A STUDY OF SUBSURFACE WATER FLOW IN A SOUTASTERN MINNESOTA KARST DRAINAGE BASIN A THESIS SUBMITTED TO THE FACULTY OF THE GRADUATE SCHOOL OF THE UNIVERSITY OF MINNESOTA BY ERIC HERBERT MOHRING IN PARTIAL FULFILLNT OF THE REQUIRENTS FOR THE DEGREE OF MASTER OF SCIENCE AUGUST 1983
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
A STUDY OF SUBSURFACE WATER FLOW
IN A SOUTHEASTERN MINNESOTA KARST
DRAINAGE BASIN
A THESIS
SUBMITTED TO THE FACULTY OF THE GRADUATE SCHOOL
OF THE UNIVERSITY OF MINNESOTA
BY
ERIC HERBERT MOHRING
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF
MASTER OF SCIENCE
AUGUST 1983
By June our brook's run out of song and speed. Sought for much after that, it will be found Either to have gone groping underground Or flourished and come up in jewel-weed, Weak foliage that is blown upon and bent Even against the way its waters went. Its bed is left a faded paper sheet Of dead leaves stuck together by the heatA brook to none but who remember long. This as it will be seen is other far Than with brooks taken otherwhere in song. We love the things we love for what they are.
- Robert Frost
ABSTRACT
The Root River drainage basin in Fillmore county, southeastern
Minnesota has well developed karst topography and karst groundwater flow
in carbonate sedimentary rocks of upper Ordovician age. In the upper
carbonate aquifer subsurface water flows rapidly through solution
enlarged fractures and conduits, and is intimately connected to surface
water. As such it is very susceptible to pollution.
An area was chosen in the drainage of the South Branch of the Root
River, southeast of the town of Spring Valley, for detailed hydrologic
study. The area has one of the highest densities of karst features in
Minnesota.
The first part of the study involved quantitative fluorometric
tracing of subsurface water using the fluorescent dye Rhodamine WT and a
field fluorometer. The tracing studies delineated subsurface flowpaths,
revealed travel times and dispersion along the flowpaths, and permitted
mass balance calculations of inflowing and outflowing water. The traces
defined the recharge area of Moth and Grabau Springs at the head of
Forestville Creek, an important trout stream.
A gauging station was installed to measure the discharge of these
springs, and so far has produced two years of continuous record. A net
work of rain gauges was installed to measure precipitation over the
recharge area. Data from these installations describe the way the karst
system responds to recharge events. Several sub-environments of flow
exist within the aquifer. Initial estimates of transmissi vi ty and
aquifer diffusivity can be derived from the data.
-i-
ACKNOWLEDGMENTS
This research was supported by grants from the Legislative
Commission on Minnesota Resources to the Minnesota Geological Survey.
Financial support was also provided by a University of Minnesota
Institute of Technology Corporate Associate Fellowship, and by a Lando
Fellowship.
First of all I would like to thank Calvin Alexander for his
unceasing support and guidance over the last few years, and for many
productive and lively discussions during all phases of this project.
This study would not have been possible without the previous mapping and
subsurface water tracing work of Ron Spong, Ramesh Venkatakrishnan, and
many others. I am grateful to all those who assisted me in the field in
various ways, including Bridget O'Brien, Warren Beck, Paul Book, Phil
Hewitt, Dave Rogers, Stephen Mohring, Mike Mccrum, Kate Mader, Shiela
Grow, and Barb Silverman. I wish to thank the United States Geological
Survey Water Resources Division in St. Paul, and especially Jerry Hicks,
for loaning a water level recorder and assisting in the construction of
the gauging station.
I would like to thank the residents of Fillmore county who helped in
this project. Thanks to Niel Davie for giving me access to Mystery
Cave. Thanks to Mike Hellerud for operating the recording rain gauge,
and to other residents around Spring Valley for faithfully keeping daily
precipitation records. Thanks to Ken Hadland for lodging and for the
use of his house as a field base station. Finally, special thanks to
the Root family for letting me tromp on their land at all hours of the
day and night, for permitting me to construct the gauging station on
-ii-
their land, and for all-around hospitality.
-iii-
TABLE OF CONTENTS
Page
ABSTRACT •••••••••••••••••••••••••••••••••••••••••••••••••••••••••• i
ACKOWLEDGMENTS •••••••••••••••••••••••••••••••••••••••••••••••••••• ii
TABLE OF CONTENTS •••••••••.••••••••••.•••.••...••.....•••....•••.. iv
LIST OF FIGURES ..•••....••.•.•...•.•..•.•••.••••••.••••.•..••..••• vii
I • INTRODUCTION ••••••••••••••••••••••••••••••••••••••••••••••••• 1
II. GEOLOGICAL SETTING AND DESCRIPTION OF STUDY AREA ••••••••••••• 6
There are many sinkholes in the upland plain north of Fairview Blind
Valley. One of the larger of these is Red Tail Sink, which accepts flow
from surficial seeps to the west (Fig. 5). Another major feature is
Lefevre Blind Valley, where an intermittent stream flowing from the
southwest sinks (Fig. 5). Natural Well is a small cave in the Prosser
Member of the Galena Formation (Fig. 10). There are many other karst
features in the study area, but the ones described above are most rele
vant to the hydrologic studies of this paper.
Figure 10. Map and cross-section of Natural Well. Adapted from an ori
ginal map prepared by R.C. Spong.
WELL HOUSE ,r:: ENTRANC& (Apprex.el4evflt1on) .tt!~ 12.i 5 .. T. MSL ....... __ _
±
GAlENAFM PROSSER MOO (ls)
~ ' ~· 19E,5
MA'1,. l)l!CLIN.
NATURAL WELL fllllll\ORE CO., MINNESOTA
SURVE.'< 21 Oct 1977 BY T.MARSHALL,R. VEN'<ATh
KRISHNAN ( R.SPONG. CAlttOGRAPHV R.SPONG.
BASE: USGS WVKOFF7!i.
SCALE
~ •5z -~,0 FT. o I :J. 3 M.
