GROUND WATER PROTECTION
STATE OF INDIANA DRAFT GENERIC
PESTICIDE MANAGEMENT PLAN
Appendix B of the IN SMP
PRESENTATION OF
MAPS OF HYDROGEOLOGIC TERRAINS AND
SETTINGS OF INDIANA - NORTHERN PART
"IT MUST BE EMPHASIZED THAT THIS PRODUCT IS NOT
A SUBSTITUTE FOR SITE-SPECIFIC STUDIES,
...."
Nature of Ground-Water Flow and Sensitivity
in Glacial Sequences
The movement of ground water through geologic sequences is a direct
manifestation of the distribution of hydraulic potential. Hydraulic potential
is defined by hydraulic head-the level to which water will rise in a properly
constructed well developed at any particular point in the sequence. The
overall pattern of ground-water movement through a sequence is known as
a flow system, which typically includes both distinct and indistinct areas
of ground-water recharge and discharge. 'The distribution of hydraulic potential
is controlled by the interaction of gravity (manifested by pressure and
elevation within the system) and the geometric arrangement of bodies of
contrasting permeability. Settings associated with highly heterogeneous
sequences are likely to be characterized by large and abrupt permeability
contrasts, resulting in complex and highly localized flow patterns. On the
other hand, settings dominated by sequences having few major permeability
contrasts are likely to characterized by relatively more uniform hydraulic
potential and flow patterns of a simpler and more regional aspect.
The concept of 'aquifer sensitivity' relates to the intrinsic
hydrogeologic susceptibility of an aquifer to contamination from the universe
of surface or near-surface sources. In contrast, the idea of 'vulnerability',
adds to sensitivity the potential for contamination from actual land-use
practices or specific contaminants and sources. Both of these concepts ultimately
represent the forecasting of the probability of contamination events occurring
at some time in the future, and are thus entirely time dependent. Stated
slightly differently, sensitivity to contamination for the great majority
of agricultural chemicals is largely dependent on the fact that ground water
moves from place to place, and relative differences in sensitivity are directly
attributable to differences in travel time that characterize flow systems
in different settings, or in different parts of the same setting. The measurement
of travel time of ground water and various contaminants in a particular
system is largely indirect, costly, and time-consuming. Even when such measurements
are made, the results are commonly problematic and may be characterized
by large standard deviations. Therefore, most of the major predictive sensitivity
models (e.g., DRASTIC, SEEPAGE) attempt to use a combination of physical
properties (represented by mapped "factors') as a proxy for travel
time. In some hydrogeologic settings (including some of those mapped in
this investigation), travel time (and thus sensitivity) are likely to be
extremely high, that is 'water and (or) contaminants moving vertically will
reach the aquifer in hours to weeks". In contrast, in other settings
they are likely to be very low, that is 'decades or centuries may pass before
water moving vertically reaches the aquifer'. Such assessments are typically
based upon specific knowledge of the hydrogeologic setting, that is thickness,
permeability, and integrity of aquifer-capping units, depth and configuration
of the water table, hydraulic gradients, etc.
Veracity and Application of Sensitivity
Assessment Techniques
'The basic intent of recent sensitivity assessment efforts is
to generally identify regions of potentially elevated sensitivity such that
appropriate planning and management authorities can protect ground-water
quality by implementing best management practices and educational initiatives,
or in extreme cases, by limiting or specifying the design criteria of certain
practices perceived to have a negative impact on ground-water quality. The
general basis for these assessments is typically some type of 'stacking'
of broad, 'area-based' criteria, each interpreted from disparate sets of
data of widely varying quality and scale. The very large range of variation
inherent in most geologic terrain in general, and glaciated terrains in
particular, makes it unwise to apply the results of these exercises at anything
other than the broadest regional scale, i.e., mainly for comparative purposes
of units defined by broad regional terrain contrasts. Such comparisons are
mainly useful for planning and educational purposes. Even when they are
limited to these applications, however, the results of sensitivity assessments
may be ambiguous as well as problematic. The veracity of such broad-based
assessments is still in doubt; few have ever received rigorous testing to
authenticate their relative accuracy (National Research Council, 1993).
