Ca models as containing the boron signature also appear to have a basalt component on the boron plot Most of the sandstone wells with a basalt component in the Sr/Ca model ( 6, 13, 17, 28, 36, 39) also lie in the basalt mixing region in Figure 48 The Cl/Br ratios in many of the sandstone wells were indeterminable due to low bromide levels II D Iron Springs Of the 33 mapped springs, 28 are along the Kettle River In the field, these springs can be classified in two groups : clear and orange The clear springs were often oxygenated Most of the sampled clear springs were ambiguous for Cl/Br because they were very dilute Several clear springs with measurable bromide had anthropogenic Cl/Br ratios These springs had all had Sr/Ca ratios in the sandstone/Quaternary range, and low concentrations of boron The sampled clear springs appear to produce waters with short residence times The orange springs are very reduced : they contain Feel in solution This iron is used by Theobacillus ferrooxidans, an iron oxidizing bacteria These bacteria form colonies that are bright orange and have the consistency of stringy pudding The colonies can be several feet in diameter, several feet thick, and commonly line the streambeds of steams that flow from the iron springs Springs and seeps along the banks of the Kettle River can be identified from a distance by the bright orange streaks Also, in shallow places, iron springs can be identified underwater in the Kettle River bed by their T ferroxidans colonies Like the clear springs, the orange springs have a range of values for most parameters that have been discussed in Chapter 2 The orange springs include the lowest Cl/Br ratios of any 1 0 4

PAGE 113

Figure 48 Boron is available to groundwater through reactions with basalt Boron is not a significant component of surface recharge or water that is only stored in till and sandstone Boron is more soluble as high pH, and it does not precipitate easily at low temperatures As seen in Figure 43, basalt groundwater often has high pH For this reason, most of the basalt wells have high boron concentrations, almost 2 orders of magnitude greater than the drinking water standard (600ppb) The two basalt wells that lie in the "mixing" region have moderate pH (about 8), and appear to have a significant component of surface or Quaternary-derived water in the CaMg-Na plot (Figure 47) 1 0 5

PAGE 114

samples in the data set, and also samples too dilute to measure bromide They have 14 C ages that range from very old (3,000 y .b .p) to recent, and both vintage and physical mixtures of 3 H Of the sampled orange springs, Sr/Ca ratios plotted both in the basalt mixing region and squarely in the center of the sandstone/Quaternary region, and also contained a range of low and moderate concentrations of boron (17 253 .ppb) The orange springs appear to produce a variety of waters, with a wide range of residence times and flow paths These two field classifications of springs along the Kettle River (clear and orange) do not correlate with age of water, or even their aquifer sources (as with the Sr/Ca ratios) The orange springs are reduced The redox states of these springs apparently depend heavily on factors prior to aquifer storage, such as the early recharge environment The study area contains many wetlands, and this redox phenomenon might reflect recharge through wetlands instead of through fields or grassy areas A detailed study of stable carbon isotopes might shed light on the redox question II E Samples Excluded from the Discussion and Analysis Three wells are shown in a Piper diagram in Figure 49 that fall far outside the sandstone distribution Their chemistries are so different that they are not included the discussion of compositional ranges of different aquifers or the strontium model These three wells are included in the chloride/bromide model and discussion of residence times The chemistry of all three wells has been overwhelmed by the results of human activities The evidence that the abnormal chemistry is human-derived is modern tritium concentrations, coliform bacteria, and extremely high Cl/Br ratios 1 0 6

PAGE 115

Figure 49 Samples Excluded from Water Chemistry Discussion Figure 49 Three wells that were significantly impacted by human activities have been excluded from the sandstone groups The chemistry of these samples is so different that they are not included the discussion of compositional ranges of different aquifers or the strontium model These three wells are included in the chloride/bromide model and discussion of residence times The chemistry of all three wells has been overwhelmed by the results of human activities 1 0 7

PAGE 116

REFERENCES Alexander, S .C ., and Alexander, E .C ., Jr ., 2002 Field and Laboratory Methods, Hydrogeochemistry Laboratory, Department of Geology and Geophysics, University of Minnesota, 20pp This document can be obtained by writing to : Scott Alexander, Hydrogeology Laboratory, 310 Pillsbury Drive SE, Minneapolis, MN, 55455 Alexander, S .C ., and Alexander, E .C ., Jr ., 1989 t Residence Times of Minnesota Groundwater, Journal of the Minnesota Academy of Science, vol 55, number 1, p 48-52 Brown, R .M ., 1961 Hydrology of tritium in the Ottawa Valley Geochemica et Cosmochimica Acta, 21 : 199-216 von Buttlar, H, and Libby, W .F ., 1955 Natural distribution of cosmic ray produced tritium Journal of Inorganic Nuclear Chemistry, 1 :75-91 Beta Analytical, 2002 Beta Analytical, Inc : Laboratory Methods and Quality Control 15 May 2002 available at w ww .radiocarbon .co m Clark, I ., and Fritz, P ., 1997 Environmental Isotopes in Hydrogeology Lewis Publishers, New York, 328pp Craig, H ., and D Lal, 1961 The production rate of natural tritium Tellus, 13 : 85-105 Eriksson, E ., 1965 An' account of the major pulses of tritium and their effects in the atmosphere Tellus, 17 : 118-125 Faure, G ., 1986 Principles of Isotope Geology, second edition John Wiley and Sons, New York, 589 p Gat, J R ., 1980 The Isotopes of hydrogen and oxygen in precipitation In P Fritz and J Ch Fontes, eds ., Handbook of Environmental Isotope Geochemistry, vol 1A, p 21-47 Elsevier, Amsterdam, 545p Grosse, A ., Johnston, W .M ., Wolfgang, R .L ., and Libby, WF ., 1951 Tritium in nature Science, 113 :1-2 Heemskerk, A .R ., 1998 Environmental Isotope Laboratory Methodology : Tritium Online Internet 10 May 2002 available at w ww .sciborg .uwaterloo .ca/research-groups / eilab/Methodology International Atomic Energy Agency, 2002 Database of the Global Network for Isotopes in Precipitation (GNIP) and the Isotope Hydrology Information System (ISOHIS) Available 20 May 2002 at w ww .isohis .iaea .or g Kaufman, S ., and Libby, W .F ., 1954 The natural distribution of tritium Physical Review, 93 : 1337-1344 1 0 8

PAGE 117

Knaeble, A .R ., Patterson, C .J ., and Mayer, G .N ., 2001 Plate 5-Quaternary Stratigraphy, County Atlas Series, Atlas C-13, Part A Minnesota Geological Survey, St Paul, Minnesota Libby, W .F ., 1946 Atmospheric helium-3 and radiocarbon from cosmic radiation Physical Review(s), 69 : 671-672 Morris, A .W ., and Riley, J .P ., 1966 The bromide/chlorinity and sulphur/chlorinity ratios in sea water Deep Sea Research Oceanography Abstracts 13, p .699 National Atmospheric Deposition Program (NADP), 2001 Cooperative Research Support Program of the State Agricultural Experiment Stations (NRSP-3), Federal and State Agencies and Non-Governmental Research Organizations Available 10 February 2002 at h ttp ://nadp .sws .uiuc .edu / Nir, A ., Kruger, S j ., Lingerfelder, R .E ., and Flamm, E .J ., 1966 Natural tritium Reviews of Geophysics 4 : 441-456 Smart, P .L ., and Hobbs, 1986 Characterisation of carbonate aquifers : a conceptual base In : Proceedings of the Environmental Problems in Karst Terranes and their Solutions Conference, Bowling Green, KY, 1-14 National Well Water Association, Dublin, Ohio Tamers, M .A ., 1975 Chemical Yield Optomization of the Benzene Synthesis for Radiocarbon Dating International Journal of Applied Radiation and Isotopes, vol .26, p .676682 Taylor, C .B ., 1977 Tritium enrichment of environmental waters by electrolysis : Development of cathodes exhibiting high isotopic separation and precise measurement of tritium enrichment factors Proceedings of the International Conference of LowRadioactivity Measurements and Applications, Slovenski Pedagogicke Nakladatelstvo, Bratislava : 131-140 Tryhorn, A .D ., and Ojakangas, R .W ., 1972 Sedimentation of the Hinckley Sandstone of east-central Minnesota In Sims, P .K ., and Morey, G .B ., eds ., Geology of Minnesota : A centennial volume Minnesota Geological Survey, St Paul, Minnesota, 416-424 Unterweger, M .P ., Coursey, B .M ., Schima, F .J ., and Mann, W .B ., 1980 Preparation and calibration of the 1978 National Bureau of Standards tritiated water standards International Journal of Applied Radiation and Isotopes, 31 : 611-614 Zhu, C and Murphy, W .M ., 2000 On Radiocarbon Dating of Ground Water Ground Water, 38 : 2609-2620 Minnesota Department of Health, 2002 Drinking Water Protection Available 3 June, 2002 at h ttp ://www .health .state .mn .us/divs/eh/water/index .html 1 0 9

PAGE 118

III Heterogeneous Permeability & Quartz Solution in the Hinckley Sandstone III A Introduction There are karst features developed in and over the Hinckley Sandstone Surface karst features, such as sinkholes, are an indication that the subsurface is heterogeneous : if the subsurface was a homogenous porous medium, sinkholes could not form As discussed in Chapter 1, bedrock type is a major control in sinkhole development This observation prompts a consideration of the permeability structure of the Hinckley sandstone The mathematical definition of permeability (specific or intrinsic permeability) is that of proportionality constant which describes the ability of fluid to move through a medium based only on the characteristics of that medium Permeability is defined mathematically as k = Cd 2 where d is the diameter of ideal spherical grains, and C is another proportionality constant designed to account for the variability of real-world materials : the distribution of grain sizes in rocks or soil, sphericity and roundness of grains, and the type of grain packing Hydraulic conductivity is a proportionality constant used in Darcy's Law ; hydraulic conductivity combines matrix permeability with fluid parameters Hydraulic conductivity is defined mathematically as K = (kpg)/, where p is density and  is dynamic fluid viscosity Permeability also has a physical meaning, which is simply how well a fluid moves through some matrix In other words, permeability is how well the void spaces in a material are connected This physical meaning is exactly what the equations above attempt to describe The epoxy experiments described below display how a fluid (warm epoxy) moves through a material (Hinckley Sandstone samples) The description of the epoxy experiments tries to convey the, motion of fluid in more than one dimension, based on the observations that fluid motion is anisotropic in these samples, and that fluid motion appear to be heterogeneous within the aquifer (see Chapter 2) 1 1 0

PAGE 119

The Hinckley Sandstone has a marked variability in strength : in places it has been quarried for building stone, while outcrops near the old quarries can be friable sandstone that crumbles at the touch The Hinckley is similar to some of the rocks described by Young (1986, 1987, 1988) in several aspects These similarities are : that an otherwise wellindurated rock can be crumbled by hand ; that such a mechanically weak rock can maintain vertical or even over-steepened faces in outcrop ; that the rocks show karstic development that follows sedimentary or structural (deformation) features ; and that surface weathering of the rock can be inhibited by a resistant weathering crust on the rock The first two observations suggests that the rock was once uniformly wellindurated, and that the selective dissolution of cements allows the rock to still be interlocking and strong in compression, but weak in shear (Young, 1988) The third observation implies that small-scale sedimentary and structural features have significant control on the permeability structure of the rock Young (1986, 1987, 1988) argues persuasively that the karst features found in Australia are due to quartz solution III B Outcrop Observations Fieldwork between the fall of 1999 and the fall of 2001 focused on the many Hinckley Sandstone outcrops along the Kettle River Most of the small Hinckley Sandstone caves are located in this area of outcrops Some outcrops, particularly in the historical quarries, are hard, well-indurated rock Others, some within a couple of hundred meters of the quarries, consisted of natural cliffs where the rock is soft and friable Within these outcrops, there are noticeable variations in rock strength defined by bedding planes Many of these outcrops have morphological features that are strikingly similar to those seen in limestone karst : enlarged bedrock fractures, conduits and curvilinear solution forms Since the examined outcrops are located along the Kettle River valley, is it hard to separate the effects of mechanical river erosion or valley wall extension from those due to solution Most of the enlarged fractures and conduits narrow quickly as they penetrate the rock mass, because surface erosion has exploited and enlarged the already existing features Once they become too narrow for the human body it is impossible to tell how far they 1 1 1

PAGE 120

continue, and at what size For enlarged fractures and conduits to dominate the permeability of a rock, these features do not need to be human size A bedrock conduit that is merely a centimeter in diameter can transport significant volumes of water quickly, in turbulent flow Figure 50 shows an outcrop that had several types of high-permeability bedrock features In this figure, the outcrop is divided into three main units, which will also be used in section Permeability Observations section This division is based field observations, only applies to this outcrop, and is not intended to serve as stratigraphic nomenclature The uppermost of the three units (AH : above the rock with holes) is about 2 meters thick, with meter scale cross beds It is well indurated, the hardest sandstone, and the most resistant to weathering The middle unit is subdivided in upper and lower parts (UH, LH) This rock is about four meters thick, and is thinly bedded to laminated with very long, low amplitude cross bedding It is the least resistant to weathering, more friable, and is recessed compared to the other rocks UH and LH both contain bedrock conduits and fractures that are more enlarged than in the rocks above and below (AH, BH) The conduits are located at the intersection of two fractures, or the intersection of fractures and selected bedding planes In Figure 50, the fractures are shown by heavy dashed lines, while the conduits are shown as shaded areas The lowest unit is BH (below the rock with holes) It is at least two meters thick (the base is not exposed), friable, and did not have any sedimentary structures visible at outcrop or hand sample scale III C Hand Sample Permeability Observations Several samples were collected from the outcrop in Figure 50 Several of the samples were so friable that transportation from field to lab was challenging In the lab, the samples were placed on individual trays and oven dried for several days, at 60€C Portions of these samples were then placed in a vacuum chamber and impregnated with thin section epoxy The samples were oven dried again to insure complete curing When the epoxy was cured, the samples were cut in with a rock saw and the depth of epoxy 1 1 2

PAGE 121

Figure 50 Hinckley Sandstone Outcrop along the Kettle River, divided into three units The uppermost of the three units (AH) is about 2 meters thick, with meter scale cross beds It is well indurated and the most reisistant to weathering The middle unit is subdivided in upper and lower parts (UH, LH) This rock is about four meters thick, and is thinly bedded to laminated with very long, low amplitude cross bedding It is the least reisistant to weathering, and is recessed compared to the other rocks UH and LH both contain bedrock conduits and fractures that are more enlarged than in the rocks above and below (AH, BH) The conduits are located at the intersection of two fractures, or the intersection of fractures and selected bedding planes The lowest unit is BH It is at least two meters thick (the base is not exposed), and did not have any sedimentary structures visible at outcrop or hand sample scale Fractures are shown by heavy dashed lines, while the conduits are shown as shaded areas 113

PAGE 122

penetration was recorded Samples that were not completely saturated by the epoxy were dried (because of water from the rock saw), and subjected to another cycle of vacuum impregnation, cut, and observed This cycle was repeated until all samples were completely saturated with epoxy Epoxy is more viscous than water, moves slower, and solidifies in place, which allows observation of specific flow paths The samples exhibited several patterns One pattern is that surfaces exposed to subaerial weathering can block epoxy penetration, suggesting an impermeable residual crust similar to that described by Young (1986) Small sedimentary features (such as laminations) can have a large effect on fluid flow, either facilitating or impeding the movement of fluid Small fractures can have a similar effect A final observation is that fluid flow in some samples was markedly anisotropic In some cases, the anisotropy could be related to visible sedimentary and structural features, while in other cases the anisotropy appeared to be responding to features not visible to the naked eye or with a hand lens Such "invisible" features can exert major control over the permeability structure of the rock Sample AH-1 was a roughly round rock 3 to 4 centimeters in diameter, from above the enlarged fractureand conduit-bearing strata Epoxy was able to soak into the sample in a relatively uniform fashion, as shown in Figure 51 No sedimentary or structural features were seen in this sample, although the parent rock has well defined meter-scale cross beds Sample BH-1 was taken from below the conduit-bearing strata The sample had visible laminations The dashed line in Figure 51 shows a weathering surface that obscured the laminations This sample was almost completely saturated by epoxy in the first application, which indicates that it is very permeable Permeability was greatest parallel to the laminations, although several laminations were much less permeable Sample UH-1 (Figure 52) was taken from the upper part of the conduit-bearing strata The outcrop at this level had cross bedding and other sedimentary structures, and had very heterogeneous weathering patterns This samples had a top surface with centimeter-scale anastomosing channels developed along a bedding plane Epoxy penetration was preferential when parallel to sedimentary structures, but appears to be reduced adjacent to a weathering surface 1 1 4

PAGE 123

Sample AH-1 was above the zone of enlarged fractures and conduits This layer of rock was massive and did not have the cross beds often seen in the Hinkcley Gray shows the areas saturated by epoxy This sample has relatively uniform permeability This sample did not have any sedimentary and deformation structures visible at hand lens scale, although the parent rock has obvious meter-scale cross beds Sample BH-1 was below the zone of enlarged fractures and conduits This sample had visible laminations The dashed line shows a weathered surface, where the sedimentary structures were obscured It was almost completely saturated by epoxy on the first try Permeability was greatest parallel to laminations, although a couple of ,laminations are markedly less permeable 115

PAGE 124

Sample UH-1 was in the zone of enlarged fractures and conduits This part of the outcrop has two distinct layers, referred to here simply as "upper" and "lower" UH-1 was from the upper layer This layer of rock had cross beds, other sedimentary stuctures, and showed very heterogeneous weathering patterns There were many enlarged fractures and even conduits in the rock This sample had a top surface with cm-scale anastomosing channels developed along a bedding plane Epoxy penetration was more effective parallel to sedimentary structures, but appears partially blocked when adjacent to a weathering crust Sample LH-2 was in the zone of enlarged fractures and conduits This part of the outcrop has two distinct layers, referred to here simply as "upper" and "lower" LH-2 was from the lower layer This layer had cross beds, other sedimentary structures, and showed very heterogeneous weathering patterns There were many enlarged fractures and even conduits in the rock There is a small vertical fracture in the center of this sample that was not visible before the application of epoxy This fracture was very effective at transporting fluid The sample was almost impermeable perpendicular to the laminations and only slightly permeable parallel to the laminations 116

