Thesis Proposal, revision 2

From GeoMod

Contents 1 Introduction 2 Literature Review 3 Description of Study Area: 4 Problem Description: 5 Proposed Solution: 6 Methodology

Introduction

This study will attempt to identify patterns (del-that may exist Lurbano) in river systems where exchange with groundwater systems has altered the geomorphic form of the river. A multi-method approach will be used to study the effect of groundwater/river water exchange to ensure accuracy and integrity of the study. Field observations, (deemphasize Lurbano) laboratory sand box models, (eliminate Lurbano) and computer simulated models will be integrated into the study to determine actual, idealized, and theoretical effects of groundwater intrusion in river systems. Once (areas of high groundwater discharge Lurbano) are identified (outline how you identify these locations Lurbano), bed forms and larger scale features will be statistically analyzed and compared to non-exchange river sections. A channelized meandering river system will be used for the study area for ease of comparison of large scale features, such as reinitiated meander bends, and for ease of sampling in lower velocity waters. Field data will then be correlated to observed features obtained by sand box modeling to ensure accuracy of scale. The goal is to form a better understanding of the processes at work between the surface and subsurface environment.

Literature Review

Previous work has shown that a correlation can be made between river geomorphology and groundwater movements. A Texas study by Larkin and Sharp (1992) found that streams with a significant base flow tended to have a sinuosity greater than 1.5 (when compared to similar streams- Lurbano). Lower gradients where also associated with base flow dominated systems. The reason for this behavior was not investigated.

Previous studies done on the exchange of groundwater and surface water have mainly studied fluid transport. Although the geomorphic effect of this exchange is often overlooked (look at Media:Lautz seigel2006.pdf and add a reference Lurbano), the methods used to determine amount of exchange are very useful. Becker (2004) describes the following three approaches to estimating the amount of groundwater flux and the problems involved with each approach.

• Traditional stream gauging is done by determining river flow volume at various locations along the river. Any discrepancy in the flow volume is perceived as groundwater discharge or recharge. The traditional method does not account for other possible sources of added river water, such as runoff, and are made inaccurate by any kind of precipitation.
• A second method uses macroscopic surveying of river water for temperature change. River water that is closer to the groundwater temperature is presumed to have a groundwater influence. This method can determine location of groundwater exchange but does not determine rate.
• Another point temperature based method of acquiring discharge rate is said to be accurate (said by whom? Lurbano), independent of precipitation, but only measures rate at a single location. This is done by burying temperature data loggers at various depths and determining the temperature gradient with depth over time.
• (What about seepage meters Lurbano)

Since the goal of this study is to determine geomorphic features of varying scale, it is necessary to determine precise areas of groundwater exchange. The macroscopic temperature methods will be utilized for its ability to locate these finite locations. The traditional method will be used to determine general volume of discharge. The point temperature method will not be used due to its (suspect accuracy --> cite a reference or say that it does not give good areal data Lurbano) and the long (del-amount of Lurbano) time needed to obtain data.

A study by Wells(2002) in Western New York compared the accuracy of the stream gauging method with the temperature gradient discharge method. It was found that a temperature flux existed in cobble sized sediment to a depth of 0.5 meters while in bedrock substrate, the flux was only present to a depth of 0.1 meters. The flux in temperature was due to daily solar heating of stream water and subsequent cooling at night. Deeper water should have less flux in temperature (in the above section you should have a change in temperature or a flux in heat, remember flux is the movement of heat across a boundary Lurbano). The discharge rates found from stream gauging were an order of magnitude higher than those found by temperature gradient in Wells’ study. It is possible that the temperature sample location was not in a peak discharge area. This kind of non-uniform discharge may create unique bedfroms or channel pathways.

Bank erosion processes have been the focus of several studies and may be an important factor in meander propagation. (del-A study conducted in 1999 by Lurbano) Toth (1999) found that positive pore pressure had the effect of loosening the grain matrix in sediment while negative pre pressure had the effect of increasing the cohesiveness of sediment. A later study found that increased pore pressure created by effluent groundwater reduced the overall bed friction and may therefore increase erosive ability of the river (Simon et al., 2000).