I N 0 I
-21-
III. QUANTITATIVE FLUOROMETRIC DYE TRACING
Tracers of various kinds have been used for many years in the study
of karst groundwater flow. A tracer is introduced into a sinkhole or
sinking stream and detected or recovered at one or several spring
resurgences. Fluorescent dyes, minute biological materials such as
Lycopodium spores or bacteria, soluble salts, and radioactive materials
are examples of tracers used in various studies (Aley and Fletcher,
1976). An ideal tracer should have the following properties: 1) its
introduction into the subsurface should be easy; 2) it should be easily
detectable or recoverable at the resurgence points; 3) its flow velocity
should approximate that of its host water; 4) its adsorption onto the
aquifer material should be minimal; and 5) its potential for harmful
environmental effects should be small.
The fluorescent dye Rhodamine WT, used in this study in conjuction
with a field fluorometer, adequately satisfies these requirements. The
fluorometer (a Turner Designs model 10-005) is capable of accurately
measuring dye concentrations down to a fraction of a part per billion.
The advantage of the fluorometric technique is that the concentration of
dye as it emerges from a spring resurgence can be monitored, yielding
quantitative travel time and dispersion data.
Rhodamine WT can also be used semi-quantitatively by placing a
packet of activated charcoal in a spring resurgence. If dye emerges
from the spring, it will adsorb onto the charcoal. The packet can then
be collected and analyzed for dye.
-22-
Methods
The basic procedure used in this study was as follows:
1. An input point (sinkhole or sinking stream) and resurgence points
(springs) were chosen.
2. Packets of activated charcoal were placed in all the suspected
resurgences, since it was not possible to monitor every spring with
the fluorometer.
3. A known quantity of Rhodamine WT was injected at the input point.
4. Discharge measurements of the spring(s) were taken. At first this
was done by measuring the cross-sectional area of the stream and
measuring the velocity of floating or suspended objects at intervals
along this cross-sectional plane. After the first several traces,
flow measurements were done using a dye dilution technique. Dye of
known concentration C0 was injected in the spring at constant rate
q, using a Mariette vessel constant head injector device (see Cobb,
1968). After equilibrium was established, the downstream con
centration C was measured with the fluorometer. The discharge Q was
then determined using the formula:
Q = q ( 1) c
5. The concentration of the dye as it emerged from one spring or
several springs was monitored for as long as possible.
6. The final step on leaving the field was to change the charcoal
packets in all the springs in order to catch any dye that might not
-23-
have been detected and to measure long-term washout. The charcoal
packets were analyzed for Rhodamine using the technique of Aley and
Fletcher (1976).
Data Analysis
The fluorometric data were plotted as concentration versus time
curves. These pulse response curves contain information about the tra-
vel time and dispersion of the subsurface water. Also, the area under
such a curve is proportional to the amount of dye discharged from the
spring:
(X)
m = JQ(t)c(t)dt
0
where m = mass of dye recovered c(t) = dye concentration in spring Q(t) = spring discharge
(2)
If spring discharge is constant over the period of the trace (a tenuous
assumption), Q can be taken outside of the integral. Using this rela-
tionship it is possible to perform mass balance calculations, thereby
determining what proportion of the water entering a given sink point
emerges at a given resurgence. When several springs are being moni-
tored, it is possible to determine how the water entering a given sink
is partitioned among several resurgences.
The concentration versus time curves were integrated numerically,
using a simple trapezoidal formula. The tails of the curves were extra-
polated using an exponential decay rule, and the areas under the extra-
polated tails were determined analytically. The tails contained a
-24-
significant proportion of the total area. The total area under each
curve was multiplied by the measured spring discharge to determine the
mass of dye recovered (equation 2). This was compared to the amount of
dye injected to derive a percentage dye recovery. The greatest uncer
tainties in this analysis are in the discharge measurements.
Results
Figure 5 shows all the quantitative fluorometric dye traces
accomplished in the study area. The traces will be discussed one by
one, with most of the information appearing in figures. The data from
the dye traces is in Appendix A.
Disappearing River to Seven Springs, Oct • .!.Q..z. 1979 (Figure 11)
This was the first of several traces to Seven Springs. Dye was
introduced at the commercial entrance to Mystery I Cave, where the South
Branch first starts losing water to the subsurface. The flow conditions
were relatively low, with the South Branch going completely underground.
It is interesting that the various outlets at Seven Springs, which are
only a few meters apart, responded differently from one another. The
response at outlet /11 peaked at a lower concentration and was more
dispersed than the responses of the other outlets. When the curves are
averaged, the resulting curve represents an 89.5% dye recovery. Given
the uncertainty of the flow measurements and the inaccuracies introduced
by a simple averaging of the pulses, this represents essentially
complete dye recovery.
Figure 11. Disappearing River to Seven Springs, Oct. 10, 1979 Time of dye drop: 3:13 a.m., October 10, 1979, at Mystery I
Cave entrance. Amount, type of dye: 914 g Rhodamine WT 20% solution Horizontal distance: 2.38 km (1.48 miles) Vertical drop: 18 m (60 ft) Arrival of leading edge of dye pulse: 7.75 hours Arrival of peak dye concentration: 10 hours Flow measurements: 259 liters/sec (measured with tape and
stopwatch) Dye budget calculations: 89.5% of dye recovered Springs where dye did not emerge: Moth Spring, Grabau
Spring, Cold Spring The numbering of the outlets is shown in Figure 7.
8 10 12 14 16 28 30 32 40 40
35 SEVEN SPRINGS 35
..... 30 I- • ro - -i 30 m a. a. ---~ 25· I- .. 1....:-· 1 ' \: -i 25 .... c .... .... c
Disappearing River to Seven Springs, Sept. £!. 1980 (Figure 12)
The flow conditions were much higher during this trace than for the
previous trace from Disappearing River to Seven Springs, with the South
Branch flowing all the way along its surface channel. The spring
response differed markedly from that of the first trace. The responses
for the different outlets form a nestled family of curves, with the
higher numbered outlets showing the highest concentrations. All curves
show the same dispersion through time, and all have a second peak with a
lag time of 3 hours after the first peak.
The double-peaked response curves can be best explained by a
divergence in the flowpath, followed by a later reconvergence, with dif
ferent travel times associated with each pathway. The fact that the
double peak occurs under high flow, but not under low flow conditions
suggests that the alternate pathway is an overflow channel available
only under high flow conditions.