PAGE 125

Sample LH-2 was in the zone of enlarged fractures and conduits, and is also shown in Figure 52 This part of the outcrop has two distinct layers, referred to here simply as "upper" and "lower" LH-2 was from the lower layer This layer of rock had cross beds, other sedimentary structures, and showed very heterogeneous weathering patterns There were many enlarged fractures and even conduits in the rock Sample LH-2 had visible laminations This rock was effectively impermeable perpendicular to the laminations, and slightly more permeable parallel to laminations There is a small vertical fracture in the center of this sample that was not visible before the application of epoxy, and served as the most permeable flow path, in contrast to the fracture seen in LH-1 Sample LH-1 (Figure 53) was from the same strata as LH-2 The boundary between the epoxy saturated and unaffected areas was sharp This sample had epoxy applied three times before it was completely saturated Small fractures and laminations exert a strong control on fluid flow One fracture runs diagonally from the top right of the sample (as shown) to the bottom left This fracture inhibited fluid movement, but was only visible after epoxy had been applied Surfaces exposed to surficial weathering did not effect epoxy movement in this sample The finer laminations near the base of the drawing seem to be the least permeable Sample LH-3 had been part of a narrow fin of rock inside one of the voids in the irregularly weathered section of this outcrop In LH-3, epoxy moved preferentially along several laminations and a very small fracture (Figure 54) Some laminations, such as those near the top, are effectively impermeable, while other laminations were easily penetrated by epoxy Fluid did not seem to move perpendicular to laminations A small fracture, oriented horizontally on the figure, inhibited fluid movement This fracture was also not visible prior to the application of epoxy III D SEM Observations Individual grains from the several of the outcrop units in Figure 50 (UH, LH and BH) were imaged with an electron microprobe with scanning electron micrograph (SEM) capability The purpose of the SEM work was to observe grain surface textures The 1 1 7

PAGE 126

Figure 53 Sample LH-1 was in the zone of enlarged fractures and conduits This part of the outcrop has two distinct layers, referred to here simply as "upper" and "lower" UH-1 was from the upper layer This layer of rock had cross beds, other sedimentary stuctures, and showed very heterogeneous weathering patterns There were many enlarged fractures and even conduits in the rock Gray areas were saturated with epoxy This sample had epoxy applied three times before it was saturated Small fractures and laminations exert a strong control on fluid flow The finer laminations near the base seem the least permeable The sample had the lowest permeability perpendicular to the plane of the figure 118

PAGE 127

Figure 54 Sample LH-4 was in the zone of enlarged fractures and conduits This part of the outcrop has two distinct layers, referred to here simply as "upper" and "lower" UH1 was from the upper layer This layer of rock had cross beds, other sedimentary stuctures, and showed very heterogeneous weathering patterns There were many enlarged fractures and even conduits in the rock In LH-3, epoxy moved preferentially along several laminations and a very small fracture LH-3 had been part of a narrow fin of rock inside one of the voids in the irregularly weathered section of this outcrop Some laminations are very impermeable, as is a small near-vertical fracture that was not visible prior to the application of epoxy 119

PAGE 128

SEM work was exploratory, and does not represent an exhaustive survey of the Hinckley sandstone The SEM findings are meant to compliment the permeability observations on rocks from the outcrop in Figure 50 The grains imaged had spontaneously fallen off the samples during the oven drying described in III C The samples were on separate trays, so the loose grains from different samples could be identified Removing grains with a tool might have introduced artificial textures Young (1988) describes a number of surface textures on quartz grains from arenite sandstones that he believes are solutional in origin He describes two classes of textures : surface reaction controlled and transport controlled A schematic drawing of the textures he discusses is shown on the bottom half of Figure 55 On the basis of this preliminary investigation, the solutional textures were equally common in both parts of the conduit-bearing unit (UH and LH) The grains in the central conduit-bearing unit (UH and LH) were characterized by extensive etching, pitting and embayments (Figures 55-58) These solutional features were also seen on BH grains, but to a much lesser extent A few scanning electron micrographs were taken of selected grains Figure 55 shows a grain from UH with an embayment on the grain surface running vertically through the middle of the image This embayment is about 300 microns long and 100 microns wide It is broad and shallow The grains surface outside of the embayment is smooth and euhedral, while the floor of the embayment is rough and pitted This surface texture is very similar to examples B and D from Young's cartoon below the micrograph Figure 56 shows micrographs from another grain from UH The larger scale view (A) shows a grain that is extensively pitted with only a few remnants of the smooth crystal surface The zoomed-in view (B) shows a small portion from the center of view A that hosts several angular notches These notches may be analogous to the V-etch pits from Young's cartoon (Figure 55, A) Figure 57 shows about half of a large quartz grain from LH Near the end of the grain, remnants of the original crystal surfaces can still be seen Most of the grain is heavily pitted Image B shows a small portion from the center of A, where some crystal faces and their intersections have been preserved The flat crystal faces appear to be removed first, 1 2 0

PAGE 129

Figure 55 Scanning Electron Micrograph of a grain from UH Figure 55 Scanning electron micrographs of a quartz grain from UH The photomicrograph shows a quartz grain with a long shallow embayment in the center of the photo The surrounding grain or overgrowth surfaces are smooth, while the surface of the embayment is pitted Scale in microns is shown on the image This feature is very similar to images B and D on Young's (1988) figure showing solutional textures on quartz (below) SURFACE REACTION CONTROLLED DISSOLUTION' 1 2 1 TRANSPORT CONTROLLED DISSOLUTION

PAGE 130

Figure 56 Scanning electron micrographs of a quartz grain from UH Photomicrograph A shows a grain that has been extensively etched The center of A has a small area with notches These notches are shown at a bigger scale in photomicrograph B These notches are similar to the V -notches, described by Young (1988) Scale in microns is shown by a bar on each image 1 2 2

PAGE 131

Figure 5 7 : Scanning Electron Micrograph of a grain from LH Figure 57 Scanning electron micrographs of a quartz grain from LH Photomicrograph A shows a grain that has been extensively etched Only rememants of grain or overgrouwth surfaces have not been removed by embayments Photomicrograph B shows detail of a former grain corner, where embayments have eaten in towards the corner This closis very similar to image C on Young's solution textures figure (Figure 50) Scale in microns is shown by a bar on each image 123

PAGE 132

1 2 4 leaving their intersections as smooth ridges in a fashion similar to C on Young's cartoon (Figure 55) Figure 58 displays two grains from LH Both A and B show remnants of the smooth euhedral grain surfaces surrounded by embayments and pitted surfaces III E Theoretical Solution of Quartz The Hinckley sandstone is a quartz arenite, composed dominantly of quartz grains and quartz cements (Tryhorn and Ojakangas, 1972) The features described above suggest that cement has been removed in places The removal of cements means solution of quartz Proposing solution as a major weathering mechanism appear to be at odds with the conventional wisdom that quartz is very resistant to chemical weathering (Goldich, 1938, in Langmuir, 1997) However, "resistant to weathering" is not the same as insoluble Comparison of S10 2 and CaCO 3 solubilities in Table 9 illustrates the range of solubilities between quartz and the typical host rock for karst, limestone (approximated as pure calcite) Table 10 (after Langmuir, 1997) As shown in Table 9, the equilibrium solubility (25C) of calcite at typical soil P C ,02 is only two times greater than amorphous silica, and about forty times greater than the solubility of crystalline SiO 2 Modern conditions in Minnesota are better described at 10C, where calcite is almost 4 times more soluble than amorphous silica and about 80 times more soluble than quartz Limestone is not pure calcite, but contains aragonite, magnesium up to the range of stoichiometric dolomite, as well as other detritus and trace elements The impurity in limestone generally decreases its solubility Mineral Solubility (ppm) @ 25C Solubility (ppm) @ 10C Si0 2 (quartz) 6 .6 4 SiO 2 (amorphous) 116 85 CaCO 3 ( C02 = 10-1 .5 bar) 250 318

PAGE 133

Figure 5 8 Scanning Electron Micrographs of grains from LH Figure 58 Scanning electron micrographs of two quartz grains from LH Photomicrograph A shows a grain that has been extensively etched Only remnants of grain or overgrowth surfaces have not been removed by embayments, surrounded by a pitted surface Photomicrograph B shows etching and embayments on another quartz grain Scale in microns is shown by a bar on each image 125

PAGE 134

* where X is : t nH2O + Si-O = n(H20)(Si-O ) 1 2 6 In water, quartz forms the aqueous species monomeric silicic acid by the reaction shown below The solubility constant for this reaction has been modeled as a function of temperature (Barnes, 1980, in Langmuir, 1997) SiO2( g tz) + H 2 O = H 4 SiO4 € log Ksp = 1 .8814 2 .028x10" 3 T 1560 .46/T t (Kelvin) Quartz solubility increases with temperature The Hinckley and Fond du Lac sandstones are sediments deposited adjacent to the 1 .1 Ga Midcontinent Rift System, which was the site of volcanic activity for about 40 million years The solubility of Si02 is also a function of pH For quartz, solution rates are pHindependent near 7 (Brady and Walther, 1990) Above 7 .5, solution rates increase with increasing pH Brady and Walther have modeled this behavior with a rate law : Kb = kT/h K(X) al-120 [Si-O"]y Organic acids and humic substances have the capacity to enhance quartz solution There are two proposed mechanisms The first involves the formation of an aqueous organic-silica complex, while the second is a decrease in activation energy for the quartz solution reaction (Bennet, 1991) An aqueous organic-silica complex could increase the solubility of quartz directly, by removing molecules from the effective ionic strength These complexes could be the byproducts of microbial activity Bennet and Siegel (1987) note that quartz was observed quickly in water at a petroleum spill site that had been subject to biological remediation Unfortunately, the complex composition of organic acids and humic substances make is difficult to distinguish between competing electrolyte reactions (Bennet, 1991 ; Milne et al, 1995) A suggested two-part process is based on UV-difference spectroscopy : a weak electron donor-acceptor forms quickly at neutral pH, followed by the formation of a stronger complex (Bennet, 1991) Humic substances may also decrease the activation energy of the solution reaction Activation energy is decreased when framework bonds in the quartz crystal lattice are

PAGE 135

broken by strong interactions between the organic acid and the SiO 2 surface Terminal hydroxyl protons of the organic acid interact with the crystal surface, forming either a hydrogen bond or organic-silica ester (Bennet, 1991) Both the aqueous complexes and activated surface-complexes described above could be effective at increasing quartz solution at near-neutral pH The result of the reduced activation energy was an approximate 100% increase in apparent solubility (Bennet, 1991) Quartz is the most resistant of the common minerals to surface weathering It is also soluble enough to develop karst permeability structures 1 2 7

PAGE 136

REFERENCES Bennet, P .C ., 1991 Quartz dissolution in organic-rich aqueous systems Geochimica et Cosmochimica Acta 55 : 1781-1797 Bennet, P .C ., and Siegel, D .I ., 1987 Increased solubility of quartz in water due to complexation by dissolved organic compounds Nature, 326 : 684-687 Brady, P .V ., and Walther, J .V ., 1990 Kinetics of quartz dissolution at low temperatures Chemical Geology, 82 : 253-264 Goldich, S .S ., 1938 A study in rock weathering Geology, 46 : 17-58 Langmuir, D ., 1997 Aqueous Environmental Geochemistry Prentice-Hall, Inc ., New Jersey, 600 pp Milne, C j ., Kinniburgh, D .G ., DeWit, J .C .M ., VanRiemsdijk, W .H ., and Koopal, L .K ., 1995 Analysis of proton binding in a peat humic substance using a simple electrostatic model Geochimica et Cosmochimica Acta, 59 : 1101-1112 Rimstidt, J .D ., and Barnes, H .L ., 1980 The kinetics of silica water reactions Geochimica et Cosmochimica Acta, 44 :1683-99 Tryhorn, A .D ., and Ojakangas, R .W ., 1972 Sedimentation of the Hinckley Sandstone of east-central Minnesota In Sims, P .K ., and Morey, G .B ., eds ., Geology of Minnesota : A centennial volume Minnesota Geological Survey, St Paul, Minnesota, 416-424 Young, R .W ., 1986 Tower Karst in Sandstone : Bungle Bungle massif, northwestern Australia Zeitshrift fur Geomorpholgie, 30 :189-202 Young, R .W ., 1987 Sandstone landforms of the tropical East Kimberley region ; northwestern Australia Journal of Geology, 95 : 205-218 Young, R .W ., 1988 Quartz etching and sandstone Karst : Examples from the East Kimberleys, Northwestern Australia Zeitshrift fur Geomorphology, 32 : 409-423 1 2 8

PAGE 137

Discussion The recurring theme in this thesis is heterogeneity The Hinckley Sandstone in the study area is a sandstone karst : karst systems are defined by their heterogeneity Karst systems are inherently more than landforms shaped only by surface weathering Dynamic processes at work below the ground surface drive the development of karst features The existence of karst features implies that transportation in the underlying hydrology is dominated by turbulent flow, even though porous media flow is present in parts of the karst aquifer The Hinckley Sandstone aquifer has a significant component of turbulent flow moving through voids : caves, conduits, enlarged fractures and/or high-permeability zones Turbulent flow can move quickly and transport large volumes of water and suspended material The karst features in the study area (caves, sinkholes, streamsinks, and springs) are closely analogous to the geomorphologic features found in traditional karst systems Sinkholes in the study area function as karst features : they have clearly crosscutting stratigraphy due to the transportation of water and material directly into the subsurface ; during precipitation events, sinkholes have been directly observed quickly draining water into the subsurface A number of the sinkholes can be identified as active or seasonal streamsinks, as they have either active or dry streambeds leading into them The subsurface heterogeneity implied by the existence of karst features is supported by the hydrochemistry of the study area as well as by the observed permeability structure of the Hinckley Sandstone : the hydrochemistry portrays the Hinckley Sandstone as a complex aquifer that hosts a variety of waters with both short and long residence times, has distinct flow paths that do not mix significantly, has some flow paths that are obviously mixed and displays the movement of water over long distances These are characteristics of karst flow and not of porous media flow The sinkholes contribute to just one of several flow systems found in the Hinckley Sandstone Water in the karst system is moved quickly into the subsurface for this reason, these waters would probably be oxygenated, and similar to rainwater and surface waters Near areas of human occupation, these waters would show the impact of human 1 2 9

PAGE 138

activities, such as high Cl/Br ratios, elevated .'conceifttrations -of chloride, nitrates, and possibly human contaminants such as coliform bacteria In the Hinckley Sandstone, there are many samples of Ca/Mg/HCO 3 waters with modern tritium and anthropogenic Cl/Br ratios some samples have been contaminated with coliform bacteria ; these waters have short residence times There are also Ca/Mg/HCO 3 waters in the Hinckley Sandstone that are very reduced Carbon radiometric dating suggests that a significant number of samples have been reduced enough to undergo methanogenesis Many of the areas in central Pine County that are not cultivated are wetlands ; the wetlands can be areas of recharge or of groundwater discharge During recharge through a wetland, the abundance of organic material will promote reduction reactions The highly reduced wells and springs had a range of ages, depending of what flow paths they have followed There is also a group of samples in the Hinckley Sandstone with very old characteristics in the age models ( 3 H, Cl/Br, and 14 C) that seem to have a component of basalt groundwater This water is characterized by higher Sr/Ca and Sr/Mg ratios, as well as high concentrations of boron These waters were only found in springs from the Hinckley Good drinking water is found at relatively shallow depth in the Hinckley, so that wells are not drilled to the depths where these waters are surely found in the bedrock It is significant that oxygenated and very reduced waters, as well as waters that are effectively modern and up to 4,000 ybp can discharge within meters of each other This type of spring distribution could not occur unless the individual waters were following relatively discreet flow paths in close proximity in the bedrock This represents an impressive heterogeneity in the sandstone with respect to water flow The heterogeneity in this system is present at scales from hand sample to regional flow paths The large-scale heterogeneity is seen in the hydrochemistry discussed above The small-scale heterogeneity was seen in outcrop and in hand samples In outcrop, the Hinckley sandstone can have very different strength and weathering patterns over a few meters This heterogeneity can be either vertically (in the same outcrop) or horizontally (between outcrops) Permeability experiments on samples from the Hinckley Sandstone also suggest that this rock is heterogeneous to fluid flow Outcrop and hand sample scale 1 3 0

PAGE 139

sedimentary and structural features such as cross beds, laminations and fractures can exert a strong control on fluid movement, alternately facilitating or inhibiting flow The non-uniform nature of the Hinckley Sandstone seen in groundwater chemistry and permeability experiments is the result of the same bedrock heterogeneity that drives this active karst system 1 3 1

