Klaus Beyrle Undergraduate @ SUNY
Brockport
Paleo Flow Analysis of a Meltwater Channel
Feeding
During the last glaciation of
The sand pits are located in the
7.5 minute Holley Quadrangle of
During
the Wisconsinan glacier advance a lowland was carved
out of the
Insert Pic
The paleoflow of a meltwater
channel from the Holley quadrangle in
This is a
close-up picture of the four pits. The
red dots represent the location of the pit and are on the fringe of the
delta. The protruding finger of the
The picture on the left is one
that has an outline of the delta in black, and the picture on the right is one
that has an outline of the lake in blue.
The hypothesis of bifurcated flow depositing the sand pits is supported
entirely from these topographic maps. It
is obvious that a finger of the
Fluvial Analysis
Field Work
Lithologic description
of each pit
Describe grain size,
sedimentary structures
Measurements of crossbed orientation and crossbed
dip
Samples were collected
for grain size analysis
Laboratory Work
Prepared samples
Dry
Analyzed samples using
the Coulter particle size analyzer
Estimation of stream
flow direction
Estimation of stream
flow regime
Estimation of stream
flow velocity
Analysis of maximum
clast size
At right is a picture of one of
the pits. Each pit had to have a fresh
surface prepared for analysis.
Graph 1 is the average grain size plotted against
depth. There is no discernable pattern
between grain size and depth and a large variation can be observed. The texture class, according to the Wentworth
size classification, ranges from coarse sand (0phi) to coarse silt (5phi).
Sorting
is represented as the standard deviation graphed against average grain
size. According to Folk, 1.0-2.0 SD
represents a poorly sorted sample, and 2.0-4.0 represents a very poorly sorted
sample. The majority of the samples in
graph 2 from all four pits showcased poor sorting with standard deviations
between 1.0 and 2.0, and a few samples from pit 1 were very poorly sorted with
standard deviations from 2.0-2.50. High
standard deviations reflect the large variability in the samples.
Skewness is a measure of the
symmetry of distribution. Folk reports
that positive skew values from 1.00 - 0.30 dictate that the sample is strongly
fine-skewed, and positive skew values from 0.30 0.10 indicate that the sample
is fine-skewed. Graph 3 illustrates that
most skewness values for pits 2, 3, and 4 were between 0.1 and 0.3. Therefore, these samples are fine-skewed. The
majority of samples from pit 1 have skewness values
from 0.3 to 1.0 and are therefore strongly fine-skewed. A sample that is fine-skew is one that has
excess fine sediment and a tail to the right in the distribution.
Graph
4 is a comparison of sorting (standard deviation) to skewness. A relationship
appears to exist between samples with high standard deviations and samples with
high positive skewness values. Samples
that are poorly sorted also seem to have excess fine material.
This is a diagram of Hjulstroms
curve. The graph is one comparing flow
velocity with grain size. In the plot
area of the curve, sedimentary processes of entrainment, transport, and
deposition are represented. The important
curve is the one separating the deposition area from the transport area. This line represents the velocity needed to
start transporting the grain. For
example, if the average sample size was -4phi (16mm) then a flow velocity of
100cm/s is needed to transport that grain to that location and deposit it. Hjulstroms curve
is one that determines magnitudes of flow.
In other words, the purpose is to determine whether the flow was 10cm/s,
100cm/s, or 1000cm/s. This is an estimation and is used to ballpark the velocity of the
water that produced the deposit.
The estimated velocities based on grain size are
represented in graph 5. From the graph
one can observe that the maximum velocity for pit 1 was about 3.75 cm/s, for
pits 2 and 3 approximately 2.75 cm/s, and for pit 4 a little greater than
1cm/s. When these sediments were
deposited the flow velocity of the meltwater channel was quite low, less than
4cm/s. All the clasts larger than -1 phi
were used to produce velocity estimates for the pits. Theoretical models by Allen (1985) predict
that larger grains require stronger flows to move them. Also Miller et al. (1977) conducted
experiments showing that coarser beds need stronger flows to start their
transport. Therefore, it is expected
that these velocities will be larger since they are based on larger
particles. The maximum velocity for pit
1 based on maximum clast size was approximately 150cm/s, and for pits 3 and 4
it was approximately 35-40cm/s. Pit 2
did not have any clasts larger than -1 phi, and is not represented on the
graph.
The analysis of angular data revealed that the major flow
in all four pits was to the south with slight variance either east or
west.
|
Table 1 |
Pit 1 |
Pit 2 |
Pit 3 |
Pit 4 |
|
|
Orientation |
Orientation |
Orientation |
Orientation |
|
Vector Mean |
178.3 |
223.1 |
218.3 |
157.4 |
As reported in table 1, the mean
orientation for pit 1 was 178.3°, for pit 2 was 223.1°, for pit 3 was 218.3° and for pit 4 was 157.4°.