Techniques for the assessment of sensitivity fall under two general
categories: area-based and point-based. Area-based methods are exemplified
by the now well-known 'DRASTIC' approach (Aller and others, 1987; also see
discussions in Zaporozec, 1989, and in preparation; and Geologic Sensitivity
Project Workshop, 1991), in which various 'factors' are quantified, weighted
(i.e., interpreted as to level of relative importance), and mapped. Numerous
such map layers are superposed and a composite map of sensitivity map is
generated by computer or by manual interpretation. The area approach is
utterly dependent upon individual interpretation in creating each
of the various map layers. Stated somewhat differently, it is highly unlikely
that two (or more) qualified professionals will produce the same sensitivity
map using the area approach, assuming the same data distribution to begin
with. Thus, this method stacks layer upon layer of interpretation, with
little or no attention given to assessing the resulting total variation
of mapped sensitivity units or of the data used to compile individual layers.
Indeed, the very structure of the approach is not amenable to rigorous statistical
analysis or other measures that might help to quantify variance or indicate
the relative accuracy of mapped sensitivity units.
In point-based methods, by contrast, individual data points
are analyzed with respect to travel time of water and contaminants (see
Passero and others, 1988; Zaporozec, 1999, and in preparation). Various
weighting systems may be applied and sensitivities may be mapped or categorized
in some fashion, commonly in the context of broadly defined hydrogeologic
settings. In this approach, a sensitivity index (SI) score is derived
for every well record by tallying aggregate thickness of reported "clay
layers' above a 'used' aquifer, modified perhaps by other factors, such
as saturated thickness of aquifer and position of clay units and aquifers
in the total sequence. This approach is not influenced by interpretational
biases of individual map-makers; instead, it assesses the basic nature of
the geologic terrain in a purely analytical fashion, and thus provides a
unified criteria for comparisons over a particular area or between different
areas. Perhaps of greatest significance, individual scores are directly
related to actual "used" aquifers in any given area, unlike the
results of area-based methods, which commonly make no reference to any particular
aquifer. SI scores then form the basis for geostatistical treatment of mapped
hydrogeologic settings, which can suggest trends and help to quantify the
all-important element of variance. SI scores also afford a direct means
for statistical comparisons with databases showing the nature and distribution
of known ground-water contamination incidences-in other words, a somewhat
quantitative means of testing and validating a sensitivity analysis, as
well as directing future ground-water monitoring and management efforts.
A significant conceptual shortcoming of virtually all of the sensitivity models is that they do not directly take into account the relative hydraulic potential-i.e., the force that drives ground-water flow and that ultimately gives the physical properties of a setting their significance. Instead, most of the commonly employed area-based methods substitute such factors as 'depth to water' and 'recharge potential' generally simplistic concepts that commonly have no real meaning for hydrogeologic systems in many terrains. The distinction between these 'proxies' and the true character of ground-water flow in a setting is of no small consequence-in some discharge areas, owing to upward flow, it is physically impossible for contaminants (other than those heavier than water) to penetrate subjacent aquifers. Yet, many discharge areas are characterized by very high 'recharge potential' and are wetlands (i.e., emergent water table)-factors that would tend to produce highly elevated sensitivity scores in a conventional 'stacked layer' type of assessment. The 'depth-to-water' component is also relatively meaningless in the typically
complex glacial terrains of the upper Midwest, which contain multiple
confined aquifers in various positions as well as one or more water tables
perched at relatively shallow depth in tight till confining units. A partial
solution to this difficulty may be to include relative position(s) in the
ground-water flow system as part of the conceptual definition of a 'hydrogeologic
setting'. Settings defined in such a way (Fleming, 1992; 1994) yielded sensitivity
results that were substantially different from the 'expected' results using
settings defined in the conventional manner. In particular, sensitivities
of hydraulically defined ground-water recharge areas (downward flow) were
significantly elevated relative to those of ground-water discharge areas
(upward flow), regardless of the nature of the sequence or other factors.
Hydraulic definition of distinct recharge and discharge areas
is commonly problematic unless relatively abundant hydraulic head data are
available. Such data, when compiled in both map and cross-sectional formats,
provide useful three-dimensional representations of hydraulic potential
and thus allow areas where recharge and discharge are actually taking place
to be identified. Unfortunately, potentiometric surface maps are available
for only a small percentage of the area covered by this study, and the time-frame
of the study did not permit additional maps to be compiled. It is generally
possible, however, to make reasonable inferences about hydraulic conditions
based on a combination of surface topography, the nature and configuration
of surface water bodies, and the hydraulc properties (and contrasts thereof)
within the sequence below. Descriptions of all of the hydrogeologic settings
presented herein thus contain at least qualitative references to their likely
ground-water flow patterns; on the other hand, the position in the flow
system can be defined rather well for some of the settings based on existing
potentiometric surface maps. This information, however inferential, is critically
important for interpreting the relative sensitivities of the settings as
well as for many other tasks, such as evaluating the results of ground-water
monitoring networks.