The nestled pattern of the response curves is best explained by
varying degrees of dilution of essentially the same pulse by undyed
water from another source. The diluting water could be sinking in
streamsinks in the South Branch downstream from Disappearing River, such
as those between Matheson Sink and Seven Springs (see Fig. 5), and pre
ferentially travelling to the lower numbered (northernmost) outlets of
Seven Springs. Under high flow conditions, when water is flowing all
the way along the surface channel and sinking into these downstream
sinks, the response curves show increasing dilution from high to low
numbered outlets at Seven Springs. Under low flow conditions, there is
little or no water infiltrating in these sinks, so the response curves
Figure 12. Disappearing River to Seven Springs, Sept. 2, 1980 Time of dye drop: 6:00 a.m., September 2, 1980, at
Mystery I Cave entrance Amount, type of dye: 266.3 g Rhodamine WT 20% solution Horizontal distance: 2.38 km (1.48 miles) Vertical drop: 18 m (60 ft) Arrival of leading edge of dye pulse: 6.75 hours Arrival of peak dye concentration: 8.25 hours Flow measurements: Seven Springs - 309 liters/sec (measured
by dye dilution) The numbering of the outlets is shown in Figure 7.
-27-
!?
.....
'° ...
ID Cit > a: D c: • ... Cl • Q. Q. Cl ,,, • Q
c: ... .,, ~ • Cl) Q.
Q. ~ 0
~ ... Q
Q: • ~ >-
Q
~ ... •
2 -.. ~ c '4.a • Cl) ...
:II
~ 0
%
N CID
-28-
do not show as pronounced a pattern of increasing dilution from high to
low numbered outlets.
Formation Route Creek to Seven Springs, Aug. E2l. 1980 (Figure 13)
Formation Route Creek flows through the northwestern portion of the
Mystery Cave system, or "Mystery III Cave". Dye was injected beneath a
room known as the "First Triangle Room", where the creek is accessible
(see Fig. 6). Water was flowing all the way along the surface channel
of the South Branch at the time.
The pulse response curves are nestled as in the second trace from
Disappearing River to Seven Springs, except that there is no second
peak. Again, the lower numbered (northernmost) outlets had the lowest
concentration. This further supports the idea that the lower numbered
outlets are preferentially fed by a source not associated with the
Mystery Cave system, such as the sinks on the South Branch between
Matheson Sink and Seven Springs.
Matheson Sink to Seven, Moth, and Grabau Springs, Oct. 20-21, 1979 (Figure 14_)_
In this dye trace water was simultaneously traced from Matheson Sink
on the South Branch to Seven Springs, Moth Spring, and Grabau Spring,
and all three pulses were monitored. Under low flow conditions,
Matheson Sink can act as the terminal sink of the South Branch before it
resurges at Seven Springs.
The dye pulse first arrived at Seven Springs. The response was
similar to that of the first (low flow) trace from Disappearing River to
Seven Springs. The pulse at outlet #1 had a slightly lower peak and was
slightly more dispersed than at the other outlets.
Figure 13. Formation Route Creek {Mystery Cave) to Seven Springs, August 25, 1980 Time of dye drop: 10:19 a.m., August 25, 1980, under First
Triangle Room in Mystery III cave Amount, type of dye: 309.3 g Rhodamine WT 20% solution Horizontal distance: 1.63 km (1.02 miles) Vertical drop: 6 m (18 ft) Arrival of leading edge of dye pulse: 5.0 hours Arrival of peak dye concentration: 6.25 hours Flow measurements: Seven Springs - 220 liters/sec (measured
by dye dilution) Springs where dye did not emerge: Moth and Grabau Springs The numbering of the outlets is shown in Figure 7.
-.a a. Q, -
I&
-29-
SEVEN SPRINGS
15 Houn After Dye Dropped In Formation Route Creek
Figure 14. Matheson Sink to Seven Springs, Moth Spring and Grabau Spring, October 20-21, 1979 Time of dye drop: 6:35 a.m., October 20, 1979 Amount, type of dye: 1118 g Rhodamine WT 20% solution Horizontal distance: Seven Springs - 1.72 km (1.07 miles)
Grabau Spring - 3.86 km (2.40 miles) Moth Spring - 3.82 km (2.37 miles)
Vertical drop: Seven Springs - 12 m (40 ft) Grabau Spring - 26 m (85 ft) Moth Spring - 26 m (85 ft)
Arrival of leading edge of dye pulse: Seven Springs - around 8.5 hours Grabau Spring - 13.17 hours Moth Spring - 18.67 hours
Arrival of peak dye concentration: Seven Springs - 10.17 hours Grabau Spring - 16.58 hours Moth Spring - 23.67 hours
Flow measurements: Seven Springs - 242 liters/sec Grabau Spring - 157 liters/sec Moth Spring - 381 liters/sec (measured with tape and stopwatch)
Dye budget calculations: Seven Springs - 5.8% (4.5%) of dye Grabau Spring - 34.3% (26.6%) of dye Moth Spring - 89.0% (68.9%) of dye (numbers in parentheses have been scaled down to
100% dye recovery) Springs where dye did not emerge: Cold Spring The numbering of the outlets of Seven Springs is shown in Figure 7.
20
m t 15 I-~
c: 0 ....
I c .... .... c: Q)
0 c: 0
u 10 I-I-3: Q)
c: -E c
"C 0
.s:: a::
5 I-
10
;:)t:.Vt:.N
15
GRABAU SPRING
I
I I
I
\
20
I
\ j \ J
I '\..
25
MOTH SPRING
,
\ \
\
30
20
--l 15
I -I 10
-I 5
IY ·~ I -- I 0 , f* , e , 0
10 15 20 25 30 Hours After Dye Injection At Matheson Sink
I lN 0 I
-31-
Most of the dye emerged at Moth and Grabau Springs. The subsurface
connection between the South Branch of the Root River and Forestville
Creek was suggested in 1958 when souvenirs from the Mystery Cave con
cession stand, which had been washed away in a flood, were found below
Moth Spring. It is surprising that the Moth Spring response curve lags
behind that of Grabau Spring by almost six hours. Not only is Grabau
Spring farther away from Matheson Sink than Moth Spring, but it is
across the deeply incised stream valley of Forestville Creek (see Figs.