PAGE 140

Table 1 Sinkhole Database 'All UTM locations use the 1983 North American Datum AD83 132 Sinkhole # t Name t Township Range Section UTME UTMN MN58 :D0001 Big Sinkhole/MN58 :B0001 N 43 W 19 31 516234 5113122 MN58 :D0002 MN58 :B0002 N 43 W 19 29 516998 5114173 MN58 :D0003 MN58 :B0017 N 43 W 19 29 516964 5114132 MN58 :D0004 Kroon/ MN58 :B0003 N 43 W 19 29 517540 5114388 MN58 :D0005 Kroon/MN58 :B0014 N 43 W 19 29 517549 5114356 MN58 :D0006 Kroon/MN58 :B0015 N 43 W 19 29 517440 5114222 MN58 :D0007 MN58 :B0004 N 43 W 20 36 514461 5113192 MN58 :D0008 Sahlen N 43 W 19 31 515420 5112006 MN58 :D0009 MN58 :130016 N 43 W 19 30 515243 5113405 MN58 :D0010 LDC2 N 43 W 20 13 513285 5117646 MN58 :D0011 LDC3 N 43 W 20 13 513291 5117653 MN58 :D0012 N 43 W 19 10 519564 5119014 MN58 :D0013 N 43 W 19 10 519554 5118933 MN58 :D0014 MN58 :B0024 N 43 W 19 10 519660 5118978 MN58 :D0015 N 43 W 19 10 519858 5118890 MN58 :D0016 N 43 W 19 10 519788 5119067 MN58 :D0017 N 43 W 19 10 519567 5119191 MN58 :D0018 N 43 W 19 10 519897 5119224 MN58 :D0019 N 43 W 19 9 518528 5119238 MN58 :D0020 Field Stone, Thienesson N 43 W 19 21 518745 5115279 MN58 :D0021 Thienesson N 43 W 19 20 518368 5115750 MN58 :D0022 Thienesson N 43 W 19 20 518346 5115761 MN58 :D0023 Old Ford Truck, Thienesson N 43 W 19 20 518314 5115775 MN58 :D0024 Thienesson N 43 W 19 20 518292 5115762 MN58 :D0025 Thienesson compound w\ D26 N 43 W 19 20 518303 5115782 MN58 :D0026 Thienesson compound w\ D25 N 43 W 19 20 518298 5115786 MN58 :D0027 Thienesson compound w\ D28 N 43 W 19 20 518286 5115791 MN58 :D0028 Thienesson compound w\ D27 N 43 W 19 20 518274 5115797 MN58 :D0029 Thienesson this is two collapses N 43 W 19 20 518315 5115836 MN58 :D0030 Thienesson N 43 W 19 20 518345 5115849 MN58 :D0031 Thienesson N 43 W 19 20 518334 5115857 MN58 :D0032 Thienesson N 43 W 19 20 518320 5115880 MN58 :D0033 Thienesson N 43 W 19 20 518365 5115853 MN58 :D0034 ditch, east side N 43 W 19 14 521610 5117416 MN58 :D0035 ditch, east side N 43 W 19 14 521610 5117409 MN58 :D0036 ditch, east side N 43 W 19 14 521612 5117398 MN58 :D0037 MN58 :B0018 N 43 W 19 30 516339 5113231 MN58 :D0038 LDC1, small, side of trail N 43 W 20 13 513274 5117411 MN58 :D0039 sick skunk area N 43 W 19 29 518323 5114397 MN58 :D0040 sick skunk area N 43 W 19 29 518294 5114413 MN58 :D0041 sick skunk area N 43 W 19 29 518270 5114435 MN58 :D0042 sick skunk area N 43 W 19 29 518178 5114481 MN58 :D0043 sick skunk area N 43 W 19 29 518194 5114516

PAGE 141

Table 1 Sinkhole Database 1 3 3 Sinkhole # t Name t Township Range Section UTME UTMN MN58 :D0044 sick skunk area N 43 W 19 29 518172 5114544 MN58 :D0045 sick skunk area N 43 W 19 29 518101 5114564 MN58 :D0046 sick skunk area N 43 W 19 29 518144 5114522 MN58 :D0047 sick skunk area N 43 W 19 29 518240 5114579 MN58 :D0048 Boulder drain, SE of Askov N 43 W 19 29 518334 5114434 MN58 :D0049 U2, Banning East N 42 W 20 3 511512 5110577 MN58 :D0050 MN58 :B0022, Beaver Sink N 43 W 19 1 523784 5121274 MN58 :D0051 MN58 :B0023, Beaver Sink N 44 W 19 36 523803 5121283 MN58 :D0052 Beaver Sink N 44 W 19 36 523822 5121314 MN58 :D0053 Beaver Sink N 44 W 19 36 523827 5121314 MN58 :D0054 Beaver Sink N 44 W 19 36 523837 5121313 MN58 :D0057 Shown by Al Jensen N 43 W 19 9 519394 5119037 MN58 :D0058 N 43 W 19 9 519409 5118922 MN58 :D0059 N 43 W 19 9 519294 5118997 MN58 :D0060 N 44 W 18 27 529717 5124455 MN58 :D0061 N 43 W 19 2 521644 5121119 MN58 :D0062 in yard N 43 W 19 2 521644 5121058 MN58 :D0063 N 43 W 19 2 521668 5121087 MN58 :D0064 N 43 W 19 2 521636 5121001 MN58 :D0065 N 43 W 19 2 521655 5120959 MN58 :D0066 N 43 W 19 2 521469 5121027 MN58 :D0067 N 43 W 19 2 521214 5121232 MN58 :D0068 Foyt N 43 W 19 15 520015 5117995 MN58 :D0069 Foyt N 43 W 19 15 520042 5117710 MN58 :D0070 Anderman N 43 W 19 22 520953 5116236 MN58 :D0071 Anderman N 43 W 19 22 521016 5116332 MN58 :D0072 Anderman N 43 W 19 22 520910 5116417 MN58 :D0076 sinks above Robinson N 42 W 20 10 510611 5109929 MN58 :D0077 sinks above Robinson N 42 W 20 10 510634 5109852 MN58 :D0080 woods north of sewage ponds N 43 W 19 29 517080 5114261 MN58 :D0081 woods north of sewage ponds N 43 W 19 29 517087 5114271 MN58 :D0082 woods north of sewage ponds N 43 W 19 29 517285 5114340 MN58 :D0083 east of sewage ponds N 43 W 19 29 517315 5114156 MN58 :D0084 due north of Petersen Spring N 43 W 19 22 521565 5115977 MN58 :D0085 due north of Petersen Spring N 43 W 19 22 521544 5115926 MN58 :D0086 Ken Nelsen N 43 W 19 29 517679 5113914 MN58 :D0087 Ken Nelsen N 43 W 19 29 517733 5113939 MN58 :D0088 Ken Nelsen N 43 W 19 29 517745 5113966 MN58 :D0089 Ken Nelsen N 43 W 19 29 517910 5113827 MN58 :D0091 barnyard, filled N 43 W 19 18 515223 5116585 MN58 :D0092 barnyard, filled N 43 W 19 18 515203 5116610 MN58 :D0093 barnyard, filled N 43 W 19 18 515224 5116562 MN58 :D0094 pasture, filled All sort of weird N 43 W 19 18 515279 5116450 MN58 :D0095 MN58 :B0012, Banning N 43 W 19 26 513234 5113787 MN58 :D0096 _Banning park, south N 43 W 19 26 513234 5113790

PAGE 142

Table 1 Sinkhole Database 1 3 4 Sinkhole # t Name t Township Range Section UTME' UTMN MN58 :D0097 Banning park, south N 43 W19 26 513239 5113863 MN58 :D0098 MN58 :B0010, Banning N 43 W 19 26 513243 5113863 MN58 :D0099 Banning park, south N 43 W 19 26 513273 5113911 MN58 :DO101 Campground, discovery sinkhole N 42 W 20 3 511358 5111350 MN58 :D0102 Banning West, near campground N 42 W 20 3 511339 5111253 MN58 :D0103 Banning West, near campground N 42 W'20 3 511345 5111227 MN58 :D0104 Banning West, near campground N 42 W 20 3 511301 5111121 MN58 :D0105 Banning West, near campground N 42 W 20 3 511248 5111095 MN58 :D0106 Banning West, near campground N 42 W 20 3 511246 5111088 MN58 :D0107 Banning West, near campground N 42 W 20 3 511243 5111086 MN58 :D0108 Banning West, near campground N 42 W 20 3 511246 5111081 MN58 :D0109 Banning West, near campground N 42 W 20 3 511245 5111077 MN58 :D0110 Banning West, near campground N 42 W 20 3 511244 5111076 MN58 :D0111 Banning West, near campground N 42 W 20 3 511236 5110920 MN58 :D0113 MN58 :B0011, Banning South N 43 W 19 26 513248 5113888 MN58 :D0116 near rest stop N 43 W 20 14 512709 5117453 MN58 :D0117 near rest stop N 43 W 20 14 512716 5117402 MN58 :D0118 near rest stop N 43 W 20 14 512744 5117143 MN58 :D0119 MN58 :B0006, Eklund N 43 W-19 28 518920 5114653 MN58 :D0120 W2 Banning East N 42 W 20 3 511525 5110570 MN58 :D0121 W3 Banning East N 42 W 20 3 511528 5110573 MN58 :D0122 W4 Banning East N 42 W 20 3 511534 5110579 MN58 :D0123 W5 Banning East N 42 W 20 3 511539 5110583 MN58 :D0125 Af2 Banning East N 42 W20 3 511367 5110271 MN58 :D0126 Zinkhan MN58 :B0007 N 43 W 19 11 521816 5119233 MN58 :D0127 Zinkhan Excavation 4 N 43 W 19 11 521773 5119235 MN58 :D0128 Petersen Spring/Anderman area/so N 43 W 19 22 521039 5115971 MN58 :D0129 Petersen Spring/Anderman area/s< N 43 W 19 22 521014 5115915 MN58 :D0130 Petersen Spring/Anderman area/s< N 43 W 19 22 520985 5115889 MN58 :D0131 30 meters futher along same headir N 43 W 19 22 520962 5115870 MN58 :D0132 MN58 :B0013 N 43 W 19 22 521077 5115935 MN58 :D0133 Petersen Spring/Anderman area N 43 W 19 22 520950 5116003 MN58 :D0134 Petersen Spring/Anderman area N 43 W 19 22 520398 5116169 MN58 :D0135 Petersen Spring/Anderman area N 43 W 19 22 520438 5116308 MN58 :D0136 Petersen Spring/Anderman area N 43 W 19 22 520435 5116352 MN58 :D0137 Petersen Spring/Anderman area N 43 W 19 22 520414 5116318 MN58 :D0138 near rest stop N 43 W 20 28 512053 5116011 MN58 :D0139 Petersen Spring/Anderman area N 43 W 19 22 521081 5116038 MN58 :D0140 Log drive creek area N 43 W 20 14 513578 5116535 MN58 :D0141 Log drive creek area N 43 W 20 14 513594 5116525 MN58 :D0142 Log drive creek area N 43 W 20 14 513603 5116510 MN58 :D0143 Log drive creek area N 43 W 20 14 513625 5116530 MN58 :D0144 dig #3, log drive creek N 43 W 20 13 513597 5116584 MN58 :D0145 Log drive creek area N 43 W 20 14 513624 5116631 MN58 :D0146 mud hole, log drive creek N 43 W 20 14 513637 5116611

PAGE 143

Table 1 Sinkhole Database 1 3 5 Sinkhole # t Name t Township Range Section UTME UTMN MN58 :D0147 Log drive creek area N 43 W 20 14 513641 5116637 MN58 :D0148 Log drive creek area N 43 W 20 14 513843 5116633 MN58 :D0149 Log drive creek area N 43 W 20 14 513672 5116365 MN58 :D0180 Zinkhan N 43 W 19 11 521874 5119228 MN58 :D0181 Zinkhan N 43 W 19 11 521908 5119261 MN58 :D0182 Zinkhan N 43 W 19 11 521943 5119266 MN58 :D0200 Banning East A N 42 W 20 3 511583 5110628 MN58 :D0201 Banning East B N 42 W 20 3 511596 5110630 MN58 :D0202 Banning East C N 42 W 20 3 511579 5110670 MN58 :D0203 Banning East D N 42 W 20 3 511578 5110667 MN58 :D0204 Banning East E N 42 W 20 3 511804 5110824 MN58 :D0205 Banning East El N 42 W 20 3 511778 5110834 MN58 :D0206 Banning East F N 42 W 20 3 511801 5110847 MN58 :D0207 Banning East G N 42 W 20 3 511799 5110863 MN58 :D0208 Banning East H N 42 W 20 3 511798 5110868 MN58 :D0209 Banning East I N 42 W 20 3 511803 5110866 MN58 :D0211 Banning East K N 42 W 20 3 511691 5110682 MN58 :D0212 Banning East L N 42 W 20 3 511638 5110637 MN58 :D0213 Banning East N N 42 W 20 3 511378 5110452 MN58 :D0214 Banning East 0 N 42 W 20 3 511368 5110420 MN58 :D0215 Banning East P N 42 W 20 3 511393 5110455 MN58 :D0216 Banning East Q N 42 W 20 3 511420 5110491 MN58 :D0217 Banning East R N 42 W 20 3 511459 5110551 MN58 :D0218 Banning East S N 42 W 20 3 511473 5110569 MN58 :D0219 Banning East T N 42 W 20 3 511498 5110556 MN58 :D0220 Banning East U1 N 42 W 20 3 511508 5110571 MN58 :D0221 Banning East V N 42 W 20 3 511511 5110574 MN58 :D0222 Excavation #1, W1 N 42 W 20 3 511530 5110575 MN58 :D0223 Banning East X N 42 W 20 3 511546 5110570 MN58 :D0224 Banning East Y N 42 W 20 3 511361 5110381 MN58 :D0225 Banning East Z N 42 W 20 3 511357 5110383 MN58 :D0226 Banning East AA N 42 W 20 3 511339 5110410 MN58 :D0227 Banning East AB N 42 W 20 3 511333 5110340 MN58 :D0228 Banning East AD N 42 W 20 3 511328 5110271 MN58 :D0229 Banning East AE N 42 W 20 3 511346 5110270 MN58 :D0230 Banning EastAF2 N 42 W 20 3 511359 5110271 MN58 :D0231 Banning East AG N 42 W 20 3 511290 5110314 MN58 :D0232 Banning East AH N 42 W 20 3 511239 5110341 MN58 :D0233 Banning East AI more to east? N 42 W 20 3 511204 5110319 MN58 :D0234 Banning East AJ-cluster of 6 N 42 W 20 3 511172 5110280 MN58 :D0235 Banning East AK-cluster N 42 W 20 3 511162 5110264 MN58 :D0236 Banning East AL2 N 42 W 20 3 511088 5110231 MN58 :D0237 Banning East AL3 N 42 W 20 3 511141 5110248 MN58 :D0238 Banning East AM N 42 W 20 3 511115 5110224 MN58 :D0239 Banning East AM N 42 W 20 3 511095 5110245,

PAGE 144

Table 1 Sinkhole Database 1 3 6 Sinkhole # t Name t Township Range Section UTME UTMN MN58 :D0240 Banning East AM1 N 42 W 20 3 511110 5110241 MN58 :D0241 Banning East AN N 42 W 20 3 511022 5110207 MN58 :D0242 Banning East BA N 42 W 20 3 511238 5110251 MN58 :D0243 Banning East BB N 42 W 20 3 511236 5110255 MN58 :D0244 Banning East BC N 42 W 20 3 511229 5110239 MN58 :D0245 Banning East BD N 42 W 26 3 511187 5110218 MN58 :D0246 Banning East BE N 42 W 20 3 511179 5110206 MN58 :D0247 Hell's Gate NE, HR2 N 42 W 20 3 511830 5111337 MN58 :D0248 Hell's Gate NE, HR1 N 43 W 20 35 512087 5111686 MN58 :D0249 Log Drive Creek 1 of 2 N 43 W 20 14 513457 5116440 MN58 :D0250 Log Drive Creek, 2 of 2 N 43 W 20 14 513454 5116471 MN58 :D0251 Log Drive Creek N 43 W 20 13 513578 5116539 MN58 :D0252 Log Drive Creek N 43 W 20 13 513604 5116522 MN58 :D0253 Log Drive Creek N 43 W 20 13 513617 5116488 MN58 :D0254 Log Drive Creek N 43 W 20' 13 513620 5116540 MN58 :D0255 Log Drive Creek N 43 W 20 13 513599 5116565 MN58 :D0256 Triple Hole-Log Drive Creek N 43 W 20 13 513600 5116566 MN58 :D0257 Log Drive Creek near LDC N 43 W 20 13 513463 5117181 MN58 :D0300 MN58 :B0019 N 43 W 19 19 516011 5115135 MN58 :D0301 MN58 :B0020 N 43 W 19 19 515963 5115224 MN58 :D0302 N 43 W 19 19 515902 5115224 MN58 :D0303 N 43 W 19 15 521213 5117953 MN58 :D0304 N 43 W 19 15 521175 5117955 MN58 :D0305 N 43 W 19 10 521429 5118046 MN58 :D0306 N 43 W 19 10 521435 5118079 MN58 :D0307 N 43 W 19 10 521388 5118058 MN58 :D0308 N 43 W 19 15 521540 5117242 MN58 :D0309 N 43 W 19 15 521560 5117373 MN5800310 N 43 W 19 15 521548 5117348 MN5S :D0311 N 43 W 19 15 521572 5117344 MN58 :D0312 N 43 W 19 15 521542 5117328 MN58 :D0313 N 43 W 19 15 521558 5117305 MN58 :D0314 mailbox N 43 W 19 14 521628 5117193 MN58 :D0315 across road N 43 W 19 15 521560 5117137 MN58 :130316 south of 58DO0314 N 43 W 19 14 521707 5116892 MN58 :D0317 N 43 W 19 14 521626 5118024 MN58 :D0318 N 43 W 19 14 522652 5117723 MN58 :D0319 N 43 W 19 14 522604 5117653 MN58 :D0320 With D30, D331 N 43 W 19 11 523190 5118110 MN58 :D0321 near Beaver Sink N 43 W 19 2 523183 5121189 MN58 :D0322 near Beaver Sink N 43 W 19 1 523213 5121145 MN58 :D0323 near Beaver Sink N 43 W 19 1 523220 5121098 MN58 :D0324 road detours N 43 W 19 2 521822 5119663 MN58 :D0325 road detours N 43 W 19 2 521899 5119700