GENERAL DESCRIPTION OF HYDROGEOLOGIC SETTINGS
MAPS
Heirarchy of Map Units
The maps and accompanying descriptions of hydrogeologic settings
are based on a heirarchical system of classification for the various map
units. The nature of this heirarchy is dictated by regional patterns, which
are significantly different between the far northern part of the state and
the central till plain area. Consequently, the methods used to classify
these two regions differ slightly in the ways that individual units are
defined, but the overall concept is similar. The major terrains of northern
Indiana occur in five distinct groups, or tracts, referred to as subbasins.
In turn, as many as a dozen discrete hydrogeologic settings can be identified
within a particular terrain. The terrains and settings within a particular
subbasin tend to be related in terms of the processes by which they formed,
as well as by the predominant hydrogeologic environments they represent.
Such distinctions are less evident within the central till plain,
which could be viewed as a kind of "mega-subbasin" unto itself.
The central till plain is subdivided in to nine "segments" of
widely-ranging size, some of which may or may not represent "subbasins"
in the same sense as those defined in northern Indiana. These segments are
generally distinguished on the basis of broad contrasts in features that
might be expected to affect regional hydrogeologic behavior and properties,
and are in large part delimited by geography, as opposed to being purely
terrain-defined. Most of the segments contain from two to five internal
terrains; in most cases each of these terrains is a hydrogeologic
setting.
Despite these regional differences, however, all of the major
map units north of the 40th parallel are part of a heirarchical, hydrogeologically-defined
land-system model. Progressing from the largest to smallest, the elements
of this system are defined by: the northern subbasins and the central till
plain and its segments; glacial terrains; and hydrogeologic settings and
subsettings. All of the settings and subsettings are further defined by
a set of internal elements, which are in part represented by contoured ranges
of: till or clay capping unit thickness; unconfined sand and gravel thickness;
water table depth; and bedrock depth; as well as qualitative descriptions
of ground-water flow patterns and other features. Overall boundaries of
the several subbasins, the central till plain, and its segments are not
explicitly shown on the hydrogeologic settings maps. Instead, the maps depict
the boundaries of their constituent terrains and hydrogeologic settings,
whose overall pattern defines the distributions of these larger elements.
Table 2 illustrates the relationship of specific
subbasins and till plain segments to the terrains and hydrogeologic settings
specifically mapped during this study.
It is essential to note that all these map units, and the internal
variability they represent are entirely scale-dependent; that is, the present
scale of 1: 100,000 essentially determines the ability to resolve these
units and hence has a direct affect on the range and nature of variation
represented by each. Stated somewhat differently, a unit that may appear
to be highly variable at a scale of 1: 100,000 may be much less so at a
larger map scale because the larger scale may afford the ability to resolve
the unit into various endmembers. On the other hand, some units may be so
heterogeneous that no further resolution is possible, except perhaps at
a scale of 1: 1. Likewise, the nature and time-frame of the study resulted
in a built-in bias in the sense that some units or regions are simply much
better known or have much better subsurface control than others. These situations
illustrate precisely the utility and flexibility of the glacial terrain
approach and its sequence-based map units, relative to traditional mapping
approaches wherein map units are specified by a narrow range of 'materials'.
Descriptions of Maps and Map Units
Subbasins
Subbasins are very large areas, commonly thousands of square miles
in size. They consist of numerous internal terrains that have generally
experienced a common history, origin, or pattern of deposition. Although
considerable hydrogeologic differences exist between their internal terrains,
the various subbasins are each typified by certain types of sequences and
landscape patterns. This may in turn lead to gross similarities in hydrogeological
behavior or characteristics within a particular subbasin and relatively
large differences between separate subbasins. In a sense, these subbasins
could reasonably be construed as 'ground-water provinces' of a sort. Contrasting
examples of subbasins include: the Maumee subbasin, defined by the distribution
of clayey Erie Lobe tills that occur in a variety of sequences and arrangements
with various granular sediments, and which tend to dominate shallow ground-water
flow in many terrains; and the Lake Michigan rim, which consists of a concentric
series of small to large, arcuate end moraines, outwash fans, sluiceways,
and beach features that collectively outline various ice margins and glacial
lakes that existed temporarily during the phased retreat of the Lake Michigan
glacier.