5 and 8). The discharge from Moth Spring is also much greater than that
from Grabau Spring. A partial explanation for the surprising travel
times lies in the nature of the spring openings. Inside the Moth Spring
cave entrance is a large pool which overflows to produce the spring
outflow.
outlets.
The output from Grabau Spring comes from isolated small
The mixing time for the large reservoir inside Moth Spring
cave could account for some of the increased dispersion and travel time
for the dye pulse. However the peak concentration of the Moth Spring
response is greater than that of Grabau Spring, so there is more going
on than simple lagging and routing through a reservoir. Whatever the
mechanism, the connection to Grabau Spring is shorter and involves less
dispersion than that to Moth Spring.
The mass balance calculation for this trace produced a 129.1% dye
recovery (Fig. 14). This is clearly impossible, and due no doubt to
inaccurate flow measurements. Figure 14 also gives numbers wich have
been scaled down to 100% recovery. It is interesting that over 90% of
the water which infiltrates at Matheson Sink gets pirated off to another
drainage.
-32-
Fairview Blind Valley to Moth and Grabau Springs
Two traces were done from Fairview Blind Valley, one under high flow
and one under low flow conditions. As mentioned before, under high flow
conditions the stream flows all the way to the end of the valley,
finally disappearing in a collection of sinks known as the Hellerud
Streamsinks (Fig. 9). Under low flow, the stream makes it only as far
as the Poldervaard Streamsinks.
High Flow Hellerud Sink 'A' to Moth and Grabau Springs, Nov. 17-21, 1979 (Figure 1~ - - -- -
The terminal sink on November 17, 1979 was identified as Hellerud
Sink 'A' to distinguish it from the other Hellerud Streamsinks (see Fig.
9). The response curves show several interesting features. Again, the
Moth Spring response lags behind the Grabau Spring response, this time
by two hours. Notice the similarity in shape of the two curves, and the
pronounced double peak in both. Some of the water takes a faster path,
producing the first peak, while most of the water takes a second, slower
path, producing the main, larger peak. The similarity of the two curves
indicates that the divergence and reconvergence occur before the final
split between the flows of Moth Spring and Grabau Spring.
Low Flow Poldervaard Sink 'A' to Moth and Grabau Springs, July 13-22, 1980 (Figure 16)
The terminal sink on July 13, 1980, was one of the Poldervaard
Streamsinks, and was identified as Poldervaard Sink 'A' (see Fig. 9).
Note that the travel time is twice that from the downstream (Hellerud)
sink of Fairview Blind Valley. The Moth Spring response lags by 6-9
Figure 15. Fairview Blind Valley (Hellerud Sink A) to Moth and Grabau Springs, November 17-21, 1979 Time of dye drop: 3:30 p.m., November 17, 1979 Amount, type of dye: 863 g Rhodamine WT 20% solution Horizontal distance: Moth Spring - 5.64 km (3.50 miles)
Grabau Spring - 5.66 km (3.52 miles) Vertical drop: 47 m (155 ft) Arrival of leading edge of dye pulse:
Grabau Spring - 40.25 hours Moth Spring - 42.25 hours
Arrival of peak dye concentration: Grabau Spring - 65.5 hours Moth Spring - 67.5 hours
Flow measurements: Grabau Spring - 153 liters/sec Moth Spring - 1281 liters/sec (measured with tape and stopwatch)
Dye budget calculations: Grabau Spring - 12.5% (10.9%) of dye Moth Spring - 102.4% (89.1%) of dye (numbers in parentheses have been scaled down to
100% dye recovery) Springs where dye did not emerge: Cold, Stagcoach, Freiheit,
Barr, Narcissus, Mahood Springs
40 45 50 55 60 65 70 75 80 85 90
[ 2,0 ~ ~ \ MOTH ~ 2.0 SPRING
GRABAU SPRING
c 0 .... Cl .... ....
L I/ \\ J I c l:..N
G) l:..N () I c 0 u I- 1.0 ~ r r \ \ -l 1.0 3,:
G)
c E Cl
"Cl 0
.s:::. 0::
40 45 50 55 60 65 70 75 80 85 90 Hours After Dye Injection At Fairview Blind Valley
Figure 16. Fairview Blind Valley (Poldervaard Sink A) to Moth and Grabau Springs, July 13-22, 1980 Time of dye drop: 6.36 p.m., July 13, 1980 Amount, type of dye: 1203 g Rhodamine WT 20% solution Horizontal distance: Moth Spring - 5.96 km (3.70 miles)
Grabau Spring - 5.97 km (3.71 miles) Vertical drop: 50 m (165 ft) Arrival of leading edge of dye pulse:
Grabau Spring - 92.5 hours Moth Spring - 98 hours
Arrival of peak dye concentration: Grabau Spring - 111.4 hours Moth Spring - 120.4 hours
Flow measurements: Grabau Spring - 70 liters/sec Moth Spring - 332 liters/sec (measured by dye dilution)
Dye budget calculations: Grabau Spring - 15.2% of dye Moth Spring - 74.5% of dye
Hours After Dye Dropped at Poldervaard Sink (Fairview Blind Valley Upstream Sink)
I
180 190 200 210
I (.N ~ I
-35-
hours, and there is no pronounced double peak.
The difference between the high flow and low flow response curves
are best explained by the level of water in the aquifer. Under high
flow conditions, the large conduits are more nearly full, and water is
flowing at a higher velocity. There are alternate overflow channels
available, which can produce double-peaked response curves as in Figure
15. In the case of Figure 15, most of the water travelled the slower of
the alternate pathways, while a smaller amount took a II short cut".
Under low flow conditions, the main conduits are not full, so the flow
is slower, and overflow passageways are not available. The same pattern
shows up at Seven Springs: high flow conditions produce a double peak,
while low flow conditions produce a single-peaked response.
Figure 17 gives a simplified, schematic cross section of the flow
between Fairview Blind Valley and Moth and Grabau Springs. The figure
does not by any means represent all the complexities of the system. It
simply shows the kind of conduit geometry which could produce the
observed results. Moth Spring is shown as being farther away from
Fairview Blind Valley than Grabau Spring. This is to show that the
distance along the flowpath is longer, and does not reflect the areal
geometry.
Natural Well to Moth and Grabau Springs, May 12-17, 1980 (Figure 18)
Natural Well is a small cave in the Prosser Member of the Galena
Formation (Fig. 10). A small stream of water sinks into the floor of
the cave, and this is where dye was injected.