PAGE 145

Table 1 Sinkhole Database 1 3 7 Sinkhole # t Name t Township Range Section UTME UTMN MN58 :D0327 MN58 :B0021 N 43 W 19 10 521296 5118614 MN58 :D0328 N 43 W 19 10 521513 5118607 MN58 :D0329 N 43 W 19 11 521624 5118483 MN58 :D0330 N 43 W 19 11 523189 5118110 MN58 :D0331 N 43 W 19 11 523189 5118112 MN58 :D0332 N 43 W 19 22 521355 5116352 MN58 :D0333 N 43 W 19 22 520878 5116413 MN58 :D0334 N 43 W 19 16 519795 5116482 MN58 :D0335 N 43 W 19 16 519741 5116526 MN58 :D0336 filled in 1930's N 43 W 19 10 520258 5118922 MN58 :D0337 near Kettle Rest Area N 44 W 18 32 526410 5121347 MN58 :D0338 near Kettle Rest Area N 43 W 18 6 526327 5121272 MN58 :D0339 N 43 W 19 21 519283 5115625 MN58 :D0340 N 43 W 19 21 519052 5114879 MN58 :D0341 North of road N 43 W 19 20 518363 5114842 MN58 :D0342 North of road N 43 W 19 20 518302 5114879 MN58 :D0343 Oleson N 43 W 19 30 516219 5113801 MN58 :D0344 Oleson N 43 W 19 30 516312 5113695 MN58 :D0345 Oleson N 43 W 19 30 516346 5114137 MN58 :D0346 near cemetery N 43 W 19 30 516147 5114670 MN58 :D0347 near cemetery N 43 W 19 30 516173 5114571 MN58 :D0348 along county ditch/stream N 43 W 19 29 517483 5114778 MN58 :D0349 along county ditch/stream N 43 W 19 29 517537 5114745 MN58 :D0350 near stramsink/D5 N 43 W 19 29 517566 5114391 MN58 :D0351 near road N 43 W 19 29 517297 5113271 MN58 :D0352 East of Ken Nelsen N 43 W 19 29 517718 5113902 MN58 :D0353 East of Ken Nelsen N 43 W 19 29 517770 5113931 MN58 :D0354 East of Ken Nelsen N 43 W 19 29 517677 5114130 MN58 :D0355 Excavation #2 N 43 W 19 29 517286 5114114 MN58 :D0356 Banning West near campground N 42 W 20 3 511334 5111352 MN58 :D0400 N 44 W 18 23 531417 5125381 MN58 :D0401 N 44 W 18 23 531429 5125350 MN58 :D0402 N 44 W 18 15 531109 5126109 MN58 :D0403 N 44 W 18 14 532545 5126091 MN58 :D0404 N 44 W 18 15 531144 5126135 MN58 :D0512 N 43 W 19 10 519867 5119191 MN58 :D0513 N 43 W 19 9 519224 5119224 MN58 :D0515 Also MN58 :B0025 N 43 W 19 16 518747 5117551

PAGE 146

Table 2 Streamsink Database *All UTM locations use the 1983 North American Datum (NAD83) -These sinkhole numbers are a shortened version of the sinkhole numbers defined in Table 1 1 3 8 Streamsink # t Name t Township Rang Section UTME* UTMN MN58 :'B0001 Big Sinkhole (D001N 43 W 19 31 516234 5113122 MN58 :B0002 Also D002 N 43 W 19 29 516980 5114165 MN58 :B0003 Kroon (also D004) N 43 W 19 29 517540 5114389 MN58 :B0004 Kroon (also D007) N 43 W 20 36 514461 5113192 MN58 :B0005 B5 N 43 W 19 -21 518040 5115826 MN58 :B0006 Eklund (also D119) N 43 W 19 28 518920 5114653 MN58 :B0007 Also D126 N 43 W 19 11 521816 5119233 MN58 :B0009 Also D515 N 43 W 19 16 518747 5117551 MN58 :B0010 2 of pair (also D098) N 43 W 19 26 513243 5113863 MN58 :B0011 Also D113 N 43 W 19 26 513248 5113888 MN58 :B0012 1 of pair (also D095) N 43 W 19 26 513234 5113787 MN58 :B0013 Also D132 N 43 W 19 22 521077 5115935 MN58 :B0014 Also D005 N 43 W 19 29 517549 5114356 MN58 :B0015 Also D006 N 43 W 19 29 517440 5114222 MN58 :B0016 Also D009 N 43 W 19 10 515243 5113405 MN58 :B0017 Also D003 N 43 W 19 29 516964 5114132 MN58 :B0018 2nd Drain (D037) N 43 W 19 30 516339 5113231 MN58 :B0019 Also D300 N 43 W 19 19 516011 5115135 MN58 :B0020 Also D301 N 43 W 19 19 515963 5115224 MN58 :B0021 Also D327 N 43 W 19 10 521296 5118614 MN58 :B0022 Beaver Sink (D050) N 43 W 19 1 523784 5121274 MN58 :B0023 Beaver Sink (D051) N 44 W 19 36 523803 5121283 MN58 :B0024 Also D014 N 43 W 19 10 519660 5118978 MN58 :B0025 Als) D515 N 43 W 19 16 518747 5117551

PAGE 147

Table 3 Spring Database *All UTM locations use the 1983 North American Datum (NAD83) (+) locations are estimated 1 3 9 Spring # t Name t Township Range Section UTM E* UTM N MN58 :A0001 Partridge Creek Spring N 43 W 19 22 521582 5115774 MN58 :A0002 Gushing Orange Spring N 42 W 20 22 510332 5106182 MN58 :A0003 Spring A3 N 43 W 20 26 512702 5113817 MN58 :A0004 Spring A4 N 43 W 20 26 512777 5113401 MN58 :A0005 Beaver Boil N 43 W 20 14 512789 5117938 MN58 :A0006 Ferrooxidans Rise N 43 W 20 23 512967 5115217 MN58 :A0007 Clear Spring N 43 W 20 26 512859 5114231 MN58 :A0008 Orange Mound Spring N 43 W 20 26 512874 5114217 MN58 :A0009 No Charge Spring N 43 W 20 20 513149 5113838 MN58 :A0010 Mid Point Spring N 43 W 20 26 513120 5113368 MN58 :A0011 Gray Beaver Spring N 43 W 20 25 513031 5112212 MN58 :A0012 Orange Boil Spring N 42 W 20 3 510763 5110007 MN58 :A0013 Frog Spring (Clear) N 42 W 20 15 510828 5107073 MN58 :A0014 Orange Scum Spring N 42 W 20 15 510647 5106828 MN58 :A0015 Hi Conductivity Spring N 42 W 20 15 511096 5107233 MN58 :A0016 Orange Seeps N 42 W 20 15 510923 5106969 MN58 :A0017 Orange Seeps 11 N 42 W 20 15 511118 5107323 MN58 :A0018 Hi Conduct II N 42 W 20 15 511124 5107370 MN58 :A0019 Hike Spring N 42 W 20 15 511179 5107389 MN58 :A0020 Filip Spring N 41 W 16 24 544121 5097159 MN58 :A0021 PSP-1 N 42 W 20 10 511165 5108509 MN58 :A0022 PSP-2 N 42 W 20 10 511067 5109080 MN58 :A0023 PSP-3 N 42 W 20 10 510855 5110065 MN58 :A0024 PSP-6 N 40 W 18 15 533308 5088396 MN58 :A0025 PSP-4 N 40 W 18 15 533358 5088640 MN58 :A0026 PSP-5 N 40 W 18 15 533306 5088612 MN58 :A0027 KR 708 @ Hwy 22 (+) N 43 W 21 24 504520 5115424 MN58 :A0028 KR 748 @ Hwy 48 (+) N 41 W 20 23 512235 5095386 MN58 :A0030 Banning Park 766 (+) N 43 W 20 35 512730 5112328 MN58 :A0031 Banning Park Large Bench N 42 W 20 3 511298 5110645 MN58 :A0032 Hell's Gate Boat Ramp (+) N 43 W 20 35 512778 5112305 MN58 :A0033 Banning Overflow N 42 W 20 3 510805 5110264

PAGE 148

Table 4 Water Sam le Locations 1 4 0 ID# t Name t Type Unique # Twp Range Section UTM E* UTM N 1 Askov 3 MU 449381 N 43 W 19 20 518295 5114848 2 Askov 4 MU 449382 N43 W 19 20 518314 5114848 3 Finlayson 1 MU 217182 N 43 W 20 18 506288 5116666 4 Hinckley 3 MU 538117 N 41 W21 24 504809 5095539 5 Hinckley 4 MU 562973 N 41 W20 30 506434 5094964 6 Rutledge PS 465946 N44 W20 30 510317 5123432 7 Sandstone 1 MU 242053 N 42 W20 10 510964 5108963 8 Sandstone 2 MU 219291 N42 W20 10 510995 5108786 9 Banning SP Campground PS 416555 N 42 W 20 3 511289 5111464 10 Banning SP Picnic Area PS 574578 N 43 W 20 35 512370 5112260 11 Banning SP Manager DO W03120 N 43 W20 27 511440 5113313 12 Banning SP Shop PS 574577 N 43 W 20 34 511632 5113147 13 Banning Junction PS 493869 N 43 W 20 27 511337 5113940 14 Banning Sport & Gun PS 473703 N 43 W 20 27 511183 5113931 15 Kettle Rest Area PS 219400 N 43 W 20 14 512508 5117545 16 Bruno Fire Department PS 507417 N44 W 18 19 526053 5125069 17 End Zone Bar PS 142934 N44 W20 28 510299 5123116 18 1st Presby Church, Hinckely PS 575167 N 41 W 20 20 507438 5095422 19 St Croix SP Big Eddy PS 436792 N40 W 19 19 518624 5087701 20 St Croix Little Yellow Banks PS W30064 N40 W 18 1 535437 5091442 21 St Croix Haven PS N/A N 41 W 17 24 542950 5096944 22 St Croix Girls Camp PS N/A N 41 W 17 10 540218 5099904 23 Camp St Croix PS N/A N 42 W 17 28 539327 5104051 24 Pathfinder 1 PS 467676 N 41 W 18 28 529106 5094902 25 Pathfinder 3 PS 172459 N 41 W 18 28 528325 5095157 26 Pathfinder Village PS 172468 N 41 W 18 29 527782 5094233 27 Pathfinder 2 (Golf Course) PS 467698 N 41 W 18 29 527617 5094678 28 Grand Nat! Golf IR 524766 N 41 W20 30 505795 5094978 29 Oak Lake Camp PS 530460 N 45 W 18 14 531236 5136585 30 Anderson/Markville Presby DO 482606 N42 W 16 26 551955 5104401 31 Lyle Colton DO N/A N 44 W 18 19 524989 5124829 32 L Colton / Storebo DO 582017 N 44 W 18 19 526118 5125068 33 Dennis Coveau DO 551151 N 42 W 15 30 554314 5105040 34 Charles Drilling DO 113967 N 42 W 20 12 514409 5108767 35 Darrell Fleek DO 538199 N 43 W 20 20 507498 5116143 36 Henry Hoffman DO 123992 N45 W 18 15 530961 5136347 37 Matt Hoplin (Mettler) DO 433529 N 40 W 19 2 524543 5091633 38 Blaine Jensen DO N/A N44 W 19 34 521520 5121846 39 Jason Johnson DO 599023 N44 W 20 22 511226 5124946

PAGE 149

Table 4 Water Sam le Locations 1 4 1 ID# t Name t Type Unique # Twp Range Section UTM E* UTM N 40 King/Bennett DO 452259 N 41 W 17 13 543481 5097395 41 Marcy Moran DO 532850 N 40 W 19 11 524530 5091251 42 Ken Nelson DO 581953 N 43 W 19 29 516867 5113762 43 David Nichols DO 433481 N 41 W 18 28 529048 5094818 44 Paul Oleson DO 142945 N 43 W 19 30 516657 5113795 45 Richard Pauley DO 620416 N 43 W 18 4 529631 5120460 46 Harry Petersen DO 473738 N 44 W 18 33 529018 5121345 47 Mary Agnes Peterson DO 634574 N 43 W 19 29 518368 5114386 48 Roger Peterson DO N/A N 43 W 19 22 521420 5114853 49 A Sand DO N/A N 41 W 17 31 543072 5096505 50 Lois Saxe DO 597976 N 44 W 17 24 535809 5122643 51 Herman Velsvaag DO 567565 N 43 W 19 3 521468 5121053 52 Gregory Wickstrom DO 143949 N 42 W 20 15 511782 5108232 53 Partridge Creek Spring SP A01N 43 W 19 22 521582 5115774 54 Gushing Orange Spring SP A02 N 42 W 20 22 510332 5106182 55 Beaver Boil SP A05 N 43 W 20 14 510763 5110007 56 Ferrooxidans Rise SP A06 N 43 W 20 23 512967 5115217 57 Clear Spring SP A07 N 43 W 20 26 512859 5114231 58 Orange Mound Spring SP A08 N 43 W 20 26 512874 5114217 59 No Charge Spring SP A09 N 43 W 20 20 513149 5113838 60 Mid Point Spring SP A10 N 43 W 20 26 513120 5113368 61 Gray Beaver Spring SP All N 43 W 20 25 513031 5112212 62 Orange Boil Spring SP A12 N 42 W 20 3 510763 5110007 63 Orange Scum Spring SP A14 N 42 W 20 15 510647 5106828 64 Hi Conductivity Spring SP A15 N 42 W 20 15 511096 5107233 65 Hike Spring SP A19 N 42 W 20 15 511179 5107389 66 Filip Spring SP A20 N 41 W 16 24 544121 5097159 67 PSP-1 SP A21 N 41 W 16 24 511165 5108509 68 PSP-2 SP A22 N 42 W 20 10 511067 5109080 69 PSP-3 SP A23 N 42 W 20 10 511065 5109848 70 PSP-6 SP A24 N 42 W 20 10 533308 5088396 71 KR 708 @ Hwy 22 (+) SP A27 N 43 W 21 24 504520 5115424 72 KR 748 @ Hwy 48 (+) SP A28 N 41 W 20 23 512235 5095386 73 Banning Park 766 (+) SP A30 N 43 W 20 35 512730 5112328 74 Banning Park Large Bench SP A31 N 42 W 20 3 511298 5110645 75 Banning Overflow-Middle Ben SP A33 N 42 W 20 3 510805 5110264 76 The Big Sinkhole SW D01 N 43 W 19 31 516787 5114020 77 Beaver Sink SW D52 N 44 W 19 36 523805 5121278 78 Banning Puddle-High Bench SW N 42 W 20 3 511273 5110562

PAGE 150

Table 4 Water Sam le Locations Key to Sample Type MU = Municipal Well PS = Public Supply Well DO = Domestic Supply Well SP = Spring SW = Surface Water *All UTM locations use the 1983 North American Datum (NAD83) (+) locations are estimated -These spring numbers are a shortened version of the spring numbers defined in Table 3 ID# = Sample Identification Number for this project 1 4 2 ID# t Name t Type Unique # Twp Range Section UTM E* UTM N 79 Kettle Lift Station (+) SW N 42 W 20 15 511185 5107880 80 Kettle R 622 @ Hwy 22 (+) SW N43 W 21 24 504845 5115845 81 Kettle R 648 @ Hwy 48 (+) SW N 41 W 20 23 512327 5095346 82 KR 48 @Hwy 48 (+) SW N 41 W20 23 512226 5095441 83 KR 52 @Hwy 52 (+) SW N 45 W 20 23 512061 5134062 84 Kettle R 652 @Hwy 52 (+) SW N45 W20 '23 512109 5135158 85 Banning Peat Dome SW N 42 W 20 3 511468 5110315 86 WP 174/Below Rest Area SW N 43 W20 14 512905 5117116 87 Cane Creek 603 (+) SW N 43 W 19 6 516768 5119734 88 Stream @RD 633 (+) SW N44 W 19 31 516736 5121279

PAGE 151

Table 5 Water Chemistry Analyses Cations determined by ICP/MS Anions determined by ion chromatography Alkalinity determined by acid tritration 1 4 3 Sample ID # 1 1 1 2 2 2 3 Replicate Replicate Replicate Replicate Type MU MU MU MU MU MU MU Unique # 449381 449381 449381 449382 449382 449382 217182 Aquifer PMHN PMHN PMHN PMHN PMHN PMHN PMHN Depth (ft) 200 200 200 193 193 193 170 Date 3/29/00 3/29/00 3/29/00 3/29/00 3/29/00 3/29/00 3/29/00 Temp (€C) 8 .2 8 .2 8 .2 8 .1 8 .1 8 .1 8 .6 pH 6 .65 6 .65 6 .65 6 .45 6 .45 6 .45 6 .35 Cond (pmhos) 100 100 100 70 70 70 170 Redox (mV) -344 -344 -344 -365 -365 -365 -210 D .O (ppm) <0 .01 <0 .01 <0 .01 <0 .01 <0 .01 <0 .01 <0 .01 Cations (ppm) Ca 15 .6 15 .7 15 .3 14 .9 16 .0 15 .4 20 .5 Mg 4 .43 4 .49 4 .50 4 .32 4 .28 4 .31 7 .48 Na 2 .47 2 .43 2 .85 2 .35 2 .35 2 .38 6 .01 K 1 .49 1 .47 1 .47 1 .51 1 .51 1 .53 1 .43 Al 0 .062 0 .061 0 .152 0 .052 0 .058 0 .083 0 .011 Fe 4 .44 4 .47 4 .45 2 .94 2 .95 2 .95 7 .73 Mn 0 .107 0 .107 0 .107 0 .113 0 .114 0 .116 0 .444 Sr 0 .037 0 .037 0 .037 0 .037 0 .040 0 .038 0 .063 Ba 0 .032 0 .035 0 .041 0 .036 0 .032 0 .031 0 .035 Si 5 .1 5 .2 5 .2 5 .0 5 .0 5 .0 9 .7 Anions (ppm) Alk (as CaCO3) 44 .3 44 .3 44 .3 46 .7 46 .7 46 .7 74 Cl 1 .89 1 .93 1 .87 1 .97 1 .92 1 .90 12 .2 Br 0 .01 0 .01 0 .01 0 .01 0 .01 0 .01 0 .016 N02-N <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 N03-N <0 .005 0 .018 0 .025 0 .011 0 .009 0 .010 0 .54 S04 4 .54 4 .57 4 .56 4 .53 4 .52 4 .56 5 .03 P04-P <0 .02 <0 .02 <0 .02 <0 .02 <0 .02 <0 .02 <0 .02 total P 0 .13 0 .13 0 .13 0 .07 0 .07 0 .07 <0 .06 F <0 .05 <0 .05 <0 .05 <0 .05 <0 .05 <0 .05 0 .05 Sr/Ca 2 .37E-03 2 .36E-03 2 .42E-03 2 .48E-03 2 .50E-03 2 .47E-03 3 .07E-03 CI/Br 189 193 187 197 192 190 763 Charge Balance Cations (meq/kg) 1 .29 1 .30 1 .30 1 .24 1 .29 1 .27 1 .94 Anions (meq/kg) 1 .04 1 .04 1 .04 1 .09 1 .08 1 .09 1 .97 % difference 10 .95 11 .10 10 .97 6 .69 8 .78 7 .73 -0 .78 Ionic Strength 0 .0017 0 .0017 0 .0017 0 .0016 0 .0017 0 .0017 0 .0027