Central Till Plain and Segments
The central till plain is the broad region of generally very low
relief that occupies most of central Indiana. Although it was formed as
a result of numerous glacial episodes involving different ice lobes, the
central till plain is characterized by sequences and landscapes of remarkably
similar overall aspect. Regional differences are much less pronounced than
in and between the northern subbasins, and are differentiated by relatively
subtle contrasts of transitional nature. The central till plain is thus
divided into several segments on this basis. Some segments are relatively
distinct because of terrain or geographical distribution, whereas others
are much less so. Contrasting examples of segments include the East Fork-Whitewater
segment, an area comprised in part of varied morainal topography that is
deeply dissected by a distinctive system of tunnel valleys at the heads
of several major surface drainages; and the lower Wabash segment, a relatively
nondescript plain underlain at most places by a monotonous sequence of loam
till.
Glacial Terrains
Glacial terrains are generally defined by one or more distinctive
depositional sequence(s) that are associated with a characteristic landscape
or group of landforms. Glacial terrains may be of any size, however, most
are regional-scale features covering hundreds of square miles. Each glacial
terrain in turn may contain a variety of hydrogeologic settings or the entire
terrain may constitute only one or two settings, depending on the nature
of internal variation and the ability to resolve that variation. Contrasting
examples of glacial terrains include the Topeka fan, essentially a single
massive sequence of outwash and much lesser till with a relatively simple
upland landscape; and the Huntertown terrain, which is essentially defined
by its great diversity of sequences and landscapes. In contrast to the Topeka,
the Huntertown terrain generally consists of numerous local sequences of
greatly differing character that were deposited in association with an extensive
mass of stagnant ice and which now underlie a somewhat amorphous 'dead-ice'
landscape.
Hydrogeologic Settings and Subsettings
Hydrogeologic settings are defined by the observed or expected
relationship between the geologic environment, as represented by glacial
terrain, and the occurrence and movement of ground-water. Different settings
within the same terrain are usually indicative of changes in hydrogeologic
regime due to variations in sequence or landscape. Individual settings range
widely in size. Some settings are characterized by a relatively narrow range
of conditions, whereas others are specifically defined by a wide range of
local hydrogeologic conditions. -Continuing the glacial terrain example
presented above, the Topeka tan consists of three settings that differ primarily
on the basis of relatively minor differences in the proportions and geometry
of outwash and capping till units; whereas the Huntertown terrain contains
seven settings that differ significantly not only on the basis of contrasting
sequences, but primarily because each setting is defined by a very different
range and nature of internal variation.
Subsettings represent areas of generally local extent that differ
from the overall setting in one or more aspects that may affect hydrogeologic
behavior to a certain degree but do not alter the basic definition or fundamental
characteristics of the setting. Subsettings are defined for only a relatively
small number of settings. Small areas of dunal topography within an outwash
plain are an example of a subsetting.
Nomenclature and Labeling of Terrains and Settings
Terrains and settings are identified by informal names that are
based on a combination of geologic, hydrogeologic, and geographic attributes.
It would have been much simpler to simply group all of these units into
generic categories (e.g., 'till plains', 'outwash", etc.) but such
an approach would severely hinder a discussion of nuances and characteristics
of geographically and geologically different units within a particular category.
Therefore, most of these features are given informal geographic names wherever
possible in order to better highlight specific attributes that distinguish
one unit from another and that may create distinct and significant patterns
of hydrogeologic behavior. Map unit labels generally consist of from one
to as many as four alpha-numeric characters, depending on the type of terrain,
and the general scheme of these characters is similar for all map units.
Outwash fans all are identified by the letter
'F' as the first character of their labels. Further characters differentiate
the particular type of sequence, as follows:
Fl: predominantly till-capped fan
F2: predominantly exposed (unconfined) outwash fan
F3: exposed fan deposited over highly irregular or collapsed substrate
Fx: fan-head complex characterized by complexly interbedded till
and outwash havinhg closely intermingled capped and exposed areas.
Additional upper-case letters at the end of the label identify
the specific name of the fan; i.e., FxT, FIT, and F2T all refer to the Topeka
fan.
Outwash plains and sluiceways are identified
by the letter 'O' as the first character. The second character depicts the
type of terrain, as follows:
01: sluiceway, or discrete channel
02: outwash plain
Ox: hummocky outwash complex of irregular dimensions
Additional upper-case letters at the end of the label identify
the specific name of the feature; i.e., 02S refers to the St. Joseph outwash
plain.