As in other traces to Moth and Grabau Springs, the Moth Spring
response lags behind the Grabau Spring response. There are some unusual
Figure 17. Schematic cross-section of the subsurface flowpath between
Fairview Blind Valley and Moth and Grabau Springs, showing
one kind of conduit geometry which would produce the
observed dye trace response curves.
--Q
> "O c: ·-·-> .. -
-36-
Figure 18. Natural Well to Moth and Grabau Springs, May 12-17, 1980 Time of dye drop: 4:00 p.m., May 12, 1980 Amount, type of dye: 404.8 g Rhodamine WT 20% solution Horizontal distance: 1.83 km (1.14 miles) Vertical drop: 17 m (55 ft) Arrival of leading edge of dye pulse:
Moth Spring - 55 hours Grabau Spring - 51.5 hours
Arrival of peak dye concentration: Moth Spring - 66.5 hours Grabau Spring - 61.0 hours
Flow measurements: Moth Spring - 421 liters/sec Grabau Spring - 92 liters/sec (measured by dye dilution)
Dye budget calculations: Moth Spring - 44.1% of dye Grabau Spring - 9.9% of dye
Springs where dye did not emerge: Root Spring
-37-
0 N
a ... ::, -a z c
,:,
• Q, Q,
0 ... Q
... • -... ct
-38-
oscillations near the peaks of both curves. The response curves cannot
account for all of the dye, suggesting that some of the water
infiltrating at Natural Well might be emerging elsewhere. Root Spring,
located 1 km northeast of Moth Spring, showed no positive response.
Red Tail Sink to Moth and Grabau Springs, June 28 - July 2.t_ 1980 (Figure 1gy- - -- -
Red Tail Sink is a large sinkhole north of Fairview Blind Valley,
which drains the outflow from several surficial seeps when they are
flowing. The sink is near the topographic divide between the drainages
of the Middle and South Branches of the Root River.
The travel time to Moth and Grabau Springs is relatively long and
the response curve is fairly dispersed. While the dye recovery is less
than complete, dye was not detected at any neighboring springs. The
characteristic lag of the Moth Spring response is present.
Lefevre Blind Valley to Moth and Grabau Springs, October 7-10, 1980 (Figure 20)
Lefevre Blind Valley is a major sink draining an intermittent
stream. When flowing, the stream drains into several streamsinks at the
end of the valley, and this is where dye was injected.
The response curves were fairly typical for Moth and Grabau Springs.
Data from the initial rise of the curves is missing, and was approxi-
mated by an exponential rise for the dye budget calculations.
Root River Dye Traces
The Root River dye traces were large scale dye trace experiments
performed in August of 1981 and August of 1982. In each case, a large
pulse of dye was injected into the South Branch south of Spring Valley
Figure 19. Red Tail Sink to Moth and Grabau Springs, June 28-July 9, 1980 Time of dye drop: 10:10 a.m., June 28, 1980 Amount, type of dye: 1325 g Rhodamine WT 20% solution Horizontal distance: 6.69 km (4.16 miles) Vertical drop: 58 m (190 ft) Arrival of leading edge of dye pulse:
Grabau Spring - around 77 hours Moth Spring - around 81 hours
Arrival of peak dye concentration: Grabau Spring - 99.5 hours Moth Spring - 102 hours
Flow measurements: Grabau Spring - 82 liters/sec Moth Spring - 455 liters/sec (measured by dye dilution)
Dye budget calculations: Grabau Spring - 12.6% of dye Moth Spring - 75.3% of dye
Springs where dye did not emerge: Mahood, Barr, Narcissus, Freiheit, Stagecoach, Root, Cold, Seven, Frost, and Bartsch Springs
Figure 20. Lefevre Blind Valley to Moth and Grabau Springs, October 7-10, 1980 Time of dye drop: 5:00 p.m., October 7, 1980 Amount, type of dye: 895.4 g Rhodamine WT 20% solution Horizontal distance: Moth Spring - 4.65 km (2.89 miles)
Grabau Spring - 4.68 km (2.91 miles) Vertical drop: 52 m (170 ft) Arrival of leading edge of dye pulse: missed exact arrival
time Arrival of peak dye concentration:
Grabau Spring - 46.33 hours Moth Spring - 48.5 hours
Flow measurements: Grabau Spring - 90 liters/sec Moth Spring - 512 liters/sec (measured by dye dilution)
Dye budget calculations: Grabau Spring - 8.0% of dye Moth Spring - 47.2% of dye
5
-.a
h MOTH a. a.
SPRING -~ 4 GRABAU -ID SPRING ... -c • u
II \ \ i ~ 31 I ..p. 0 I
.... ~
• c ·- 2 E ID 'a 0 .I: II:
I
40 45 50 55 60 65 70 Hours After Dye Dropped In Lefevre Blind Valley
-41-
over a period of several hours. The pulse was monitored as it travelled
down the river, and as it emerged from Seven Springs, Moth Spring, and
Grabau Spring. Residential wells in the area were also monitored. The
purpose of the traces was to determine the potential flowpath of toxic
chemicals from the Ironwood Sanitary Landfill, a facility which had
illegally received some hazardous waste.
To provide a complete description of these dye traces is beyond the
scope of this paper. I will present only the response curves for Seven
Springs, Moth Spring, and Grabau Spring, which are a small fraction of
the total data set.
The distinguishing feature of these experiments is that they traced
the entire flow of the South Branch of the Root River, rather than the
small amount which enters one particular sink. The spring responses
reflect the combination of pulses produced as dye travelling down the
South Branch sequentially entered strearnsinks in the river bed.
August 1981 Root River Dye Trace - High Flow
The dye, 50 pounds of Rhodamine WT 20% solution, was injected over a
five hour period from 10:20 a.m. to 3:20 p.m. on August 1, 1981. It was
diluted to about 450 liters of 1% solution and siphoned into the river
at about 1.5 liters/min. Figure 21 shows the injection points for both
the 1981 and 1982 Root River dye traces. The river was flowing at about
600 liters/sec and the Rhodamine WT was immediately diluted to about
400 ppb.