PAGE 152

Table 5 Water Chemistry Analyses 1 4 4 Sample ID # 3 Replicate 6 7 8 9 Type MU MU MU PS MU MU PS Unique # 217182 538117 562973 465946 242053 219291 416555 Aquifer PMHN PMHN PMHN PMHN PMHN PMHN PMHN Depth (ft) 170 375 397 80 725 392 150 Date 6/14/01 3/29/00 3/29/00 3/31/00 6/13/00 3/29/00 5/22/00 Temp (€C) 8 .7 7 .7 7 .9 9 .0 7 .8 7 .7 7 .6 pH 6 .30 6 .92 7 .33 7 .31 6 .38 6 .54 6 .96 Cond (mhos) 340 300 240 270 70 50 170 Redox (mV) NA -248 -282 -223 -72 -110 90 D .O (ppm) NA <0 .1 <0 .01 NA <0 .01 1 .69 <0 .01 Cations (ppm) Ca 17 .63 47 .9 41 .9 49 .2 12 .1 10 .3 24 .4 Mg 7 .34 17 .09 14 .5 9 .1 4 .37 2 .49 7 .48 Na 6 .24 6 .35 6 .55 3 .37 2 .79 1 .84 5 .37 K 1 .55 1 .46 1 .61 2 .08 0 .67 0 .62 0 .18 Al 5 .0 0 .039 0 .036 0 .024 0 .006 0 .022 0 .041 Fe 7 .08 2 .37 1 .98 13 .02 0 .40 1 .96 17 .1 Mn 0 .447 0 .220 0 .150 0 .677 0 .226 0 .006 0 .377 Sr 0 .079 0 .104 0 .105 0 .086 0 .028 0 .026 0 .035 Ba 0 .023 0 .177 0 .118 0 .155 0 .111 0 .012 0 .132 Si 9 .5 4 .4 8 .2 12 .6 7 .2 7 .7 10 .9 Anions (ppm) Alk (as CaCO3) 57 181 139 158 53 .2 27 .7 92 .4 Cl 12 .4 2 .87 11 .3 1 .00 0 .88 0 .92 0 .76 Br 0 .022 0 .019 0 .032 0 .01 0 .01 0 .01 0 .01 N02-N <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 N03-N 0 .54 <0 .005 <0 .005 0 .014 0 .072 0 .198 <0 .005 S04 4 .89 3 .99 8 .6 0 .07 0 .70 2 .39 0 .44 P04-P <0 .015 <0 .02 <0 .02 <0 .02 <0 .02 <0 .02 <0 .02 total P <0 .015 <0 .06 <0 .06 <0 .06 <0 .02 <0 .06 0 .30 F 0 .03 0 .16 0 .13 0 .09 0 .07 <0 .05 0 .07 Sr/Ca 4 .48E-03 2 .17E-03 2 .51E-03 1 .75E-03 2 .31E-03 2 .52E-03 1 .43E-03 Cl/Br 564 151 353 100 88 92 76 Charge Balance Cations (rneq/kg) 1 .80 4 .12 3 .61 3 .41 1 .10 0 .82 2 .07 Anions (meq/kg) 1 .82 3 .79 3 .28 3 .19 1 .11 0 .64 1 .88 % difference -0 .52 4 .10 4 .79 3 .24 -0 .32 11 .75 4 .85 Ionic Strength 0 .0025 0 .0051 0 .0046 0 .0046 0 .0013 0 .0010 0 .0030

PAGE 153

Table 5 Water Chemistry Analyses 1 4 5 Sample ID # 10 11 12 13 14 15 16 Type PS DO PS PS PS PS PS Unique # 574578 W03120 574577 493869 473703 219400 507417 Aquifer PMHN PMHN PMHN PMHN PMHN PMHN PMHN Depth (ft) 120 80 90 267 100 128 345 Date 5/22/00 5/22/00 5/22/00 3/31/00 3/31/00 3/27/00 3/27/00 Temp (€C) 8 .7 8 .7 8 .2 8 .1 8 .9 8 .4 7 .6 pH 6 .63 7 .15 6 .86 6 .64 7 .40 6 .32 7 .38 Cond (mhos) 115 130 220 110 800 360 170 Redox (mV) 210 193 125 -362 -88 72 -76 D .O (ppm) 1 .10 5 .00 1 .60 NA 9 .10 8 .0 NA Cations (ppm) Ca 19 .5 28 .6 46 .7 9 .7 91 .0 15 .8 28 .2 Mg 5 .70 9 .49 11 .1 2 .36 36 .6 4 .83 7 .38 Na 5 .84 6 .43 8 .35 1 .72 52 .3 61 .8 3 .00 K 0 .59 0 .58 0 .75 1 .10 1 .28 1 .40 1 .40 Al 0 .023 0 .033 0 .113 0 .035 0 .022 0 .066 0 .036 Fe 5 .70 0 .41 11 .2 19 .0 <0 .15 0 .33 3 .42 Mn 0 .600 0 .024 0 .198 0 .433 0 .005 0 .017 0 .023 Sr 0 .040 0 .034 0 .040 0 .027 0 .106 0 .067 0 .067 Ba 0 .115 0 .082 0 .086 0 .044 0 .053 0 .071 0 .044 Si 7 .1 8 .9 8 .8 6 .7 12 .5 4 .5 9 .2 Anions (ppm) Alk (as CaCO3) 68 .6 80 .3 101 47 .7 199 17 .2 94 Cl 1 .79 9 .3 3 .50 1 .45 193 117 1 .23 Br 0 .016 0 .01 0 .017 0 .01 0 .022 0 .01 0 .01 N02-N <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 N03-N 0 .008 0 .259 0 .197 0 .008 0 .65 0 .55 0 .010 S04 1 .17 8 .2 11 .8 3 .26 10 .2 8 .2 0 .05 P04-P <0 .02 0 .02 <0 .02 <0 .02 0 .07 <0 .02 <0 .02 total P 0 .20 0 .04 0 .20 <0 .06 0 .13 <0 .06 0 .16 F 0 .12 0 .14 0 .11 0 .01 0 .13 <0 .05 0 .13 Sr/Ca 2 .05E-03 1 .19E-03 8 .57E-04 2 .78E-03 1 .16E-03 4 .24E-03 2 .38E-03 Cl/Br 112 930 206 145 8,773 11,700 123 Charge Balance Cations (meq/kg) 1 .71 2 .50 3 .63 0 .78 9 .86 3 .91 2 .18 Anions (meq/kg) 1 .45 2 .06 2 .39 1 .06 9 .69 3 .86 1 .92 % difference 8 .22 9 .62 20 .64 -15 .26 0 .90 0 .73 6 .32 Ionic Strength 0 .0022 0 .0031 0 .0046 0 .0017 0 .0127 0 .0045 0 .0027

PAGE 154

Table 5 Water Chemistry Analyses 1 4 6 Sample ID # 16 Replicate 16 Replicate 16 17 18 18 19 Type PS PS PS PS PS PS PS Unique # 507417 507417 507417 142934 575167 575167 436792 Aquifer PMHN PMHN PMHN PMHN QBAA QBAA PMVU Depth (ft) 345 345 345 110 200 200 Date 3/27/00 3/27/00 6/14/01 3/31/00 3/29/00 6/12/01 6/12/01 Temp (€C) 7 .6 7 .6 8 .0 9 .2 7 .9 9 .7 7 .9 pH 7 .38 7 .38 7 .61 7 .80 7 .92 8 .06 8 .51 Cond (mhos) 170 170 280 240 300 240 300 Redox (mV) -76 -76 NA -170 -264 NA NA D .O (ppm) NA NA NA <0 .01 <0 .01 0 .20 2 .00 Cations (ppm) Ca 28 .8 28 .5 22 .8 46 .2 46 .2 37 .1 11 .26 Mg 7 .34 7 .32 7 .22 9 .07 14 .2 13 .6 5 .08 Na 2 .94 2 .92 3 .23 4 .81 4 .88 4 .54 69 .7 K 1 .34 1 .28 1 .26 2 .58 1 .10 0 .80 2 .83 Al 0 .036 0 .144 0 .020 0 .035 0 .033 0 .003 0 .006 Fe 3 .41 3 .44 3 .71 1 .40 1 .13 2 .36 <0 .005 Mn 0 .024 0 .023 0 .024 1 .35 0 .061 0 .068 0 .013 Sr 0 .068 0 .068 0 .089 0 .110 0 .074 0 .068 0 .154 Ba 0 .033 0 .064 0 .028 0 .108 0 .063 0 .043 0 .052 Si 9 .1 9 .0 9 .0 7 .6 9 .6 9 .1 5 .9 Anions (ppm) Alk (as CaCO3) 94 94 92 149 159 153 152 Cl 1 .09 0 .96 1 .05 1 .80 1 .12 0 .81 32 .8 Br 0 .016 0 .01 0 .01 0 .01 0 .01 0 .01 0 .108 N02-N <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 N03-N 0 .012 0 .005 0 .049 0 .040 <0 .005 0 .006 <0 .005 S04 0 .09 0 .03 0 .12 0 .11 6 .68 5 .97 7 .86 P04-P <0 .02 <0 .02 <0 .015 <0 .02 <0 .02 <0 .015 <0 .015 total P 0 .16 0 .17 0 .16 <0 .06 0 .06 <0 .015 <0 .015 F 0 .13 0 .12 0 .11 0 .11 0 .11 0 .10 0 .68 Sr/Ca 2 .36E-03 2 .39E-03 3 .90E-03 2 .38E-03 1 .60E-03 1 .83E-03 1 .37E-02 Cl/Br 68 96 105 180 112 81 304 Charge Balance Cations (meq/kg) 2 .21 2 .19 1 .91 3 .33 3 .72 3 .19 4 .09 Anions (meq/kg) 1 .92 1 .92 1 .88 3 .04 3 .36 3 .21 4 .17 % difference 6 .91 6 .54 0 .69 4 .55 5 .11 -0 .33 -0 .95 Ionic Strength 0 .0027 0 .0027 0 .0024 0 .0040 0 .0046 0 .0041 0 .0039

PAGE 155

Table 5 Water Chemistry Analyses 1 4 7 Sample ID # 20 21 22 23 24 25 26 Type PS PS PS PS PS PS PS Unique # W30064 467676 172459 172468 Aquifer PMVU PMVU QUUU QUUU PMVU QBAA PMVU Depth (ft) 400 43 75 73 575 Date 6/12/01 6/22/00 3/29/00 6/22/00 6/18/01 6/18/01 6/18/01 Temp (€C) 8 .8 15 .2 8 .8 9 .7 15 .7 11 .2 15 .8 pH 8 .63 7 .67 7 .73 6 .12 9 .91 10 .05 9 .89 Cond (mhos) 290 320 480 210 312 290 300 Redox (mV) NA -105 -240 -36 -25 -14 12 D .O (ppm) NA 4 .86 0 .19 8 .40 0 .82 0 .29 0 .60 Cations (ppm) Ca 16 .18 47 .0 87 .1 31 .9 3 .05 2 .41 3 .65 Mg 4 .87 16 .2 31 .2 16 .5 0 .65 0 .38 0 .69 Na 49 .8 16 .9 9 .04 5 .52 81 .4 89 .0 76 .1 K 2 .44 2 .34 1 .83 1 .04 0 .45 0 .30 0 .42 Al 0 .005 0 .003 0 .032 0 .008 0 .005 0 .016 0 .003 Fe 0 .63 0 .02 0 .11 0 .02 <0 .015 <0 .015 <0 .015 Mn 0 .025 0 .064 1 .200 0 .200 0 .001 0 .001 0 .066 Sr 0 .126 0 .291 0 .209 0 .052 0 .023 0 .031 0 .031 Ba 0 .030 0 .479 0 .082 0 .101 0 .047 0 .046 0 .051 Si 4 .42 10 .4 10 .9 12 .1 10 .6 11 .3 10 .0 Anions (ppm) Alk (as CaCO3) 84 185 335 135 106 109 104 Cl 49 .5 14 .1 1 .05 2 .47 33 .8 36 .8 31 .5 Br 0 .252 0 .061 0 .01 0 .01 0 .113 0 .131 0 .108 N02-N <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 N03-N 0 .012 0 .146 0 .007 0 .124 <0 .005 <0 .005 0 .083 S04 14 .1 3 .44 0 .27 9 .3 22 .7 24 .0 21 .7 P04-P <0 .015 0 .02 0 .02 0 .03 <0 .015 <0 .015 <0 .015 total P <0 .015 <0 .06 0 .06 <0 .06 <0 .015 <0 .015 <0 .015 F 0 .34 0 .23 0 .13 0 .08 0 .44 0 .45 0 .43 Sr/Ca 7 .79E-03 6 .19E-03 2 .40E-03 1 .63E-03 7 .54E-03 1 .29E-02 8 .49E-03 Cl/Br 196 231 105 247 299 281 292 Charge Balance Cations (meq/kg) 3 .44 4 .49 7 .36 3 .22 3 .76 4 .03 3 .56 Anions (meq/kg) 3 .39 4 .19 6 .74 2 .97 3 .65 3 .85 3 .53 % difference 0 .66 3 .42 4 .41 3 .96 1 .47 2 .26 0 .50 Ionic Strength 0 .0038 0 .0053 0 .0088 0 .0040 0 .0035 0 .0037 0 .0033

PAGE 156

Table 5 Water Chemistry Analyses 1 4 8 Sample ID # 27 28 29 30 31 32 33 Type PS IR PS DO DO DO DO Unique # 467698 524766 530460 482606 582017 551151 Aquifer PMVU PMHN QBAA QBAA PMHN PMHN QWTA Depth (ft) 300 417 46 58 50 110 70 Date 6/18/01 6/12/01 6/19/01 5/30/01 3/27/00 3/27/00 6/18/01 Temp (€C) 17 .1 16 .1 7 .7 9 .1 8 .8 7 .9 9 .0 pH 9 .95 7 .73 .7 .88 7 .94 7 .70 7 .44 8 .36 Cond (mhos) 332 250 231 239 390 160 150 Redox (mV) -42 NA -205 -208 -70 -149 -165 D .O (ppm) 0 .74 NA <0 .01 <0 .01 1 .83 <0 .01 0 .22 Cations (ppm) Ca 3 .13 32 .7 35 .9 28 .1 62 .1 27 .7 21 .23 Mg 0 .61 13 .44 11 .2 9 .80 17 .33 7 .46 7 .08 Na 80 .9 5 .69 5 .03 4 .11 14 .2 2 .67 4 .24 K 0 .38 1 .48 1 .50 1 .12 1 .39 0 .61 0 .56 Al 0 .005 0 .009 <0 .001 0 .028 0 .095 0 .024 0 .001 Fe <0 .015 0 .77 2 .04 0 .56 9 .1 11 .16 <0 .015 Mn 0 .001 0 .012 0 .308 0 .087 0 .779 0 .123 0 .044 Sr 0 .027 0 .090 0 .141 0 .145 0 .079 0 .059 0 .072 B a 0 .020 0 .090 0 .281 0 .029 0 .189 0 .029 0 .064 Si 10 .3 4 .8 11 .6 12 .7 6 .9 9 .3 7 .7 Anions (ppm) Alk (as CaCO3) 109 118 185 114 193 112 104 Cl 33 .8 11 .3 0 .59 6 .38 11 .7 0 .47 1 .24 Br 0 .112 0 .026 0 .01 0 .01 0 .030 0 .01 0 .01 N02-N <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 N03-N 0 .08 0 .075 0 .006 0 .005 0 .840 <0 .005 0 .017 S04 22 .8 11 .1 0 .03 7 .25 23 .1 0 .15 6 .51 P04-P <0 .015 <0 .015 <0 .015 <0 .015 <0 .02 <0 .02 <0 .015 total P <0 .015 <0 .015 <0 .015 0 .04 0 .09 0 .34 <0 .015 F 0 .43 0 .10 0 .11 0 .08 0 .12 0 .10 0 .05 Sr/Ca 8 .63E-03 2 .75E-03 3 .93E-03 5 .16E-03 1 .27E-03 2 .13E-03 3 .39E-03 Cl/Br 302 435 59 638 390 47 124 Charge Balance Cations (meq/kg) 3 .74 3 .03 2 .98 2 .42 5 .18 2 .13 1 .84 Anions (meq/kg) 3 .72 2 .92 3 .72 2 .62 4 .74 2 .26 2 .25 % difference 0 .15 1 .80 -11 .10 -3 .89 4 .48 -2 .99 -10 .04 Ionic Strength 0 .0035 0 .0039 0 .0038 0 .0031 0 .0068 0 .0030 0 .0024

PAGE 157

Table 5 Water Chemistry Analyses 1 4 9 Sample ID # 34 35 36 37 38 39 40 Type DO DO DO DO DO DO DO Unique # 113967 538199 123992 433529 599023 452259 Aquifer QBAA PMHN PMHN PMCV PMHN PMFL PMVU Depth' (ft) 167 80 115 205 100 105 285 Date 3/28/00 3/31/00 6/19/01 6/19/01 3/27/00 3/28/00 3/29/00 Temp (€C) 7 .5 7 .6 8 .6 9 .3 7 .8 8 .3 8 .6 pH 7 .77 7 .01 7 .94 8 .22 6 .76 7 .68 8 .36 Cond(mhos) 380 340 245 188 140 270 400 Redox (mV) -200 -224 -210 -202 103 -170 -293 D .O (ppm) <0 .01 NA <0 .01 0 .03 2 .9 <0 .01 <0 .01 Cations (ppm) Ca 53 .4 48 .4 37 .7 21 .9 18 .9 50 .0 28 .7 Mg 21 .2 15 .88 12 .74 10 .62 6 .86 12 .26 11 .1 Na 9 .97 12 .39 5 .09 11 .1 4 .14 3 .24 61 .6 K 1 .82 0 .97 1 .51 2 .46 0 .69 1 .20 3 .53 Al 0 .025 0 .022 0 .002 <0 .001 0 .021 0 .027 0 .023 Fe 1 .01 7 .94 1 .52 0 .15 <0 .15 4 .04 0 .34 Mn 0 .221 1 .036 0 .217 0 .053 0 .003 0 .172 0 .025 Sr 0 .119 0 .076 0 .173 0 .183 0 .038 0 .152 0 .266 Ba 0 .074 0 .075 0 .234 0 .066 0 .039 0 .098 0 .041 Si 10 .8 13 .5 11 .0 5 .8 10 .4 10 .5 6 .1 Anions (ppm) Alk (as CaCO3) 208 171 195 140 58 167 108 Cl 0 .91 18 .3 0 .52 0 .58 5 .47 0 .64 88 .5 Br 0 .01 0 .022 0 .01 0 .01 0 .01 0 .01 0 .47 N02-N <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 N03-N <0 .005 0 .018 0 .019 <0 .005 1 .09 <0 .005 <0 .005 S04 11 .3 5 .06 0 .08 1 .97 7 .9 <0 .02 14 .1 P04-P <0 .02 <0 .02 <0 .015 <0 .015 <0 .02 <0 .02 0 .02 total P <0 .06 0 .11 <0 .015 <0 .015 <0 .06 <0 .06 <0 .06 F 0 .26 0 .19 0 .11 0 .16 0 .05 0 .23 0 .32 Sr/Ca 2 .23E-03 1 .57E-03 4 .59E-03 8 .36E-03 2 .01E-03 3 .04E-03 9 .27E-03 Cl/Br 91 832 52 58 547 64 188 Charge Balance Cations (meq/kg) 4 .89 4 .29 3 .20 2 .52 1 .71 3 .68 5 .12 Anions (meq/kg) 4 .43 4 .05 3 .92 2 .87 1 .56 3 .37 4 .97 difference 4 .94 2 .85 -10 .17 -6 .42 4 .53 4 .41 1 .47 Ionic Strength 0 .0060 0 .0055 0 .0041 0 .0030 0 .0022 0 .0046 0 .0058