All of the pits had ripple sedimentary structures. According to Harms and Fahnestock,
this would categorize the flow in the lower energy regime.
All of the
pits contained a large amount of variability.
This inference is concluded from that fact that there is no recognizable
vertical pattern in the mean grain size for any of the pits (Graph 1). This is further supported by the fact that
samples are all, at best, poorly sorted (Graph 2). In other words the distribution of grain
sizes is heterogeneous and reflects the immaturity of the deposits. Poor sorting could be reflective of velocity
changes, rapid sedimentation, or initially poorly sorted material supplied to
the river (by the glacier). The high
skewness values, as stated previously, indicate that deposits have excess fine
material (Graph 3). This could have been
caused by changes in flow velocity. For
example, as the velocity of a flow decreases it is unable to carry large sizes
of sediment. Therefore, a low enough
drop in velocity would cause the finer particles to be deposited. There does appear to be a relationship
between sorting and skewness (Graph 4).
In graph 4 above a trend line has been added to 
illustrate
the relationship. It appears that
increasing standard deviation coincides with increasing skewness. However, this relationship seems
logical. A sample with excess fine sediment
has a bimodal distribution which means there are two
grain size modes represented in the sample.
If a sample has two modes it is probable it will also have high
variability (a high standard deviation).
The aforementioned data is analogous to the sediment deposited by a
braided river. J.R.L. Allen (1970)
remarks that braided streams are composed of texturally varying sediments, but
are dominated by sediments void of clay and mud. All four of the pits had large variation and
were for the most part deficient in mud and clay. The clay and mud act as a binding agent
between sand grains in the sediment.
Therefore, absence of sediment this size is what allows the banks of
braided rivers to be eroded so easily.
Easily erodible banks allow the establishment of braided flow. Schumm (1977)
related high percentage of sand and gravel to braided river morphology. The sediment size from all the pits ranged
from coarse sand to coarse silt. The pits
are primarily composed of sands. Braided
streams are also known for having flashy discharges, or fluctuating
velocities. This would cause the
deposits of a braided river to have excess fine material. Braided flow is an appropriate label for the
meltwater channel that deposited these four pits.
Evidence
of flashy discharge is also offered by velocity estimates based on grain size
and maximum clast size. The maximum
grain size velocity wasnt over 4cm/s, but the maximum clast velocity was about
150cm/s. The likely scenario is that
there was predominately a flow of 4cm/s or less with the occasional burst of
water and sediment depositing the clasts.
On another note, flashy discharge, or at least fluctuating velocities,
would be expected since the system is influenced by the rate at which the
Laurentide ice sheet melted. Therefore,
in warmer months the flow would be faster than in colder months.
The
hypothesis of a bifurcated flow depositing the sand pits is supported entirely
from the topographic map. In figure 1
(re-shown below) it is obvious that a finger of the
Characterizing
the flow of the meltwater channel engaged many lines of evidence to determine
that it was likely a braided channel.
This determination was based on sorting values, average grain size
values, as well as skewness and velocity values. The objective in proving a historical
hypothesis is to gather as many supporting lines of evidence as possible. These multiple lines can be woven together to
form strong arguments validating the proposed hypothesis. Since many of the characteristics typifying
braided flow could be exhibited in the sediments deposited by the meltwater
channel, one can conclude that the meltwater channel also exhibited braided
river morphology.
Based
on the characteristics of the grain size data it was determined that the flow
velocity was predominately 4cm/s or less with occasional surges reaching
150cm/s. Also because all the preserved
bed forms in all the pits were ripples and cross-strata, the flow can be
characterized as being in the lower-flow regime. The flow direction was determined from the
analysis of the strike orientation of the cross strata and was primarily toward
the south.
Lastly,
based on topographic analysis, it is very likely that a bifurcated flow
deposited the pits instead of one continuous channel.
Bibliography
Allen, J.R.L. 1970. Physical Processes of
Sedimentation. George Allen and Unwin,
Allen, J.R.L. 1985. Principles of Physical
Sedimentology. George Allen and Unwin,
Harms, J.C., and Fahnestock,
R. K. 1965. Stratification, Bed
Forms, and Flow Phenomena (With an Example from the
Miller, M.C., McCave, I.N., and Komar, P.D. 1957. Threshold of Sediment Motion Under Unidirectional Currents. Sedmentology 24:507-527.
Muller, Ernest H., and Calkin,
Parker E. 1988. Later Pleistocene and Holocene Geology of the Eastern