From about 2:00-3:00 p.m., a heavy thunderstorm dropped over an inch
of rain over the study area. Etna Creek, which empties into the South
Branch 2 km below the injection point, flash flooded. High water con-
Figure 21. Map of the study area showing the dye injection points for
the 1981 and 1982 Root River dye traces and the location of
Ironwood Sanitary Landfill.
\
\ \
\. ·, \
,,.. .. - .. _ . .r<J./
·., ~
{ \ ,..
\ I
·,. ·,
/·'-. . ...,
r·· ·...... ·r ·-··,..··-··-· \
/ / ./ ..... · /
/
~ .. ./
r .,.
i
·-··- i
\
'-· \
i ( ..... i
/
~ LANDFILL\ ~ .
,..(7
l'.lo~
,..,.1·
/',,... (. ) i
{
/ i
(
/ j
/
!' ... -./ (
I
,, (
< "
rZ. \. .......
Red Tail
·,. ·,
I
\.) i ·.'l- .. ,.
\
' ·,
/
j'
\ ·,. :r··-...
· ........ ·,.
\ \ ·,.
( >(>- '··-'~- ~~, f)
j
\
'· \ /
·,.
EXPLANATION
..,., spring
-3- streamsink
~ cave entrance
+- dye trace connection
.i. gaging station
® recording rain gage
scale 0 .5 1mile
0 .5 ~
-.. ,··-... .-.·- .~ ........ ,..,.,,,) ·,
1.
I .I
.'.
1 km
\··-··.--··-··-,
!/~.,.,.·-··-··: /
,«:
N
t
I ..,. N I
-43-
di tions prevailed for the duration of the experiment, with the South
Branch flowing continuously in its surface channel from Matheson Sink to
Seven Springs.
The response curves at Seven Springs (Fig. 22) have higher peak con
centrations and lower peak dispersion for the lower numbered outlets, a
markedly different response than that of traces from Disappearing River
and Formation Route Creek under high flow conditions (Figs. 12 and 13).
The response is consistent with the earlier interpretation that sinks on
the South Branch between Matheson Sink and Seven Springs preferentially
feed the lower numbered outlets. In this case, the sinking water was
augmenting rather than diluting the responses at the lower numbered
outlets since the entire flow of the South Branch was being traced. The
response curves are more dispersed at the higher numbered outlets
because the water had to travel a longer distance underground.
The response at Moth and Grabau Springs is fairly typical (Fig. 23).
The response curves are similar in shape, with that of Moth Spring
lagging by about an hour, and having a higher peak.
August 1982 Root River Dye Trace - Low Flow
The dye, 50 pounds of Rhodmine WT 20% solution, was injected from
8:45 AM to 9:10 PM on August 7, 1982. Conditions were very dry, with
the South Branch going completely underground just downstream from
Matheson Sink.
The Seven Springs response curves (Fig. 24) are similar to those
of the low flow dye trace from Disappearing River to Seven Springs. The
outlet /11 response has a slightly lower peak and is slightly more
dispersed than that of the higher outlets. "Outlet /10 11 , a small,
Figure 22. Seven Springs response curves for the 1981 Root River dye
trace (high flow conditions).
CD u 0 !'.....
E-i
CD ::T)
D
!'..... CD > _ _)
0::::
~ 0 0
0::::
D n
(f)
m c
_ _)
!'..... 0....
en c 0)
> 0)
en
LI)
0J
-44-
l
I
i I I :
# Lfil 9£1
c.__ ....____ --------- ----- -
0 (\J
--~~
0 Lr) ...... D
Q)
-i-> 0
0
_...)
0) :J m :J
a:
Figure 23. Moth and Grabau Springs response curves for the 1981 Root
River dye trace (high flow conditions).
7
I (\
t 61 I \. 1981 Rool RLver O~e Trace
I \ c 5 0
I \ •. .J .+J 0
I \ L .,_) 4 c
I
\0 (I)
Molh Sprlng 0 I I
c I
..,. (.11
0 I
u '3
I Grabau S~~
f:-; ---.3:
(I) I \ c .. .J 2
I \ E 0 "'( u
t 0 ..c ~ Cl:: 1 /J
Ou 2 3 4 5
Augusl 1981 Dale
Figure 24. Seven Springs response curves for the 1982 Root River dye
1----------------------------------:;~-------------------'-o C __ _;)
8 0 - -
m E
_..:,
+>
-78-
Q = 1.966 m3/sec
l:o.sc = 0.0036 m (26)
2.3 1.966 T = = 99.7 m2/sec
4,r 0.0036
For sub-regime 2:
Q = 0.85 m3/sec
6Sc = 0.086 m (27)
2.3 0.85 T = = 1.8 m2/sec
4,r 0.086
For sub-regime 3:
Q = 0.483 m3/sec
l:o.sc = O. 109 m (28)
2.3 0.483 T = = o.81 m2/sec
4,r o. 109
There are several problems with this analysis. Dividing the incre
mental change in volume remaining by the recharge area probably
underestimates the drawdown by a considerable amount, thus overesti
mating the transmissivity. A piezometer well in the recharge area would
have solved this problem, and also would have permitted the deter
mination of the storage coefficient S from the Theis equation. Another
problem is the term 4 'TT in the expression for transmissi vi ty. This
comes from assuming 360° radial flow towards a well. In the case of
these springs, a closer approximation would be goo - 180°, or 'TT - 2 TI
radians.
Still, the technique has provided order-of-magnitude estimates of
the transmissivity at minimal expense. The results of pumping tests in
-79-
karst aquifers are often quite undependable. It is only by chance that
a pumping test well will penetrate the system of secondary porosity in
such a way as to accurately reflect the transmissivity and storage coef
ficient of the entire aquifer. Springs have the advantage of being
hydraulically connected to a karst aquifer, and thus have the potential
of producing more dependable and accurate results.
Summary and Conclusions
The gauging station and rain gauge network have produced data useful
in investigating the way a karst drainage system responds to recharge.
The system responds differently than either surface water or typical
groundwater environments, and in fact has some characteristics of both.
The aquifer has several sub-environments of flow which are revealed by
the storm response hydrographs and recession curves. The springs are
capable of responding almost instantaneously to precipitation on the
recharge area. The larger storm responses (those for which the peak
discharge is greater than 3000 liters/sec) have a prolonged delayed
response component due to recharge of the network of smaller fractures
and openings. The smaller storm response hydrographs are more akin to
the response of a small surface watershed. Initial estimates of aquifer
transmissivity and diffusivity can be derived fram the recession data.