PAGE 158

Table 5 Water Chemistry Analyses 1 5 0 Sample ID # 40 41 42 43 44 45 46 Type DO DO DO DO DO DO DO Unique # 452259 532850 581953 433481 142945 620416 473738 Aquifer PMVU PMCV PMHN QWTA PMHN PMVU PMFL Depth (ft) 285 317 90 64 95 165 85 Date 6/13/00 6/19/01 3/29/00 6/18/01 3/28/00 5/24/01 5/31/01 Temp (€C) 9 .0 8 .8 7 .9 8 .6 7 .9 7 .6 7 .6 pH 8 .36 9 .21 6 .34 7 .67 6 .57 6 .80 7 .26 Cond (mhos) 390 311 170 200 210 212 206 Redox (mV) -170 -178 -399 -139 -237 143 -158 D .O (ppm) <0 .01 0 .02 <0 .01 2 .58 <0 .01 2 .46 <0 .01 Cations (ppm) Ca 29 .1 2 .77 20 .7 37 .1 28 .0 15 .6 23 .5 Mg 12 .8 1 .17 6 .91 7 .46 9 .23 7 .83 8 .14 Na 59 .1 94 .3 11 .01 3 .1 9 .14 16 .3 2 .70 K 3 .84 2 .13 2 .70 0 .63 1 .82 0 .63 0 .87 Al 0 .121 0 .004 0 .048 <0 .001 0 .027 0 .005 0 .006 Fe 0 .27 <0 .015 3 .80 1 .46 5 .61 <0 .015 7 .07 Mn a 0 .025 0 .009 0 .478 0 .363 0 .148 0 .004 0 .081 Sr 0 .276 0 .056 0 .052 0 .058 0 .062 0 .061 0 .041 Ba 0 .140 0 .038 0 .043 0 .108 0 .038 0 .007 0 .056 Si 5 .8 4 .3 6 .1 15 .3 7 .6 11 .5 12 .8 Anions (ppm) Alk (as CaCO3) 115 138 68 164 84 83 114 Cl 84 44 .9 17 .4 0 .61 16 .8 14 .2 0 .49 Br 0 .47 0 .161 0 .01 0 .01 0 .023 0 .081 0 .01 N02-N <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 N03-N <0 .005 0 .033 <0 .005 0 .01 <0 .005 0 .862 0 .004 S04 13 .5 16 .2 5 .38 1 .23 6 .69 6 .30 0 .16 P04-P 0 .02 <0 .015 <0 .02 <0 .015 <0 .02 <0 .015 <0 .015 total P <0 .06 <0 .015 0 .12 <0 .015 0 .08 <0 .015 0 .04 F 0 .36 0 .40 0 .04 0 .04 0 .04 0 .07 0 .07 Sr/Ca 9 .48E-03 2 .02E-02 2 .51E-03 1 .56E-03 2 .21E-03 3 .91E-03 1 .74E-03 Cl/Br 179 279 1,740 61 730 175 49 Charge Balance Cations (meq/kg) 5 .18 4 .39 2 .15 2 .62 2 .60 2 .15 1 .98 Anions (meq/kg) 4 .98 4 .40 1 .97 3 .32 2 .30 2 .26 2 .30 % difference 1 .99 -0 .12 4 .52 -11 .87 6 .29 -2 .43 -7 .37 Ionic Strength 0 .0059 0 .0040 0 .0027 0 .0034 0 .0034 0 .0026 0 .0027

PAGE 159

Table 5 Water Chemistry Analyses 1 5 1 Sample ID # 47 48 49 50 51 52 53 Type DO DO DO DO DO DO Spring Unique # 634574 A Sand 597976 567565 143949 Al Aquifer PMHN PMVU PMCV PMHN PMHN Spring Depth (ft) 65 130 165 155 150 Date 3/28/00 7/11/97 6/22/00 5/22/01 3/27/00 3/28/00 7/11/97 Temp (€C) 8 .6 8 .0 8 .4 8 .2 7 .8 7 .9 7 .5 pH 6 .74 6 .83 7 .45 7 .62 6 .85 7 .08 6 .60 Cond (mhos) 140 180 220 298 150 280 60 Redox (mV) -36 -89 -105 -269 28 -40 -24 D .O (ppm) 6 .42 0 .4 5 .85 <0 .01 5 .50 7 .0 7 .80 Cations (ppm) Ca 22 .7 40 .7 46 .6 35 .3 24 .3 43 .2 9 .60 Mg 6 .45 12 .1 15 .3 11 .8 7 .71 13 .8 3 .67 Na 3 .36 3 .18 7 .76 4 .54 2 .79 5 .85 2 .91 K 1 .32 0 .91 0 .43 1 .37 0 .72 0 .84 1 .13 Al 0 .146 0 .002 0 .007 0 .005 0 .025 0 .030 0 .001 Fe <0 .15 0 .26 0 .96 3 .47 0 .27 <0 .15 <0 .01 Mn 0 .004 1 .87 0 .113 0 .298 0 .008 0 .004 0 .004 Sr 0 .048 0 .067 0 .064 0 .139 0 .039 0 .073 0 .029 Ba 0 .028 0 .123 0 .717 0 .052 0 .031 0 .039 0 .018 Si 8 .5 15 .6 11 .8 17 .3 11 .8 13 .3 9 .9 Anions (ppm) Alk (as CaC03) 60 156 168 142 81 115 29 .8 Cl 4 .79 0 .86 1 .00 4 .19 0 .96 19 .6 3 .38 Br 0 .01 <0 .015 0 .01 0 .016 0 .01 0 .023 <0 .015 N02-N <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 N03-N 1 .60 0 .01 0 .275 <0 .005 0 .233 1 .49 0 .90 S04 6 .21 3 .26 6 .10 16 .4 4 .91 11 .2 5 .44 P04-P <0 .02 <0 .02 <0 .02 <0 .015 <0 .02 <0 .02 <0 .02 total P <0 .06 <0 .02 <0 .06 <0 .015 <0 .06 <0 .06 <0 .02 F 0 .03 0 .12 0 .17 0 .13 0 .05 0 .12 0 .05 Sr/Ca 2 .11E-03 1 .65E-03 1 .37E-03 3 .94E-03 1 .60E-03 1 .69E-03 3 .02E-03 Cl/Br 479 >57 100 262 96 852 >225 Charge Balance Cations (meq/kg) 1 .85 3 .19 3 .94 2 .97 1 .99 3 .57 0 .94 Anions (meq/kg) 1 .58 3 .22 3 .54 3 .30 1 .77 3 .20 0 .87 % difference 7 .59 5 .39 -5 .35 5 .86 5 .49 3 .69 Ionic Strength 0 .0023 0 .0040 0 .0048 0 .0041 0 .0025 0 .0046 0 .0012

PAGE 160

Table 5 Water Chemistry Analyses 1 5 2 Sample ID # 54 55 55 56 56 57 58 Type Spring Spring Spring Spring Spring Spring Spring Unique # A2 A5 A5 A6 A6 A7 A8 Aquifer Depth (ft) Spring Spring Spring Spring Spring Spring Spring Date 6/14/01 5/17/00 6/13/00 5/26/00 6/14/01 5/22/00 5/22/00 Temp (€C) 7 .0 8 .8 8 .3 7 .7 7 .1 7 .7 9 .6 pH 6 .29 6 .95 6 .91 7 .07 6 .98 6 .85 7 .52 Cond (mhos) NA 750 560 140 NA 120 220 Redox (mV) NA -56 -114 -100 NA -50 -148 D .O (ppm) NA NA NA <0 .01 NA NA NA Cations (ppm) Ca 16 .52 57 .0 52 .8 22 .4 19 .11 15 .2 20 .6 Mg 6 .49 17 .0 17 .4 8 .2 7 .56 5 .97 7 .73 Na 2 .25 68 .5 70 .2 4 .47 4 .11 14 .5 15 .0 K 0 .71 3 .90 3 .18 0 .98 0 .89 0 .56 1 .33 Al 0 .005 0 .026 0 .005 0 .093 0 .005 <0 .005 0 .003 Fe 11 .64 5 .21 5 .33 15 .9 16 .7 <0 .02 6 .54 Mn 0 .607 0 .374 0 .376 0 .449 0 .437 <0 .001 0 .283 Sr 0 .037 0 .769 0 .752 0 .055 0 .05 0 .037 0 .186 Ba 0 .022 0 .155 0 .177 0 .206 0 .088 0 .085 0 .134 Si 9 .5 7 .5 7 .1 9 .4 9 .6 7 .3 7 .1 Anions (ppm) Alk (as CaCO3) 74 113 109 99 .6 85 40 .5 92 .7 Cl 0 .43 170 168 4 .24 2 .30 28 .1 31 .3 Br 0 .01 1 .11 1 .21 0 .017 0 .01 0 .022 0 .218 N02-N <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 N03-N <0 .005 0 .045 <0 .005 0 .011 0 .026 0 .288 0 .012 S04 0 .52 7 .03 6 .97 1 .0 1 .04 6 .04 0 .24 P04-P <0 .015 <0 .02 <0 .02 <0 .02 <0 .015 <0 .02 <0 .02 total P <0 .015 0 .05 <0 .06 0 .12 0 .09 <0 .02 <0 .02 F 0 .05 0 .24 0 .11 0 .17 0 .11 0 .08 0 .09 Sr/Ca 2 .24E-03 1 .35E-02 1 .42E-02 2 .46E-03 2 .62E-03 2 .43E-03 9 .03E-03 Cl/Br 43 153 139 249 230 1,277 144 Charge Balance Cations (meq/kg) 1 .48 7 .34 7 .22 2 .02 1 .78 1 .90 2 .36 Anions (meq/kg) 1 .50 7 .23 7 .08 2 .14 1 .79 1 .75 2 .75 % difference -0 .96 0 .77 0 .97 -3 .07 -0 .37 3 .95 -7 .68 Ionic Strength 0 .0022 0 .0091 0 .0089 0 .0031 0 .0028 0 .0023 0 .0032

PAGE 161

Table 5 Water Chemistry Analyses 1 5 3 Sample ID # 59 60 61 62 63 64 65 Type Spring Spring Spring Spring Spring Spring Spring Unique # A9 AlO All A12 A14 A15 A19 Aquifer Depth (ft) Spring Spring Spring Spring Spring Spring Spring Date 6/2/00 6/2/00 6/2/00 11/20/99 8/28/01 8/28/01 8/28/01 Temp (€C) 7 .4 7 .5 7 .5 7 .8 18 .1 15 .8 9 .4 pH 6 .07 6 .25 6 .62 6 .41 6 .85 7 .35 7 .11 Cond (pmhos) 8 8 17 150 160 70 200 Redox (mV) 75 -46 -104 NA NA NA NA D .O (ppm) 4 .70 NA NA NA NA 5 .1 8 .7 Cations (ppm) Ca 11 .3 12 .7 38 .7 20 .9 12 .7 38 .81 25 .97 Mg 4 .70 5 .2 13 .84 6 .08 4 .71 13 .12 9 .36 Na 2 .91 3 .70 4 .3 6 .60 4 .40 5 .49 4 .55 K 0 .75 1 .08 0 .88 0 .52 0 .30 1 .46 0 .68 Al 0 .006 0 .004 0 .006 0 .059 0 .013 0 .654 0 .049 Fe <0 .02 <0 .02 <0 .02 11 .8 9 .36 2 .49 0 .08 Mn <0 .001 0 .001 0 .010 0 .496 0 .396 0 .075 0 .009 Sr 0 .025 0 .032 0 .060 0 .055 0 .046 0 .074 0 .052 B a 0 .152 0 .099 0 .135 0 .684 0 .027 0 .053 0 .038 Si 6 .9 6 .1 9 .4 10 .6 10 .0 12 .1 8 .64 Anions (ppm) Alk (as CaCO3) 42 .9 46 .4 137 .0 74 .5 51 160 105 Cl 2 .21 2 .99 4 .86 1 .36 6 .79 0 .14 0 .89 Br 0 .01 0 .01 0 .01 0 .01 0 .03 0 .02 <0 .015 N02-N <0 .005 <0 .005 0 .006 <0 .005 <0 .005 <0 .005 <0 .005 N03-N 0 .651 1 .144 0 .463 <0 .005 <0 .005 0 .099 0 .271 S04 3 .7 6 .5 7 .6 0 .62 0 .23 2 .23 5 .45 P04-P <0 .02 0 .02 <0 .02 <0 .02 <0 .02 <0 .02 <0 .02 total P <0 .02 <0 .02 <0 .02 0 .05 0 .22 0 .10 0 .03 F 0 .02 0 .04 0 .08 0 .11 0 .02 0 .14 0 .12 Sr/Ca 2 .21E-03 2 .52E-03 1 .55E-03 2 .63E-03 3 .62E-03 1 .91E-03 2 .00E-03 Cl/Br 221 299 486 136 238 7 >59 Charge Balance Cations (meq/kg) 1 .10 1 .25 3 .28 1 .86 1 .22 3 .29 2 .28 Anions (meq/kg) 1 .04 1 .23 3 .07 1 .55 1 .22 3 .29 2 .26 % difference 2 .59 0 .89 3 .34 9 .03 0 .18 0 .12 0 .41 Ionic Strength 0 .0014 0 .0016 0 .0041 0 .0025 0 .0018 0 .0041 0 .0028

PAGE 162

Table 5 Water Chemistry Analyses 1 5 4 Sample ID # 66 67 68 69 70 71 72 Type Spring Spring Spring Spring Spring Spring Spring Unique # A20 A21 A22 A23 A24 A27 A28 Aquifer Depth (ft) QBasalt Spring Spring Spring QBasalt Spring QBasalt Date 3/29/00 3/14/01 3/14/01 3/14/01 3/15/01 11/11/99 11/11/99 Temp (€C) 7 .3 7 .2 7 .2 7 .8 '8 .7 3 .4 5 .9 pH 8 .00 7 .00 6 .49 6 .93 6 .87 7 .18 6 .93 Cond (umbos) 160 255 143 194 70 550 530 Redox (mV) -143 -20 286 N/A N/A 240 129 D .O (ppm) NA 6 .00 8 .20 <0 .01 9 .30 11 .0 3 .7 Cations (ppm) Ca 38 .6 30 .5 14 .3 15 .1 7 .4 67 .9 29 .6 Mg 10 .38 10 .97 5 .67 5 .72 2 .42 28 .4 27 .6 Na 3 .03 3 .7 3 .1 3 .1 1 .8 8 .67 12 .6 K 0 .77 1 .21 1 .14 1 .02 0 .60 1 .38 1 .24 Al 0 .065 0 .002 0 .005 0 .001 0 .001 0 .028 0 .022 Fe <0 .15 <0 .015 <0 .015 20 .12 0 .02 1 .34 4 .94 Mn 0 .007 0 .001 0 .002 0 .001 0 .001 0 .405 1 .79 Sr 0 .053 0 .054 0 .035 0 .041 0 .020 0 .224 0 .075 Ba 0 .019 0 .026 0 .022 0 .067 0 .005 0 .098 0 .057 Si 8 .7 9 .0 7 .7 10 .8 7 .3 13 .5 8 .5 Anions (ppm) Alk (as CaCO3) 113 125 60 .3 104 30 .4 258 121 Cl 0 .70 3 .0 3 .5 1 .3 0 .9 1 .59 37 .7 Br 0 .01 0 .016 0 .023 <0 .015 <0 .015 0 .01 0 .035 N02-N 0 .008 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 N03-N 0 .084 1 .541 0 .393 <0 .005 <0 .005 <0 .005 <0 .005 S04 7 .31 7 .08 10 .52 0 .47 5 .99 0 .59 6 .71 P04-P <0 .02 0 .026 0 .038 <0 .015 <0 .015 0 .02 <0 .02 total P <0 .06 0 .024 0 .016 0 .171 <0 .015 0 .13 0 .09 F <0 .05 0 .11 0 .04 0 .05 0 .02 0 .25 0 .11 Sr/Ca 1 .37E-03 1 .77E-03 2 .44E-03 2 .72E-03 2 .71E-03 3 .30E-03 2 .53E-03 Cl/Br 70 190 155 >87 >60 159 1,077 Charge Balance Cations (meq/kg) 2 .93 2 .62 1 .35 1 .38 0 .66 6 .14 4 .33 Anions (meq/kg) 2 .44 2 .85 1 .55 2 .13 0 .76 5 .23 3 .63 % difference 9 .18 -4 .15 -7 .08 -21 .21 -6 .99 8 .07 8 .83 Ionic Strength 0 .0036 0 .0034 0 .0018 0 .0026 0 .0009 0 .0073 0 .0055,