Till-cored settings are generally identified by the first letter of:
A) the major terrain they are associated with and, where possible,
B) a formally named till-stratigraphic unit (Wayne, 1963) that
caps or cores the setting. The latter are few and include:
1) the Lagro Formation: clayey tills of the Erie Lobe, designated by the letter 'L'
2) the Trafalgar Formation: loam tills of the Huron-Erie Lobe, designated by "T*
3) the Snyder till member: silty to clayey tills of the Lake Michigan Lobe, designated by 'S'
4) the Wedron Formation (undivided): predominantly loam tills
of the Lake Michigan Lobe, designated by the latter 'W'
The use of a lower-case 'm" within the character string denotes
morainal terrain or ridges composed chiefly of till.
Examples of labels include:
N: till-cored rolling upland of the Nappanee upland
TK: till-capped upland of the Kosciusco morainal region, formed on loam tills of the Trafalgar Formation
TmK: end moraines of the Kosciusco morainal region cored by loam till of tile Trafalgar Formation
Ims: till ridges of the Iroquois morainal region formed on clayey
tills of the Snyder member
Other types of settings are also distinguished by their map labels,
for example:
Tt: mmel valleys
ics: ice-contact stratified sand and gravel
w: washed areas or highlevel meltwater channels
tj: James Lake trough system
Familiarity with the labelling system is useful but not essential.
The labels are specifically identified with their respective terrains and
settings in table 2 and via the explanations of the individual settings,
which give the full names of the settings and the map unit labels.
Contour Maps
Generalized thickness and depth ranges of several internal elements
are shown via contour maps. These maps are highly generalized and employ
variable contour intervals and thickness ranges according to the nature
of variation relative to the map scale of 1: 100,000 and the quality and
quantity of data used in their construction. Consequently, the mapped contour
or thickness ranges do not necessarily follow an orderly progression at
any given locality.
Capping or confining unit thickness maps were constructed for
settings characterized by surface till units or other predominantly fine-grained
sequences. Contoured thickness ranges are continuous between adjacent capped
or confined swings, however, no distinction is made on these maps as to
specific type of capping unit. The general nature of fine-grained confining/capping
units is summarized in the descriptions of individual settings. Settings
so mapped are obviously characterized by aquifers whose tops lie at some
depth below the land surface. It should be emphasized, however, that aquifers
in these settings are not necessarily confined in the strict sense
of the word; in many cases, a continuous zone of saturation may lie at depths
well below the interface between a particular aquifer and its capping unit.
In such cases, these aquifers are not under artesian conditions and are
therefore simply 'capped' rather than 'confined' by sediments of lesser
permeability.
During map construction, an attempt was made to follow provisional
25-foot thickness intervals for capping units less than 50 feet thick. The
natural variation within some of the settings made this procedure problematic
and necessitated the use of wider ranges (table 3).
Moreover, 'capping unit thickness" was defined by the first granular
unit 10 or more feet thick reported in the data point. Therefore, 'capping
unit thickness' commonly includes some granular units of less than 10 foot
thickness. Most importantly, no systematic attempt was made to resolve the
geometries of the specific granular units that define the base of 'capping
unit thickness' at any given data point. Consequently, "capping unit
thickness" ranges widely, and these maps should not be used to infer
the presence, extent, or significance of aquifers immediately underlying
the contoured capping unit.
In other words, it commonly cannot be inferred whether abrupt
contrasts in capping unit thickness are attributable to mere isolated lenses
of sand and gravel or to more robust and extensive bodies.
Unconfined sand and gravel thickness maps
were constructed for settings characterized by unconfined sand and gravel
units of mappable and apparently continuous extent at a scale of 1: 100,000.
Such units may or may not exist in a variety of other settings, but are
generally too small or of uncertain continuity to be included in these maps.
Therefore, the contoured thickness ranges generally apply to relatively
large surface sand and gravel bodies whose thickness and continuity can
be more or less recognized through subsurface data or surface expression.
The contoured thickness ranges are continuous between adjacent unconfined
settings; however, no distinction is made on these maps as to specific type
of sand and gravel unit, but the general nature of these bodies and the
larger terrains they are part of are summarized in the descriptions of individual
settings.