Description of Study Area:

Loosahatchie River, Shelby County, Tennessee

The purpose of this study is to qualitatively analyze the effects of groundwater exchange on meandering river systems. The Mississippi Embayment offers an ideal location due to the abundance of meandering rivers in contact with aquifer outcrops. The embayment itself is shaped like a south plunging syncline composed of alternating layers of sand and clay shallowly dipping towards the axis of the syncline. This axis line trends approximately below the Mississippi River. (Do you have a figure? Lurbano) The embayment was formed by transgressive and regressive sequences of coastal facies formed by the recurrent flooding the embayment. In Shelby County, Tennessee, continuous sand and shale outcrops trending in a north/south direction outcrop across the width of the county. The sand layers are known nationwide for their excellent groundwater transport ability (is there a reference? Lurbano). The shale/clay layers in the region are also known to be of varying thickness with thin spots and windows throughout. The rivers in Shelby County tend to cross these outcrops perpendicular to strike direction, flowing from East to West towards the Mississippi River. The Loosahatchie River was chosen primarily for its East/West orientation and relative straightness along its length in Shelby County. In 1968, the Loosahatchie River was channelized by the Corp of Engineers to reduce flooding and increase drainage potential. Prior to channelization, the Loosahatchie River was highly sinuous with only a general East/West orientation. It has subsequently begun to revert back to a meandering state. (You need to mention that the straightening has resulted in substantial downcutting, resulting in a steeply sided channel which restricts meandering. Lurbano) The Wolf River, a river in Shelby County similar to the Loosahatchie River, was channelized and straightened in the early 1960’s. Flow velocity subsequently doubled and the river channel incised to varying degrees along the length of the river (Van Arsdale, et al., 2003). Similar effects can be expected to have occurred along the Loosahatchie River. (Are you still planning on looking at the Wolf? Lurbano)

Problem Description:

The Corp of Engineers channelized many meandering rivers across Tennessee in the 1950’s and 60’s. Since then, these river channels have begun to revert back to a meandering state. Along with increased sediment load, destruction of habitat, loss of organic carbon sources for aquatic life, and loss of floodplain soil renewal, channelizing rivers costs money. Periodic maintenance is required to maintain the form desired by planning engineers. By removing natural streambeds and exposing aquifer recharge sediment directly to free flowing runoff sources, the potential for groundwater contamination (has Lurbano) also increased. A better understanding of the relationship between the geomorphology of channelized rivers and the underlying aquifer systems may aid in determining the potential for increased contamination.

Several models exist to explain the propagation and migration of river meanders. Einstein and Shen (1964) proposed a method by which twin, periodically reversing, surface convergent helical cells scour alternating pools, leading to stable meanders. Thompson (1986) devised a model were surface convergent flow interacts with a mobile bed, subsequently creating riffle and pool sections mimicking a stage 1 meander initiation (explain what "stage 1 meander initiation" means Lurbano). The macroturbulant flow model predicts velocity bursts at various locations along the river. These bursts create deeper pool sections and may initiate meanders (Yalin, 1971). None of these models are able to account for the existence of seemingly regular meander patterns in tandem with very irregular meander patterns (is there a reference for this statement? or is this pattern something you have observed? If the latter you will need to back up the statement, perhaps with a diagram Lurbano). The symmetry encountered in many meandering systems seems to imply that an external factor to the river system is at work. Irregularity can be accounted for by consideration of variant riverbed geology or sediment deposition, but why regular bends and irregular bends occur in similar geologic settings comes into question. Nor can any model convincingly (describe the process of meander belt initiation Lurbano) (del- account for how the process of meander bends is initiated Lurbano). The traditional view analyzes single reaches or bends and analyzes finite properties of the bend (Marham and Thorne, 1992). (The --> An Lurbano) alternative view analyzes the meanders in a collective sense, taking into account general trends and property change throughout the series of meanders (Fergason, 1984, Howard, 1992). Fergason (1979) (breaks this series approach into --> uses this approach to analyze Lurbano) three main components (of stream geometry Lurbano); a scale variable, sinuosity, and degree of irregularity. The benefit of this approach becomes apparent when analyzing property change at various locations along the river length. The traditional approach still retains value when analyzing finite properties of an individual reach.

Proposed Solution:

The effect of groundwater exchange on the geomorphology of river systems has the potential to answer the question of meander initiation as well as meander regularity or irregularity. It is important to first note that the confining layers in the Mississippi embayment are not of uniform thickness, nor are they always continuous. The possibility exists for thin locations and windows throughout the confining clay layers. Several possibilities exist for potential impact.

1. Discharge to the river form the groundwater system is not uniform along the riverbed. Locations of peak discharge may exist with a decreasing gradient of discharge radiating out around them, or possibly windows in the riverbed may exist allowing groundwater discharge solely through these windows.