PAGE 163

Table 5 Water Chemistry Analyses 1 5 5 Sample ID # 73 74 75 76 77 77 Replicate 77 Replicate Type Spring Spring Spring SW SW SW SW Unique # A30 A31 A33 Aquifer Spring Spring Spring Surface Surface Surface Surface Depth (ft) surface mixed w\ water Date 11/11/99 11/20/99 11/20/99 6/22/00 3/27/00 3/27/00 3/27/00 Temp (€C) 7 .5 6 .2 4 .8 16 .9 2 .4 2 .4 2 .4 pH 5 .94 5 .84 6 .68 7 .82 5 .30 5 .30 5 .30 Cond (mhos) 42 110 230 400 30 30 30 Redox (mV) 148 NA NA -57 103 103 103 D .O (ppm) 3 .8 NA NA 10 .4 11 .8 11 .8 11 .8 Cations (ppm) Ca 4 .3 13 .7 26 .9 23 .7 6 .41 6 .82 6 .87 Mg 2 .14 3 .19 8 .17 5 .09 1 .54 1 .50 1 .51 Na 1 .34 12 .6 17 .9 61 .8 1 .11 0 .81 0 .78 K 0 .82 0 .46 1 .16 8 .71 0 .80 0 .74 0 .82 Al 1 .28 0 .024 0 .027 0 .046 0 .302 0 .284 0 .299 Fe 19 .7 0 .17 0 .03 1 .12 0 .16 0 .13 0 .69 Mn 0 .191 0 .284 0 .386 0 .087 0 .051 0 .050 0 .057 Sr 0 .019 0 .100 0 .155 0 .045 0 .021 0 .022 0 .022 Ba 0 .079 0 .724 0 .871 0 .518 0 .028 0 .027 0 .040 Si 6 .4 6 .7 7 .5 4 .6 4 .0 4 .1 4 .0 Anions (ppm) Alk (as CaCO3) 18 17 .5 65 .2 108 1 .7 1 .7 1 .7 Cl 1 .21 22 .5 29 .4 83 0 .66 0 .64 0 .78 Br 0 .01 0 .151 0 .201 0 .040 0 .01 0 .01 0 .01 N02-N <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 <0 .005 N03-N 0 .014 <0 .005 0 .011 0 .047 1 .49 1 .48 1 .50 S04 4 .32 4 .23 2 .68 4 .57 5 .58 5 .49 5 .69 P04-P <0 .02 <0 .02 <0 .02 0 .89 <0 .02 <0 .02 <0 .02 total P 0 .28 0 .02 <0 .02 1 .42 <0 .06 <0 .06 <0 .06 F <0 .05 0 .07 0 .12 0 .56 <0 .05 <0 .05 <0 .05 Sr/Ca 4 .39E-03 7 .30E-03 5 .76E-03 1 .90E-03 3 .28E-03 3 .23E-03 3 .20E-03 Cl/Br 121 149 146 2,075 66 64 78 Charge Balance Cations (meq/kg) 0 .47 1 .52 2 .84 4 .52 0 .52 0 .52 0 .53 Anions (meq/kg) 0 .53 1 .08 2 .20 4 .63 0 .29 0 .28 0 .29 % difference -5 .78 16 .98 12 .72 -1 .19 29 .07 29 .91 28 .67 Ionic Strength 0 .0013 0 .0017 0 .0032 0 .0049 0 .0007 0 .0007 0 .0007

PAGE 164

Table 5 Water Chemistry Analyses 1 5 6 Sample ID # 78 79 80 81 82 83 84 Type Unique # SW SW SW SW SW SW SW Aquifer Depth (ft) Surface Surface River River Surface Surface River Date 11/20/99 5/22/00 11/11/99 11/11/99 11/11/99 11/11/99 11/11/99 Temp ( € C) 4 .5 NA 4 .0 5 .0 5 .1 5 .2 4 .3 pH 5 .26 8 .96 6 .90 7 .50 6 .52 5 .16 6 .72 Cond (mhos) 30 4200 50 130 260 220 107 Redox (mV) NA NA 225 140 152 -22 197 D .O (ppm) NA NA 7 .6 8 .2 5 .2 2 .5 10 .0 Cations (ppm) Ca 8 .1 19 .8 6 .5 15 .1 23 .5 15 .9 13 .1 Mg 0 .87 3 .05 2 .86 5 .59 9 .03 6 .95 4 .98 Na 5 .20 270 1 .60 4 .07 9 .9 2 .31 2 .89 K 0 .58 151 0 .87 1 .11 0 .54 2 .77 0 .80 Al 0 .711 0 .066 0 .173 0 .059 0 .035 0 .149 0 .062 Fe 1 .55 0 .20 1 .33 1 .22 10 .8 1 .41 0 .95 Mn '0 .171 0 .013 0 .016 0 .077 2 .09 0 .936 0 .027 Sr 0 .032 0 .020 0 .038 0 .068 0 .036 0 .034 Ba 0 .707 0 .014 0 .035 0 .089 0 .035 0 .027 Si' 5 .5 9 .2 4 .5 5 .2 13 .4 12 .5 5 .2 Anions (ppm) Alk (as CaCO3) 3 .86 90 18 50 57 28 42 Cl 1 .79 431 1 .43 4 .87 25 .2 2 .97 3 .09 Br 0 .01 <0 .015 0 .01 0 .01 0 .045 0 .01 0 .01 N02-N <0 .005 <0 .005 <0 .005 0 .012 <0 .005 <0 .005 <0 .005 N03-N 0 .024 0 .019 0 .116 0 .109 <0 .005 <0 .005 0 .046 S04 5 .15 74 .3 0 .78 2 .47 11 .1 4 .09 2 .23 P04-P <0 .02 19 .5 <0 .02 0 .02 <0 .02 0 .06 0 .06 total P 0 .16 27 .4 <0 .02 0 .05 0 .17 0 .12 0 .03 F 0 .04 24 .5 <0 .05 0 .05 0 .07 0 .05 0 .04 Sr/Ca 3 .95E-03 3 .10E-03 2 .52E-03 2 .89E-03 2 .26E-03 2 .60E-03 Cl/Br 179 >43,100 143 487 560 297 309 Charge Balance Cations (meq/kg) 0 .73 16 .85 0 .65 1 .42 2 .36 1 .54 1 .21 Anions (meq/kg) 0 .27 16 .80 0 .43 1 .20 2 .09 0 .74 0 .98 % difference 46 .89 0 .12 20 .27 8 .33 6 .22 35 .42 10 .51 Ionic Strength 0 .0008 0 .0171 0 .0008 0 .0017 0 .0034 0 .0018 0 .0015

PAGE 165

Table 5 Water Chemistry Analyses 1 5 7 Sample ID # 85 86 87 88 Type Unique # SW SW SW SW Aquifer Depth (ft) Peat Dome Surface Stream Stream Date 11/20/99 6/13/00 11/11/99 11/11/99 Temp (€C) 4 .2 NA 4 .3 4 .5 pH 4 .90 7 .26 6 .69 7 .05 Cond (mhos) 40 NA 85 210 Redox (mV) NA NA 123 70 D .O (ppm) NA NA 5 .1 7 .3 Cations (ppm) Ca 8 .3 18 .7 8 .0 26 .5 Mg 0 .97 6 .86 3 .98 9 .26 Na 7 .58 58 .7 2 .63 4 .04 K 3 .76 0 .93 0 .65 1 .44 Al 0 .401 0 .049 0 .040 0 .116 Fe 0 .69 0 .51 0 .572 2 .86 Mn 0 .044 0 .046 0 .013 0 .481 Sr 0 .031 0 .186 0 .020 0 .063 Ba 0 .791 0 .046 0 .022 0 .062 Si 6 .2 3 .79 4 .8 7 .1 Anions (ppm) Alk (as CaCO3) 2 .98 36 .4 30 73 Cl 4 .98 111 3 .22 9 .3 Br 0 .01 0 .042 0 .01 0 .01 N02-N <0 .005 <0 .005 <0 .005 N03-N 0 .022 0 .079 0 .165 0 .397 S04 0 .34 1 .28 3 .66 6 .83 P04-P <0 .02 <0 .02 <0 .02 <0 .02 total P 0 .16 <0 .6 <0 .02 0 .03 F 0 .59 0 .04 0 .03 0 .05 Sr/Ca 3 .73E-03 9 .95E-03 2 .50E-03 2 .38E-03 Cl/Br 498 2,643 322 930 Charge Balance Cations (meq/kg) 0 .94 4 .08 0 .86 2 .30 Anions (meq/kg) 0 .25 3 .90 0 .78 1 .90 % difference 57 .53 2 .31 4 .74 9 .54 Ionic Strength 0 .0008 0 .0046 0 .0011 0 .0029

PAGE 166

Table 6 Trace Element Analyses 1 5 8 Sample ID# Unique # Aquifer Date 1 449381 PMHN 3/29/00 2 449382 PMHN 3/29/00 3 217182 PMHN 3/29/00 4 538117 PMHN 3/29/00 5 562973 PMHN 3/29/00 6 465946 PMHN 3/31/00 7 242053 PMHN 6/13/00 Elements (ppb) Li 7 0 .9 0 .8 2 .4 7 .2 5 .4 3 .4 2 .5 Be 9 <0 .3 <0 .3 <0 .3 <0 .3 <0 .3 <0 .3 <0 .3 B11 19 .7 44 .0 49 .7 56 49 168 22 .6 Al 27 21 .5 24 .6 1 .8 3 .7 1 .35 1 .8 <1 .5 V 51 NA 1 .44 0 .37 <0 .04 0 .05 0 .87 0 .45 Cr 52 1 .94 1 .60 1 .04 1 .29 0 .97 0 .99 0 .67 Fe 54 626 2680 5380 2050 1810 NA 278 Mn 55 17 .8 103 .9 364 194 133 .9 NA 174 .5 Co 59 0 .08 0 .59 0 .36 0 .95 1 .62 0 .53 0 .55 Ni 60 1 .65 1 .54 2 .45 2 .04 1 .88 1 .63 0 .43 Cu 63 39 0 .77 1 .49 0 .66 2 .01 0 .39 0 .62 Zn 66 1560 100 .1 192 200 164 175 .7 169 As 75 0 .42 0 .50 1 .07 0 .28 0 .84 1 .9 9 .1 Br 79 NA 7 .4 21 .9 21 .8 39 .9 16 .9 10 .4 Rb 85 1 .2 0 .66 0 .90 0 .80 1 .1 0 .47 0 .27 Sr 86 29 .8 34 .1 66 .2 95 .8 98 .4 81 .8 34 .9 Mo 98 0 .11 0 .07 0 .12 0 .07 0 .59 0 .27 0 .42 Cd 114 0 .014 0 .033 0 .018 0 .026 0 .023 0 .033 0 .014 Sn 120 0 .063 0 .064 0 .077 0 .033 0 .040 <0 .015 0 .020 Sb 121 0 .054 0 .030 0 .050 0 .015 0 .027 0 .018 0 .015 1127 0 .25 0 .18 0 .92 0 .94 1 .02 1 .06 1 .4 Cs 133 <0 .012 0 .005 0 .008 0 .012 0 .009 0 .010 <0 .012 Ba 138 144 .4 489 t NA 636 581 521 132 .2 Pb 208 0 .097 0 .109 0 .152 0 .16 0 .45 0 .51 0 .47 U 238 0 .011 0 .011 0 .030 0 .43 0 .111 0 .011 <0 .012

PAGE 167

Table 6 Trace Element Analyses 1 5 9 Sample ID# Unique # Aquifer Date 8 219291 PMHN 3/29/00 9 416555 PMHN 5/22/00 10 574578 PMHN 5/22/00 11 W03120 PMHN 5/22/00 12 574577 PMHN 5/22/00 13 493869 PMHN 3/31/00 14 473703 PMHN 3/31/00 Elements (ppb) Li 7 2 .9 2 .5 3 .5 3 .8 2 .7 2 .2 14 .7 Be 9 <0 .3 <0 .3 <0 .3 <0 .3 <0 .3 <0 .3 <0 .3 B 11 43 .0 22 .6 29 .5 23 .0 23 .4 25 .0 93 Al 27 2 .2 2 .6 6 .3 0 .97 1 .3 12 .9 2 .7 V 51 0 .84 0 .34 0 .03 1 .2 3 .97 37 15 .22 Cr 52 0 .88 1 .91 1 .79 2 .12 2 .9 18 .1 1 .65 Fe 54 139 NA 4960 <6 8130 NA 50 Mn 55 4 .8 322 471 20 .2 163 NA 3 .8 Co 59 0 .027 2 .68 2 .44 0 .063 1 .05 1 .14 0 .91 Ni 60 0 .39 6 .1 2 .58 1 .01 3 .29 4 .76 4 .3 Cu 63 0 .89 0 .40 1 .41 34 .8 1 .23 0 .63 6 .17 Zn 66 136 .1 100 94 61 .4 35 47 .4 130 As 75 <0 .2 9 .1 1 .5 0 .16 1 .71 1 .55 1 .37 Br 79 12 .5 10 .4 11 .8 12 .3 14 .4 13 .5 46 Rb 85 0 .10 0 .27 0 .69 0 .44 1 .3 1 .3 0 .53 Sr 86 28 .9 34 .9 39 .9 33 .6 40 .4 22 .7 98 .2 Mo 98 0 .05 0 .42 0 .22 0 .22 0 .62 0 .22 0 .98 Cd 114 0 .082 0 .014 0 .058 0 .031 0 .025 0 .016 0 .072 Sn 120 0 .027 0 .020 0 .024 0 .016 0 .090 0 .112 0 .034 Sb 121 <0 .03 0 .015 <0 .024 0 .076 0 .090 0 .031 0 .12 1127 0 .88 1 .4 1 .99 0 .93 1 .47 0 .59 0 .54 Cs 133 <0 .012 <0 .012 <0 .012 <0 .012 <0 .012 0 .006 0 .009 Ba 138 377 132 .2 115 .7 82 86 494 403 Pb 208 0 .075 0 .47 0 .81 0 .273 0 .132 0 .35 0 .55 U 238 <0 .012 <0 .012 0 .014 0 .128 0 .161 0 .022 1 .29

PAGE 168

Table 6 Trade Element Analyses 1 6 0 Sample ID# Unique # Aquifer Date 15 219400 PMHN 3/27/00 16 507417 PMHN 3/27/00 16 Replicate 507417 PMHN 3/27/00 16 Replicate 507417 PMHN 3/27/00 17 142934 PMHN 3/31/00 18 575167 QBAA 3/29/00 19 436792 PMVU 6/12/01 Elements (ppb) Li 7 3 .3 2 .5 2 .4 3 .4 5 .2 4 .1 9 .93 Be 9 <0 .3 <0 .3 <0 .3 <0 .3 <0 .3 <0 .3 NA 1311 42 .1 5 5 16 .8 40 .4 36 .0 2160 Al 27 50 .2 NA NA 39 1 .9 2 .6 8 .0 V 51 0 .58 0 .16 0 .19 0 .23 0 .04 0 .11 NA Cr 52 0 .57 0 .62 0 .69 0 .72 0 .49 1 .05 1 .20 Fe 54 356 3080 3080 3020 1200 750 12 .1 Mn 55 13 .6 20 .4 21 .6 22 .9 NA 51 .8 8 .7 Co 59 0 .095 0 .080 0 .12 0 .099 0 .85 0 .08 0 .06 Ni 60 2 .26 1 .36 1 .94 1 .49 1 .34 1 .07 0 .86 Cu 63 28 .4 3 .11 12 .9 6 .26 1 .70 0 .83 1 .45 Zn 66 22 .5 125 138 129 147 150 33 .4 As 75 0 .50 8 .1 8 .2 8 .1 0 .10 0 .7 0 .015 Br 79 15 .9 8 .7 18 .7 11 .3 13 .1 13 NA Rb 85 2 .2 0 .68 1 .36 1 .1 0 .67 0 .28 1 .46 Sr 86 57 .7 57 .5 59 .6 58 .7 98 .6 66 .7 151 .6 Mo 98 0 .03 1 .74 1 .76 1 .69 0 .65 0 .55 NA Cd 114 0 .700 0 .148 0 .260 0 .296 0 .072 0 .060 <0 .05 Sn 120 0 .137 0 .151 0 .81 0 .23 0 .021 0 .066 NA Sb 121 0 .030 0 .047 0 .096 0 .070 0 .013 0 .018 NA 1127 0 .29 0 .6 0 .66 1 .39 1 .25 0 .42 NA Cs 133 0 .014 <0 .012 0 .012 0 .015 <0 .012 <0 .012 <0 .05 Ba 138 49 24 .6 78 .1 27 .3 260 529 50 .3 Pb 208 2 .98 3 .11 3 .33 3 .38 0 .396 0 .166 0 .35 U 238 0 .011 0 .006 <0 .012 0 .007 3 .98 0 .018 <0 .005

PAGE 169

Table 6 Trace Element Analyses 1 6 1 Sample ID# Unique # Aquifer Date 20 W30064 PMVU 6/12/01 21 N/A PMVU 6/22/00 22 N/A QUUU 3/29/00 23 N/A QUUU 6/22/00 24 467676 PMVU 6/18/01 25 172459 QBAA 6/18/01 26 172468 PMVU 6/18/01 Elements (ppb) Li 7 7 .19 6 .9 4 .3 3 .4 2 .4 2 .4 2 .3 Be 9 NA <0 .3 <0 .3 <0 .3 NA NA NA 1311 1256 690 78 70 27100 10500 9420 Al 27 4 .8 0 .4 4 .2 1 .9 9 .8 32 .5 9 .9 V 51 NA 0 .09 2 .4 2 .88 NA NA NA Cr 52 1 .94 1 .06 8 .8 1 .43 1 .85 1 .68 1 .66 Fe 54 626 <6 <6 NA 15 .7 14 .9 9 .3 Mn 55 17 .8 56 .1 NA 163 0 .99 1 .04 0 .95 Co 59 0 .08 0 .10 0 .39 0 .31 0 .05 0 .05 0 .05 Ni 60 1 .65 0 .71 1 .79 0 .85 0 .86 0 .74 0 .6 Cu 63 39 .0 3 .61 1 .57 16 .9 5 .35 4 .14 1 .74 Zn 66 1560 51 .4 47 271 11 .7 15 .5 6 .09 As 75 0 .42 0 .81 1 .4 0 .85 0 .88 0 .52 0 .99 Br 79 NA 64 .7 7 .0 12 .8 NA NA NA Rb 85 1 .17 0 .93 0 .80 0 .21 0 .68 0 .40 0 .43 Sr 86 128 269 208 53 .6 18 .7 16 .5 25 .3 Mo 98 NA 0 .46 1 .3 0 .10 NA NA NA Cd 114 <0 .05 0 .031 0 .010 0 .060 0 .18 0 .06 0 .04 Sn 120 NA 0 .044 <0 .015 0 .052 NA NA NA Sb 121 NA <0 .024 0 .040 0 .025 NA NA NA 1127 NA 0 .65 2 .1 0 .85 NA NA NA Cs 133 <0 .05 0 .015 <0 .012 <0 .012 <0 .05 <0 .05 <0 .05 Ba 138 31 .8 152 .4 325 514 58 .1 55 .2 62 .2 Pb 208 5 .4 0 .08 0 .033 0 .126 0 .74 0 .65 0 .21 U 238 0 .54 0 .017 1 .72 0 .101 0 .07 0 .05 0 .12