It should be emphasized that the surface sand and gravel units
contoured in these settings are not necessarily aquifers in the strict
sense of the word; in some settings, large parts of these bodies may be
unsaturated, and they are not utilized as sources of water. At other places,
however, they constitute significant resources. A general idea of the relation
between the zone of saturation and the thickness of surface sand and gravel
units can be determined by comparing these maps with the maps showing water
table depth. The significance of these units as a water supply or as potential
conduits for saturated contaminant migration is summarized in the
descriptions of their respective settings, at least to the extent supported
by existing information.
During map construction, an attempt was made to follow provisional
25-foot thickness intervals for surface sand and gravel bodies less than
50 feet thick. The natural variation made this procedure problematic and
necessitated the use of wider ranges in some settings (table
4). Moreover, 'unconfined sand and gravel thickness' was defined by
the first fine-grained unit 10 or more feet thick reported in the data point.
Tberefore, 'unconfined sand and gravel thickness' commonly includes some
till or other fine-grained units of less than 10 foot thickness. No systematic
attempt was made to resolve the geometries of fine-grained units that define
the base of 'unconfined sand and gravel thickness" at any given data
point. Consequently, 'unconfined sand and gravel thickness' ranges widely
depending on the distribution and characteristics of embedded or underlying
fine-grained units. This procedure may lead to somewhat misleading thickness
ranges in some of the more complex outwash bodies, where locally persistent
till units in the middle of sand and gravel sequences may give the appearance
of much lesser total sand and gravel thickness than actually exists.
Where known, the presence of such 'tiered' aquifer systems is noted in the
descriptions of individual settings.
Depth to water table maps were prepared for settings characterized
by unconfined sand and gravel bodies and for several other types of settings
characterized at least locally by relatively permeable surface sediments.
The latter are primarily in low-lying parts of the landscape, and may represent
situations where some surface sand and gravel may be present, at least on
a local, discontinuous basis but could not be resolved at the present scale.
The contoured depth ranges are continuous between adjacent settings; however,
no distinction is made on these maps as to specific type of water table
aquifer, but the general nature of these units and the larger hydraulic
regimes they are thought to be part of are summarized in the descriptions
of individual settings.
The water table depth maps were derived in part from water level
data reported by drilling contractors for wells constructed in unconfined
aquifers. It should be noted that water level data may be misleading due
to the presence of vertical gradients. Consequently, water levels measured
in wells developed at great depth in a recharge area may underrepresent
the depth to water table, whereas the opposite is likely to be true in discharge
areas. In many settings, such water level data are sparse or lacking altogether
because the water table aquifers may not generally be employed as water
supplies. In addition, water levels in some unconfined aquifers are likely
to exhibit large seasonal or year-to-year fluctuations. For all these reasons,
therefore, much inferential evidence was used to estimate the depth to the
water table in many settings or parts therein. Such evidence includes configuration
of surface topography, overall position in the regional landscape, nature
and elevations of surface water bodies, and character of the underlying
sequence.
Water table elevation is likely to vary widely in some places,
especially in some of the more complex sequences where fine-grained units
that may hold up water movement and cause 'perched' conditions are common
at different or widely ranging depths in a sequence. Ideally, a depth to
water table map should be made by comparing surface topography to a map
of water table elevation. Unfortunately, water table maps are a rare commodity
in Indiana and are available only for a few scattered areas, and time did
not permit such maps to be constructed during this study. Moreover, surface
topography is not shown on several of the 1:100,000 base maps. In most such
cases, water table depth was essentially estimated by comparing landscape
characteristics from 1:24,000 topographic maps to existing water level data.
In short, definition of water table depth is problematic in many settings
at the present scale. Map units tend to be relatively large and there is
likely to be considerable variation in each of the contoured depth ranges
(table 5). Under no circumstances should these
maps be used as the sole source of information to determine water table
depth at a specific site.
It must be emphasized that water table conditions commonly exist
in many, if not most of the fine-grained mapping units as well. There are
virtually no water level data available for these units, and the potential
for large capillary forces in some of the more clayey units creates potentially
significant local differences. The prevalence of artificial drainage ditches,
an abundance of surface water features, and/or soil series characterized
by mottling, gleying, or similar features in the lower horizons are usually
good indications of a shallow water table in these units. It is not commonly
known whether the water table at any given place in a fine-grained capping
unit represents the true top of the zone of saturation. Very tight till
units are likely to be characterized by one or more perched water tables,
especially in upland settings, whereas the water table in surface tills
in lower-lying settings may represent the true zone of saturation. Inferred
water-table conditions associated with capping units are noted in the descriptions
of their respective settings.