2. The up-flow of groundwater in gaining streams may aid in destabilizing streambed grains, aiding in entrainment and resisting deposition. The minimum energy needed to entrain sediment is found by means of calculating the critical shear stress of bed material.

Ï„c = k(Ï?s – Ï?)g D - or approximately - Ï„c = 0.73D

where: Ï„c = critical shear stress D = Particle size (Knighton, 1998) This potentially creates deeper sections, or pools, at areas of higher groundwater discharge. River water volume and velocity increases at these pools and the traditional meander propagation model developed by Einstein and Poole(1964) then applies.

3. Groundwater discharge through riverbeds may create a reduced surface or bed friction. This resistance factor can be found using Manning’s equation:

n = k(R2/3*s1/2)/v

where: n = resistance factor k = 1(SI units) R = hydraulic radius s = slope of energy gradient v = mean velocity (Knighton, 1998) River water flowing through these areas increases in velocity. The velocity increase then promotes increased bank erosion in these sections.

Methodology

The problem will be approached by means of three separate methods. The primary method will involve direct measurement and observation of field conditions along the Loosahatchie River. The direct results of channelization and meander initiation can be studied and documented. A sand box model will act as the second method of investigation. This a priori approach will allow a controlled physical environment in which contributing factors to meander propagation and migration can be tested. Computer modeling offer a third approach in which calculations and initial theoretical conclusions that are time dependent can be tested in an accelerated time frame. The computer model results will then be compared to other physical models and field observations for accuracy.

Field Observation Methodology:

Since the thickness of the confining units that outcrop throughout Shelby County have variable thickness and outcrop location, it becomes necessary to first locate areas of groundwater discharge. The primary method that will be used utilizes the temperature difference in groundwater and surface waters. Since subsurface water equilibrates with the 58° F surrounding lithology, groundwater maintains a constant ground temperature. During summer months, surface waters in lakes and rivers averages well above this, and during winter months surface water temperature can be well below the 58° F ground water temperature. Gaining streams receive water from groundwater sources, therefore areas in the river that deviate in temperature from the surface water average towards the ground water average can be inferred as areas of groundwater discharge. For the purposes of this study, this deviation in temperature will be used as a proxy for groundwater discharge location. To locate discharge zones, a panel of temperatures at varying locations, depths, and cross sectional distances will be analyzed.

1. River bed temperature will be measured at several locations at varying depths for an extended period to determine the diurnal variation in temperature. Temperature will be measured at 0 cm, 5 cm, 10 cm, and 15 cm depth. This diurnal variation will be used to calibrate further temperature readings. 2. A general temperature and depth profile will be made of the study area. A temperature data logger will be attached to the end of a one meter pole and set to sample at 2 second intervals. A GPS unit will also be set to take continuous special readings at the same 2 second interval. A boat will then travel downstream towing the pole with data logger at a 1 meter depth along the entire study area. Depth will be taken at a set interval as well. The temperature data from the data logger will then be correlated to the position data from the GPS unit to from a temperature profile of the study area. This will give a general idea of areas of groundwater discharge and allow for the testing and accuracy of the data logger/GPS setup. Depth will be correlated with GPS location as well to generate a rough depth profile. 3. From the temperature profile, an area containing a discharging location will be selected, as well as a non discharging location. Each location will contain a full meander bend pair. Using a ground penetrating thermocouple probe, sub-riverbed temperatures will be measured at a depth of 5 cm below bed surface along a cross sectional profile running perpendicular to river bank. These sub-bed temperatures will be taken at a one meter interval across the cross section. This cross sectional measurement method will be applied at riffle points, outer meander bend locations, and transition zones from riffle to meander and meander to riffle. A river water temperature profile will also be created. At the same one meter interval across the same cross sections, water temperature and depth will also be measured at 10 cm above river bed, mid river depth, and at 10 cm below water surface. This will generate a river water temperature cross section that will be compared to sub-riverbed temperatures. Areas of high discrepancy between river water and sub-bed temperature will be treated as high groundwater discharge areas. A down stream line can then be created by correlating all cross section profiles to find the downstream path of highest groundwater discharge. 4. A 10 cm long ground penetrating grounded J-type thermocouple will then be attached to the end of the sampling pole. Grounded J-type thermocouples have an accuracy of +/- .5 degrees Celsius at the temperature range in question and a response time of 0.5 seconds. A skid plate will be attached directly above the probe to prevent penetration of more than the 10 cm probe length. A standard temperature data logger will then be set 10 centimeters from the penetration probe. A second standard temperature data logger will be set at a depth of 10 cm below river surface. All data loggers and GPS will be set to a sampling interval of 2 seconds. The probe will them be drug across the stream bed along the down stream line of maximum groundwater discharge. A downstream profile can then be created of sub-bed temperature, epi-bed temperature, and subsurface temperature. These profiles will then be correlated to locate exact locations and relative rate of groundwater discharge.