PAGE 170

Table 6 Trace Eletnerit Analyses 1 6 2 Sample ID# Unique # Aquifer Date 28 467698 PMVU 6/18/01 29 530460 QBAA 6/19/01 31 N/A PMHN 3/27/00 32 582017 PMHN 3/27/00 33 551151 QWTA 6/18/01 34 113967 QBAA 3/28/00 35 538199 PMHN 3/31/00 Elements (ppb) Li 7 2 .2 6 .7 14 .6 1 .9 1 .15 7 .9 5 .8 Be 9 NA NA <0 .3 <0 .3 NA <0 .3 <0 .3 1311 17800 212 26 .0 17 .4 104 59 108 Al 27 7 .33 2 .3 NA 13 .2 2 .46 1 .95 18 .7 V 51 NA NA 4 .0 0 .07 NA 0 .39 0 .37 Cr 52 1 .53 3 .08 2 .6 0 .59 1 .46 1 .00 0 .88 Fe 54 12 .6 2960 NA NA 2 .85 109 6170 Mn 55 0 .73 409 NA 119 .4 58 .8 187 NA Co 59 0 .05 0 .12 0 .63 1 .02 0 .06 0 .116 0 .34 Ni 60 0 .53 1 .62 2 .32 1 .09 0 .87 1 .27 1 .60 Cu 63 2 .81 1 .45 6 .92 1 .02 1 .2 1 .22 0 .51 Zn 66 7 .09 28 .7 102 13 .8 22 .9 171 178 As 75 0 .8 9 .76 3 .77 15 .8 1 .7 0 .86 3 .81 Br 79 NA NA 21 t 14 .7 NA 15 .7 31 .3 Rb 85 0 .47 0 .45 1 .12 0 .75 0 .21 0 .60 0 .39 Sr 86 21 .5 153 70 .8 49 .8 81 .6 111 .2 69 .6 M698 NA NA 0 .54 2 .19 2 .5 0 .79 Cd 114 0 .04 <0 .05 0 .190 0 .027 <0 .05 0 .026 0 .015 Sn 120 NA NA 0 .28 0 .066 NA 0 .085 0 .011 Sb 121 NA NA 0 .27 0 .064 NA 0 .046 0 .020 1127 NA NA 0 .81 0 .68 NA 0 .69 1 .81 Cs 133 <0 .05 <0 .05 0 .019 0 .006 <0 .05 0 .006 0 .005 Ba 138 25 369 181 20 .7 68 .1 602 447 Pb 208 0 .420 0 .140 1 .56 0 .26 0 .220 0 .159 0 .98 U 238 0 .09 <0 .05 10 .04 <0 .012 0 .06 0 .97 <0 .012

PAGE 171

Table 6 Trace Element Analyses 1 6 3 Sample ID# Unique # Aquifer Date 36 123992 PMHN 6/19/01 37 433529 PMCV 6/19/01 38 N/A PMHN 3/27/00 39 599023 PMFL 3/28/00 40 452259 PMVU 3/29/00 40 Replicate 452259 PMVU 6/13/00 41 532850 PMCV 6/19/01 Elements (ppb) Liz 6 .7 3 4 .1 2 10 .8 10 .7 7 .4 Be 9 NA NA <0 .3 <0 .3 <0 .3 <0 .3 NA B 11 49 129 12 .8 56 .8 3600 4310 6700 Al 27 3 .5 1 .9 3 .3 1 .3 8 .6 <1 .5 10 .6 V 51 NA NA 0 .56 0 .41 0 .30 0 .34 NA Cr 52 3 .23 1 .95 1 .02 1 .04 0 .87 0 .58 20 .90 Fe 54 2260 166 17 3350 208 154 69 .2 Mn 55 280 69 .1 1 .1 151 .6 21 .5 19 .3 9 .3 Co 59 0 .13 0 .09 0 .029 0 .53 0 .10 0 .06 0 .05 Ni 60 1 .84 1 .26 1 .16 1 .60 0 .73 0 .48 0 .54 Cu 63 1 .78 1 .0 7 .95 0 .61 0 .77 1 .06 1 .43 Zn 66 29 .8 29 .7 67 .3 167 110 105 23 .2 As 75 10 .5 1 .01 0 .23 2 .8 0 .52 0 .50 0 .16 Br 79 NA NA <10 7 .2 442 445 NA Rb 85 0 .52 1 .07 1 .00 0 .34 1 .8 1 .55 1 .23 Sr 86 187 197 29 .8 144 248 251 55 .8 M098 NA NA 0 .04 0 .72 2 .16 1 .43 NA Cd 114 <0 .05 <0 .05 0 .043 0 .028 0 .018 0 .032 0 .04 Sn 120 NA NA 0 .132 0 .073 0 .012 0 .074 NA Sb 121 NA NA 0 .051 0 .011 <0 .024 <0 .024 NA 1127 NA NA <0 .3 0 .57 1 .87 3 .01 NA Cs 133 <0 .05 <0 .05 <0 .012 <0 .012 0 .059 0 .049 0 .07 Ba 138 281 70 18 .6 521 622 440 46 .8 Pb 208 0 .15 0 .090 0 .7 0 .092 0 .31 0 .08 0 .160 U 238 <0 .05 0 .22 0 .018 0 .146 0 .025 0 .024 <0 .05

PAGE 172

Table 6 Trace Element Analyses 1 6 4 Sample ID# Unique # Aquifer Date 42 581953 PMHN 3/29/00 43 433481 QWTA 6/18/01 44 142945 PMHN 3/28/00 45 620416 PMVU 5/24/01 46 634574 PMHN 3/28/00 49 N/A PMVU 6/22/00 50 597976 PMCV 5/22/01 Elements (ppb) Liz 1 .9 4 .4 2 .9 6 .4 2 .2 3 .9 7 .3 Be 9 <0 .3 NA <0 .3 NA <0 .3 <0 .3 NA B11 65 5 56 .7 84 42 .5 5 58 Al 27 21 .2 3 .6 2 .6 2 .71 2 .9 <0 .5 6 .14 V 51 2 .13 NA 0 .64 NA 0 .54 2 .11 NA Cr 52 1 .94 2 .86 1 .23 1 .59 1 .19 1 .59 2 .24 Fe 54 3220 2130 4010 0 .97 11 824 4880 Mn 55 394 468 130 .7 2 .12 0 .5 101 341 Co 59 1 .20 0 .21 0 .71 0 .04 0 .066 0 .075 0 .11 Ni 60 3 .3 2 .45 3 .51 0 .62 1 .00 0 .64 1 .61 Cu 63 1 .02 1 .56 1 .03 3 .1 3 .0 0 .58 2 .25 Zn 66 91 .1 46 .0 186 9 .36 172 32 .1 25 .7 As 75 0 .9 3 .5 2 .12 0 .46 0 .15 1 .03 1 .11 Br 79 23 .9 NA 28 NA 12 .1 9 .3 NA Rb 85 2 .7 0 .50 1 .01 0 .28 0 .60 0 .88 0 .65 Sr 86 48 .1 61 .7 55 .7 64 .5 45 .1 47 .7 140 M098 0 .05 NA 0 .16 NA 0 .09 0 .21 NA Cd 114 0 .026 0 .07 0 .044 <0 .05 0 .072 0 .035 0 .04 Sn 120 0 .051 NA 0 .066 NA 0 .054 0 .020 NA Sb 121 0 .013 NA 0 .019 NA 0 .027 <0 .024 NA 1127 2 .01 NA 1 .15 NA 0 .5 <0 .3 NA Cs 133 0 .010 <0 .05 0 .006 <0 .05 0 .007 0 .022 0 .03 Ba 138 592 141 490 6 .4 513 157 64 .4 Pb 208 0 .074 0 .830 0 .110 0 .520 0 .088 0 .087 0 .930 U 238 0 .020 <0 .05 0 .011 0 .04 0 .030 0 .156 <0 .05

PAGE 173

Table 6 Trace Element Analyses 1 6 5 Sample ID# Unique # Aquifer Date 51 567565 PMHN 3/27/00 52 143949 PMHN 3/28/00 55 A05 PMHN 5/17/00 55 A05 PHMN 6/13/00 58 A08 PMHN 5/22/00 56 A06 PMHN 5/26/00 57 A07 PMHN 5/22/00 Elements (ppb) Liz 5 .2 4 .7 21 .9 21 .5 7 .5 2 .2 2 .1 Be 9 <0 .3 <0 .3 <0 .3 <0 .3 <0 .3 <0 .3 <0 .3 B 11 9 44 .9 219 253 65 17 23 .0 Al 27 1 .3 2 .55 1 .6 <0 .5 0 .84 0 .9 3 .1 V 51 1 .19 0 .27 0 .65 0 .61 0 .21 0 .09 0 .21 Cr 52 1 .76 1 .85 1 .67 0 .54 2 .08 1 .90 0 .98 Fe 54 300 19 .5 4340 4400 NA NA <6 Mn 55 6 .5 0 .5 318 300 256 348 1 .7 Co 59 0 .035 0 .075 1 .58 1 .50 1 .15 1 .39 0 .043 Ni 60 1 .12 0 .86 2 .24 1 .1 1 .42 1 .50 0 .67 Cu 63 9 .26 3 .73 1 .65 0 .38 0 .71 0 .88 0 .99 Zn 66 6 .4 170 186 7 .6 29 28 .1 55 As 75 0 .29 0 .19 1 .37 1 .26 1 .12 3 .22 0 .12 Br 79 13 .5 22 .1 1100 1150 204 18 .9 17 .5 Rb 85 0 .84 0 .50 2 .05 2 .25 1 .41 0 .62 0 .32 Sr 86 30 .2 64 .8 671 693 186 49 .6 36 .5 Mo 98 0 .09 0 .13 0 .48 0 .46 0 .48 0 .55 0 .03 Cd 114 0 .087 0 .013 0 .108 <0 .003 0 .034 0 .076 0 .011 Sn 120 0 .060 0 .100 0 .053 0 .039 0 .006 0 .066 <0 .015 Sb 121 0 .036 0 .055 0 .040 <0 .024 <0 .024 0 .011 <0 .024 1127 0 .54 0 .78 5 .5 11 .6 3 .4 1 .07 0 .45 Cs 133 0 .011 0 .007 0 .034 0 .022 0 .009 <0 .012 <0 .012 Ba 138 0 .3 537 484 75 .3 134 79 .6 84 .6 Pb 208 0 .39 0 .160 1 .05 <0 .03 0 .67 0 .12 0 .26 U 238 0 .025 0 .305 0 .202 0 .202 <0 .012 <0 .012 0 .010

PAGE 174

Table 6 Trace Element Analyses 1 6 6 Sample ID# Unique # Aquifer Date 59 A09 PMHN 6/2/00 60 AlO PMHN 6/2/00 61 All PMHN 6/2/00 66 A20 QUUU 3/29/00 76 'DOl Surface 6/22/00 77 D52 Surface 3/27/00 Elements (ppb) Li 7 1 .6 2 .1 4 .1 1 .2 1 .2 3 .5 Be 9 <0 .3 <0 .3 <0 .3 <0 .3 <0 .3 <0 .3 B 11 23 35 37 50 .8 226 5 .6 Al 27 1 .1 1 .0 1 .6 3 .2 35 .8 NA V 51 <0 .04 0 .04 0 .84 4 .56 1 .34 0 .83 Cr 52 0 .69 1 .50 2 .10 3 .6 0 .98 0 .71 Fe 54 <6 NA NA 10 .5 899 140 Mn 55 0 .67 0 .65 10 .0 4 .0 71 .5 43 .4 Co 59 0 .046 0 .039 0 .086 0 .084 0 .435 0 .44 Ni 60 0 .41 0 .59 1 .28 0 .84 1 .86 1 .45 Cu 63 0 .72 0 .59 1 .14 0 .61 2 .12 2 .46 Zn 66 158 174 .6 205 144 .1 29 .8 12 .11 As 75 <0 .12 0 .06 0 .33 1 .42 0 .98 0 .20 Br 79 10 .6 11 .4 14 .3 <10 41 .0 <10 Rb 85 0 .10 0 .15 0 .24 0 .28 7 .81 1 .81 Sr 86 26 .7 33 .9 49 .6 44 .0, 39 .1 12 .7 Mo 98 0 .03 0 .03 0 .24 0 .17 0 .21 0 .04 Cd 114 0 .056 0 .06 0 .067 0 .019, 0 .18 0 .131 Sn 120' 0 .047 0 .037 0 .028 0 .023 0 .10 0 .070 Sb 121 0 .014 <0 .024 0 .020 0 .025, 7 .900 0 .050 1127 0 .48 0 .65 1 .05 0 .44 <0 .3 1 .0 Cs 133 <0 .012 <0 .012 0 .007 <0 .012 0 .063 0 .017 Ba 138' 423 412 492 507 NA 22 .7 Pb 208 0 .14 0 .07 0 .12 0 .26 0 .35 0 .43 U 238 <0 .012 0 .011 0 .33 0 .22 0 .013 0 .014

PAGE 175

Table 8 Isotopic Data 1 449381 3/29/00 12 .6 1 .0 ~ . 2 449382 3/29/00 13 .8 1 .1 3 217182 3/29/00 6/14/01 <0 .8 0 .5 0 .9951 0 .0058 -20 .2 Lever# -20 .2 NA 0 .995 41 40 4 538117 3/29/00 <0 .8 0 .5 _5 562973 3/29/00 5 .2 0 .6 6 465946 3/31/00 19 .8 1 .5 7 242053 6/13/00 <0 .8 0 .4 0 .7705 0 .0063 -23 Lever -28 NA 0 .938 529 500 8 219291 3/29/00 2 .9 0 .6 9 416555 5/22100 1 .1 0 .5 ~ . 10 574578 5/22/00 2 .6 0 .5 _‚ 11 W03120 5/22/00 12 .7 1 .0 ~ 12 574577 5/22/00 18 .4 1 .4 13 493869 3/31/00 14 473703 3/31/00 10 .0 0 .9 .ƒ 15 Replicate 219400 3/27/00 3/27/00 9 .8 10 .2 1 .0 0 .9 16 Replicate 507417 3/27/00 3/27/00 <0 .8 <0 .8 0 .6 0 .6

PAGE 177

Table 8 Isotopic Data Replicate 3/27/00 18 .1 1 .4 32 582017 3/27/00 <0 .8 0 .5 33 Replicate 551151 6/18/01 6/18/01 13 .4 14 .8 1 .1 1 .2 34 113967 3/28/00 12 .6 1 .0 35 538199 3/31/00 20 .6 1 .5 36 123992 6/19/01 1 .7 0 .5 37 433529 6/19/01 7 .1 0 .7 38 N/A 3/27/00 12 .1 1 .0 39 599023 3/28/00 6 .3 0 .7 40 452259 3/29/00 6/13/00 <0 .8 0 .5 0 .4221 0 .1145 -14 CH 4 -28 -65 0 .481 6058 6000 41 532850 6/19/01 <0 .8 0 .5 42 581953 3/29/00 9 .4 0 .9 43 433481 6/18/01 19 .2 1 .4 44 142945 3/28/00 13 .7 1 .1 45 620416 5/24/01 12 .3 1 .0 46 473738 5/31/01 5 .9 0 .6 47 634574 3/28/00 13 .3 1 .1

PAGE 178

Table 8 Isotopic Data 49 A Sand 6/22/00 19 .3 t 1 .4 50 Replicate 597976 5/22/01 5/22/01 31 .3 t 2 .2 27 .7 t 2 .0 51 567565 3/27/00 2 .4 t 0 .6 52 143949 t 3/28/00 20 .2 t 1 .6 54 A02 t 6/14/01 <0 .8 t 0 .6 0 .818 t 0 .004 -15 .7 CH 4 t -28 t -65 t 0 .983 t 145 100 55 Replicate A05 t 5/17/00 6/13/00 <0 .8 t 0 .6 0 .538 t 0 .003 -15 .5 CH 4 t -28 t -65 t 0 .642 t 3662 4000 56 Replicate 56 A06 t 5/26/00 5/26/00 6/14/01 2 .8 t 0 .5 3 .6 t 0 .6 <0 .8 Attempted 14 C Not Analyzable 57 A07 t 5/22/00 10 .7 t 0 .9 58 A08 t 5/22/00 0 .9 t 0 .5 59 A09 t 6/2/00 12 .2 t 1 .0 60 A10 t 6/2/00 16 .2 t 1 .3 61 All t 6/2/00 18 .7 t 1 .4 66 A20 t 3/29/00 21 .5 t 1 .6 67 A21 t 3/14/01 12 .0 t 1 .0 68 A22 t 3/14/01 17 .4 t 1 .3

PAGE 179

Table 8 Isotopic Data 69 A23 3/14/01 <0 .8 0 .5 70 A24 3/15/01 9 .2 0 .9 76 N/A 6/22/00 11 .7 0 .9 77 N/A 3/27/00 10 .5 1 .0 + See text, Ch 2 for description of age calculations Age calculations based on a methanogenesis model are desiganated by "CH4" #Age calculations based on a lever law are designated by "lever" Key To Sample Type : Enriched Tritium data from : 14C Data from : MU = Municipal Well Environmental Isotopes Laboratory Beta Analytic Laboratory PS = Public Supply Well University of Waterloo 4985 S .W 74th Court DO = Domestic Supply W Waterloo, Ontario Miami, Florida 33155 SP = Spring Canada, N2L 3G1 USA SW = Surface Water