Maps showing the depth and composition of the bedrock surface
were compiled in areas where the bedrock is 50 feet or less below the moder
land surface. Such areas are found in a wide variety of settings, most of
which are south of the Wabash River valley and west of the lower Tippecanoe
River valley. Many are associated with the major bedrock structural high
known as the Cincimati Arch, which crosses the state from southeast to northwest.
Much of the bedrock surface along this structural high is composed of limestone
and dolomite of Silurian and Devonian age, which constitute a major aquifer
system at many places.
The bedrock topography of Indiana is mapped at a scale of 1:500,000
using a 50-foot contour interval (Gray, 1982), and larger-scale work maps
from which the state map was compiled are available. A similar set of maps
showing bedrock geology (Gray and others, 1987) is also available. These
maps formed the basis for the depth-to-bedrock maps produced during this
study; the dimensions of areas of shallow bedrock shown on the older maps
were modified according to new data obtained during this study, as was the
bedrock geology at a few places.
The maps were constructed utilizing 25-foot contours to show bedrock
depth, and various symbols were added to map units to indicate the type
of bedrock (table 6). Identification of the existence
or shape of some areas of shallow bedrock was limited by the present map
scale, especially in those areas where the bedrock surface is typified by
much local relief. The problem was compounded on those quadrangles lacking
surface topography. In general, though, construction of depth to bedrock
maps proved to be the least problematic of the various contour maps. Bedrock
characteristics (where applicable) are summarized in the discussions of
individual settings.
Use of the Maps and Settings
The maps and descriptions of settings were compiled in less than
one year, and basically represent a reconnaissance of glacial-hydrogeologic
terrain in northern Indiana. Limitations of time and data did not permit
a leisurely and detailed examination of each terrain and setting; in many
cases, data were few and only a cursory analysis was possible. Consequently,
many of the terrains and settings presented herein are not well-documented,
at least not in the sense of understanding the full nature of variation
within, and how that variation might be distributed across different parts
of sequences. Undoubtedly, much better definition of glacial terrains could
be provided by additional regional- or county-scale studies of the type
recently completed by Bleuer and Woodfield (1993); Fleming (1992, 1994);
Brown (1994) and Fleming and others (1993). To that end, the results of
this study may help to highlight critical areas where additional work is
desirable or necessary to better resolve existing or anticipated environmental
problems. The more detailed studies could help to develop a better conceptualization
of the nature of hydrogeologic variation in many of the more diverse terrains
and settings, and may make further subdivision of these large units possible.
For all these reasons, the maps should never be used for site-specific purposes,
because individual sites may lie within a particular range of variation
that is outside of the 'mode' shown by individual map elements and described
under a particular setting. The settings should, however, be employed to
develop a regional context for these more site-specific problems and studies,
and may help to interpret where in the possible range of conditions a particular
site or area falls. The settings also provide a useful, graphic tool for
educating the public (and particularly agricultural chemical users and researchers)
about the nature of hydrogeologic environments and behavior.
The maps will be used for some manner of sensitivity analysis.
Sensitivity map units so derived must not in any case be of a larger scale
than the settings mapped herein or misleading results are certain to occur.
It is recommended that this sensitivity analysis be kept to the lowest common
denominator -that is the map units that result should be of a scale appropriate
to the smallest scale data that were used in the assessment. Therefore,
if the hydrogeologic data (scale 1: 1 00,000) are combined with other data
at a scale of 1:250,000, for example, then the latter scale should supercede
the former, and all other largerscale data. The temptation will be very
great to view the results at the largest scale, but this should be resisted
at all costs.
Second, it is strongly recommended that a point-based method ultimately
be implemented for the sensitivity assessment. Such a method will be readily
facilitated by the ongoing computerization of waterwell records being performed
at the Division of Water. A computerized database would support the necessary
sorting, coding, and manipulation of these records in order to characterize
the entire state in terms of SI values. The hydrogeologic settings defined
herein are a natural counterpart to a point-based method, and provide a
realistic conceptual framework for interpreting the SI ranges. Moreover,
a variety of powerful statistical analyses are possible with the output
from a point-based assessment, whereas no such analyses would be feasible
with the output of any known area-based method. Statistical treatment of
these data could be instrumental for many planning and management purposes,
such as:
1) identifying management areas for specific agricultural chemicals
based on quantitative analysis rather than on untested, regional
models.