Much of the effort in this part of the study went into the installa
tion and calibration of the gauging station. It is hoped that the sta
tion will continue to produce useful data for many years, as there are
very few continuously gauged karst springs in the country. The impor
tance of the springs to the trout population of the state further
-80-
increases the need to understand their behavior. In an excellent
review, Yevjevich (1981) emphasizes the importance of isolating par
ticular karst drainage basins for study. The gauging station at Moth
and Grabau Springs and the recording rain gauge in the recharge area are
the first steps in this direction.
-81-
v. CONCLUSIONS
Southeastern Minnesota contains large groundwater reserves, and
these are a valuable resource. Yet this resource is fragile and suscep
tible to pollution. This is especially true of the carbonate formations
overlying the Decorah shale, in which there is intimate connection bet
ween surface water and groundwater, and in which subsurface water flows
through large, solution-enlarged fractures and conduits. There is a
need to understand the way water moves through this system in order to
make intelligent decisions concerning the utilization and protection of
this resource.
This study has concentrated on a small area of Fillmore county in an
attempt to define and investigate the processes that affect subsurface
water in this terrain. Dye trace studies have defined groundwater flow
paths and travel times, and have revealed hydrologic phenomena including
( 1) complex piracy of surface streams to the subsurface, (2) anasto
mosing subsurface flow, (3) a single sinkpoint feeding three separate
springs, (4) surprising travel time relationships, and (5) subsurface
drainages which cut across topographic boundaries. A preliminary hydro
logic study of the drainage basin of Moth and Grabau Springs has pro
duced information on the way the karst aquifer responds to recharge, has
revealed something about the sub-environments within the aquifer, and
has enabled the estimation of hydrogeologic parameters. People who
attempt to model this kind of system will have to account for this kind
of basic data.
Future work should include a more detailed mathematical examination
of the dye trace response curves and the storm response hydrographs.
-82-
Statistical analysis of the curves and stochastic modeling of rainfall
and spring discharge will probably prove fruitful. The application of
systems analysis to produce physically meaningful functions which con
vert system input to system output is promising. It would be useful to
couple the spring discharge measurements with water chemistry analyses,
investigating chemical interactions in the various flow environments
within the aquifer. The results of further detailed research in this
study area will not only be applicable within it boundaries, but will
add to the knowledge of carbonate areas in general, which make up 15-20
percent of the land surface of the United States.
-83-
REFERENCES
Alexander, E.C., Jr., 1980 (ed.). An Introduction to Caves of
Minnesota, Iowa, and Wisconsin: Guidebook for the 1980 National
Speleological Society Convention. NSS Convention Guidebook No. 21,
190 p.
Alexander, E.C., Jr. and Shaw, G., 1979. Southeastern Minnesota ground
water study, in: Problems relating to safe water supply in
southeastern Minnesota: Report to the Legislative Commission on
Minnesota Resources, p. B1-B55.
Aley, T. and Fletcher, M., 1979. The Water Tracer's Cookbook. Missouri
Speleology, v. 16, no. 6, 32 p.
Austin, G.S., 1972. Paleozoic lithostratigraphy of southeastern
Minnesota, in: Sims, P.K. and Morey, G.B. (eds.), Geology of
Minnesota: A centennial volume. Minnesota Geological Survey, p.
459-473.
Barnes, B.S., 1939. The structure of discharge-recession curves.
Transactions of the American Geophysical Union, v. 20, p. 721-725.
Broussard, W.L., Farrel, D.F., Anderson, H.W., Jr., and Felsheim, P.E.,
1975. Water resources of the Root River watershed, southeastern
Minnesota: Hydrologic Investigations Atlas HA-548, United States
Geological Survey.
Cobb, E .D • , 1968. Constant-rate injection equipment for dye-dilution
discharge measurements, in: Selected Techniques in Water Resources
Investigations. Geological Survey Water Supply Paper 1892, United
States Geological Survey, p. 15-22.
Dreiss , S .J • , 1980. An Application of Systems Analysis to Karst
-84-
Aquife:rs. Ph.D. thesis, Stanford University, 193 p.
Frye, J.C., 1973. Pleistocene succession of the Central Interior United
States. Quaternary Research, v. 3, p. 275-283.
Giammona, C.P., 1973. Fluorescent dye determination of groundwater
movement and contamination in permeable :rock strata. International
Journal of Speleology, v. 5, p. 201-208.
Hobbs, H.C. and Goebel, J.E., 1982. Geologic Map of Minnesota:
Quaternary Geology. Minnesota Geological Survey, University of
Minnesota.
Kneisel, W.G., Jr., 1963. Baseflow recession analysis for comparison
of drainage basins and geology. Journal of Geophysical Research, v.
68, no. 12, p. 3649-3653.
Lively, R.S., 1983. Late Quaternary U-series speleothem growth :record
from southeastern Minnesota. Geology, v. 11, no. 5, p. 259-262.
Lively, R.S. and Alexander, E.C., Jr., 1980. 230Th/234u ages from
speleothems from Mystery Cave, Minnesota.
Society Bulletin, v. 42, p. 34.
National Speleological
Mijatovic, B. , 1968. A method of studying the hydrodynamic regime of
karst aquifers by analysis of the discharge curve and level fluc
tuations during recession. Vesnik Zavoda za Geoloska i Geofizicka
Instrazivanja, Series B, no. 8, p. 43-81.
Milanovic, P., 1976. Water regime in deep karst: case study of the
Ombla Spring drainage area, in: Yevjevich (ed.), Karst Hydrology
and Water Resources: Proceedings of the U.S.-Yugoslavian Symposium,
Dubrovnik, June 2-7, 1975. Water Resources Publications, Fo::rt
Collins, Colorado, p. 165-191.
-85-
Milske, J.A., 1982. Stratigraphy and Petrology of Clastic Sediments in
Mystery Cave, Fillmore County, Minnesota. M.S. thesis, University
of Minnesota, 111 p.
Pierce, C.H., 1941. Investigations of methods and equipment used in
stream gauging. Part 2: Intakes for Gauge Wells. USGS Water Supply
Paper 868-B.