2) analysis and interpretation of ground-water monitoring data.
The SI scores would facilitate application of several common statistical
techniques to analyse covariance; that is, to answer the questions 'is there
a relationship between sensitivity domains and the occurrence of agricultural
chemicals in ground water.?' and 'what are the nature of those relationships?'
3) the answers to #2 above would help to test the validity of
the sensitivity assessment, and thus either lend credence to the results
of the assessment or point out shortcomings that need to be remedied.
4) the ability to defend management decisions will be far better when those decisions are based on
quantifiable facts and conditions.
Future
The maps and descriptions of hydrogeologic settings represent
a first approximation of the overall hydrogeologic-terrain-framework of
the state. The process of compiling these documents emphasized certain shortcomings
in the availability and format of existing data structures, and it greatly
highlights the need for certain forms of hydrogeologic information that
are not readily available in the State of Indiana. In particular, a computerized
subsurface database (e.g., water-well records) would greatly facilitate
the display and mapping of hydrogeologic features and is probably
essential for performing a point-based sensitivity analysis or other similar
operations to statistically document trends and variance.
Workable and practical definitons of hydrogeologic settings depend almost entirely on two main pieces of information:
1) documentation of both broad characteristics of geologic terrain
and its local variation; and 2) an understanding of the three-dimensional
pattern of ground-water flow, with emphasis on identifying the locations
and magnitudes of recharge and discharge areas. Glacial terrains have received
little detailed scrutiny and documentation over the majority of the state.
Studies by Bleuer and Woodfield (1993), Brown (1994), Fleming (l992; 1994);
and Fleming and others (1993) are examples of countylevel studies that merge
the concepts of glacial terrain and hydrogeologic settings and that provide
abundant documentation of local conditions and the variation within. Few
areas elsewhere have been put into a terrain context that provides a sound
conceptual basis for understanding the relations of individual aquifers
and confining units to entire hydrogeologic systems. Much more could be
done with many of the broadly-defined terrains and settings identified in
this study, especially in parts of far northeastern Indiana and the central
till plain. Much greater local hydrogeologic variation than could possibly
be interpreted or described during the short time-frame of this study is
known to characterize these and other areas. Such documentation will become
increasingly necessary as the management of agricultural chemicals and other
point- and nonpoint contamination sources becomes more geographically focussed.
Documentation of ground-water flow patterns is absolutely essential
for characterizing the position in the flow system of particular aquifer
systems and hydrogeologic settings. An understanding of position in flow
system is in turn necessary to successfully accomplish a wide variety of
endeavors, such as installation and interpretation of results from monitoring
networks; prioritizing areas for monitoring, remediation, and protection;
and meaningful characterization of aquifer sensitivity. This information
can be derived from a combination of maps, cross-sections, and block diagrams
illustrating the three dimensional distribution of hydraulic head in large
hydrageologic systems. There are few existing maps in Indiana that show
the configuration of the water table or piezometric surfaces of ground-water
basins, and fewer still that are tied to specific aquifers or aquifer systems.
Currently, the most extensive of such illustrations are contained in large-scale
water resource assessments of four river basins prepared by the Division
of Water (Beaty and Clendenon, 1988 and 1990; Beaty and Gosine, 1994; Clendenon
and Beaty, 1987), which show 'composite' piezometric surfaces for certain
types of aquifers within those basins; whereas potentiometric surface/water
table maps tied to more specific aquifer systems are available for two counties
(Fleming, 1994; Fleming and others, 1993). Development of water level maps
at useful scales and tied to specific aquifer systems may be the single
greatest need that currently exists for characterizing hydrogeologic systems
in the state. Ideally, water level data should be tied to specific sequences
and interconnected systems, thus the development of these illustrations
is closely interrelated to further characterization of terrains. However,
a true water-table map could be constructed for the entire state at the
present time with little additional mapping of sequences or aquifer systems.
Such a map, or maps, include the water table in flne-grained surface sediments, such as Till, in addition to that in unconfined sand and gravel or bedrork aquifers. All near-surface derived contamination will first encounter the water table, regardless of the nature of its host medium, and the configuration of the phreatic surface will determine the immediate disposition (i.e., direction of migration-lateral, upwards, downwards) of those contaminants. Consequently the compilation of water-table information is crucial to many activities related to the characterization, protection, and remediation of ground water.
Table of Prelminary Monitoring
Networks and Hydrogeologic Settings.
Return to:
email: hahnl@hahn.isco.purdue.edu