Rorabaugh, M.L., 1960. Use of water levels in estimating aquifer
constants in a finite aquifer. International Association of
Scientific Hydrology Publication 52, p. 314-323.
Rorabaugh, M.I., 1964. Estimating changes in bank storage as ground-
water contribution to streamflow. International Association of
Scientific Hydrology Publication 63, p. 432-441.
Schoeller, H., 1948. Le regime hydrogeologique des calcaires eocenes du
synclinal du Dyr el Kef (Tunisie). Soc. Geol. France Bull., ser. 5,
v. 18, p. 167-180.
Sims, P.K. and Morey, G.B., 1972. Resume of geology of Minnesota, in:
Sims, P.K. and Morey, G.B. (eds.), Geology of Minnesota: A centen
nial volume. Minnesota Geological Survey, p. 3-15.
Singer, R.D., Osterholm, M.T., and Straub, C.P, 1982. Groundwater
quality in southeastern Minnesota. Water Resources Research Center
Bulletin 109, University of Minnesota, Minneapolis, MN, 45 p.
Sloan, R.E. and Austin, G.S., 1966. Geologic Map of Minnesota: St.
Paul Sheet. Minnesota Geological Survey, University of Minnesota.
Sweeting, M.M., 1973. Karst Landforms. Columbia University Press, New
York, 362 p.
Theis, C.V., 1935. The relation between the lowering of the piezometric
-86-
surface and the rate and duration of discharge of a well using
ground-water storage. Transactions of the American Geophysical
Union, 16th Annual Meeting, p. 519-524.
Torbarov, K., 1976. Estimation of permeability and effective porosity
in karst on the basis of recession curve analysis, in: Yevjevich,
v. (ed.), Karst Hydrology and Water Resources: Proceedings of the
U.S.-Yugoslavian Symposium, Dubrovnik, June 2-7, 1975. Water
Resources Publications, Fort Collins, Colorado, p. 121-136.
Trainer, F.W. and Watkins, F.A., 1974. Use of base-runoff recession
curves to determine areal transmissi vi ties in the Upper Potamac
River basin. U.S. Geological Survey Journal of Research, v. 2,
no. 1, p. 125-131.
Wopat, M., 1974. The karst of southeastern Minnesota. The Wisconsin
Speleologist, v. 13, no. 1, p. 1-47.
Wright, H.E., Jr., 1972. Physiography of Minnesota, in: Sims, P.K. and
Morey, G.B., (eds.), Geology of Minnesota: A centennial volume.
Minnesota Geological Survey, p. 561-578.
Yevjevich, V., 1976. Advanced approaches to karst hydrology and water
resource systems, in: Yevjevich, V. (ed.), Karst Hydrology and
Water Resources: Proceedings of the U .s .-Yugoslavian Symposium,
Dubrovnik, June 2-7, 1975. Water Resources Publications, Fort
Collins, Colorado, p. 209-220.
Yevjevich, 1981 (ed.). Karst Water Research Needs. Water Resouces
Publications, Littleton, Colorado, U.S.A., 260 p.
-87-
APPENDIX A Dye Trace Data
Table A1 Quantitative fluorornetric dye trace from Disappearing River to Seven Springs. 914 g of 20% Rhodamine WT solution dropped in Disappearing River (at entrance to Mystery I Cavel at 3:13 a.rn., October 10, 1979.
* The numbering of the individual rize points within Seven Springs is shown in Fig. 7 •
-88-
APPENDIX A (continued)
Table A2 Quantitative fluorometric dye trace from Matheson Sink to Seven Springs. 1,118 g. of 20% Rhodamine WT solution dropped in Matheson Sink at 6:35 a,m., October 20, 1979.
* The numbering of the individual rize points within Seven Springs is shown in Fig, 7 •
-89-
APPENDIX A (continued)
Table A3 Quantitative fluorornetric dye trace from Matheson Sink to Moth and Grabau Springs. 1,118 g, of 20% Rhodamine WT solution dropped in Matheson Sink at 6: 35 a ,rn. , October 20, 1979.
Hours after Dye concentration (ppb) Hours after Dye concentration (ppb) dye drop Moth Spring Grabau Spring dye drop Moth Spring Grabau Spring
Table A4 Quantitative fluorometric dye trace from Fairview Blind Valley (Hellerud Sink A) to Moth and Grabau Springs. 863 g, of 20% Rhodamine WT solution dropped in Hellerud Sink A, the terminal sink point, at 3:30 p,m., November 17, 1979.
Hours after Dye concentration (ppb) Hours after Dye concentration (ppb) dye drop Moth Spring Grabau Spring dye drop Moth Spring Grabau Spring
Table AG Quantitative fluorometric dye trace from Red Tail Sink to Moth and Grabau Springs. 1325 g Rhodamine WT 20% solution dropped in Red Tail 10:10 CDT, June 28, 1980.
Table A7 Quantitative fluorometric dye trace from Fairview Blnd Valley (Poldervarrd Sink A) to Moth and Grabau Springs. 1203 g Rhodamine WT 20% solution dropped at Poldervaard sink A, the terminal sinkpoint, at 18:36 CDT July 13, 1980.
Table AB Quantitative fluorometric dye trace from Formation Route Creek (Mystery Cave) to Seven Springs. 309.3 g Rhodamine WT 20% solution dropped under the first Triangle Room in Mystery III cave at 10:19 CDT, August 25, 1980.
Hours After Concentration of o:i::e <EEhl* Dye DroE 2 3 4 5 6 7 8 9
* The numbering of the individual rize points within Seven Springs is shown in Fig7
-95-
APPENDIX A (continued)
Table A9 Quantitative fluorornetric dye trace from Disappearing River to Seven Springs. 266.3 g Rhodarnine WT 20% solution dropped in Disappearing River ( at entrance to Mystery I cave) at 6:00 CDT, September 2, 1980.
Hours After Concentration of Dye <EEbl* Dye DroE 2 3 4 5 6 7 9
* The numbering of the individual rize points within Seven Springs is shown in F-ig. 7 .
-96-
APPENDIX A (continued)
Table A 10 Quantitative fluorometric dye trace from Lefevre Blind Valley to Moth and Grabau Springs. 895.4 g Rhodamine WT 20% solution dropped in Lefevre Blind Valley at 17:00 CDT, October 7, 1980.