Maps of the Loosahatchie River created prior to channelization in the early 1960’s will be analyzed for sinuosity of the river channel. Regions of high groundwater discharge will be overlaid on these pre-channelization maps. A comparison can then be made of sinuosity and groundwater discharge. A Chi Squared test will be used to statistically correlate the data. On a smaller scale, depth and location of high discharge will be correlated to identify if high groundwater discharge areas coincide with deeper pool sections in the river. This will also be preformed using a Chi Squared test. If there is no correlation, than the increased erosion at discharge points theory will be disproved.

5. Flow rate will be measured at the beginning, middle, and end of the known aquifer discharge study area. Rates will be obtained using an acoustic Doppler probe. Flow rates will also be measured in the non discharging river study area. Using Manning’s equation, the resistance of stream bed will be calculated for the discharging and non discharging study areas. The sets of data will then be processed statistically using F-test and t-test methods. This will reveal any statistical correlation that may exist. If the two data sets are statistically similar, the possibility of friction reduction in discharging areas can be eliminated.

Sand Box Model Methodology: A scaled sand box model capable of modeling ground water/surface water interactions will be constructed. Water will percolate up from below through a porous medium to simulate an aquifer discharging into a fluvial network. The box model will allow for adjustable aquifer head pressure as well as discharge area. The box will be capable of handling multiple layers of varying porosity and composition. Dimensions: 1.0 m * 1.0 m * .1m

Figure 1: AutoCAD rendering of sandbox layout in isometric view. Water resevior is located on the top right. Water will exit the box to the bottom left.

The box will be constructed of an outer wooden box frame with a fiberglass coating for waterproofing. One of the 1 m ends will have a drainage lip with a central notch 1 cm in depth. The other 1 meter end will have a reservoir with a variable depth drain spout to allow for multiple head levels. The bottom of the box will have 1.5 cm risers spaced in 10 cm intervals in grid format. A 1 cm mesh screen will be placed on the risers with an additional .2 mm mesh screen on top of it. The media used will consist of a base of medium sand topped with a surface of 50% diatomaceous earth and 50% rock flour. This mixture will accurately scale the effects and properties of water (Smith et al., 1998). Testable hypothesis utilizing Sand Box:

• Ground water discharge creates low spots in river systems due to destabilization of riverbed due to increased pore pressure and therefore increase erosion rates at this point.
• River water velocity increases in areas of ground water discharge due to reduced bed friction
• Variations in thin confining layer thickness creates windows through which ground water discharges at higher rates. These higher rates increase erosion rates of surrounding streambed, widening the river near the window.
• Groundwater discharge spikes occur where aquifer outcrops are separated by a thin confining bed.
• River channels migrate towards aquifer discharge outcrops.

Procedure for testing hypothesis:

• An even layer of sand, 4 cm thick, will be placed in the box. An overland flow component will be created in the box by means of tube supplying constant water volume to the surface of the sand. Natural waterways will be allowed to form and the water channel depths will be checked for varying depth. An artesian water pressure will then be applied to the ground water in the box and natural waterways will be allowed to form. The channels will be checked at regular time intervals for varying depth. The results form no ground water pressure will be compared to those of positive ground water pressure for variance.
• An even layer of sand, 4 cm thick, will be placed in the box. An overland flow component will be created in the box by means of tube supplying constant water volume to the surface of the sand. Natural waterways will be allowed to form. A drop of dye will be added to the upstream end of the box and timed as it moves across the box. Water volume of the current situation will be logged as well. An artesian groundwater component will be added. A second drop of dye will be added to the upstream end of the box and timed once again as it moves across the box. The water volume will be checked again and dye velocity will be corrected for the additional water volume added by groundwater. Results from each run will be compared for variance.
• An alternative method will be to create a 10 cm thick layer of sand. An artificial channel, 8 cm deep will be created. Overland flow will be created and water velocity cross sectional profile will be checked using dye tests. An artesian component can then be added and checked again at varying locations in the cross section and compared to the previous test.
• An initial layer of sand will be randomly applied in the box to an average of 2 cm thickness. A layer of dry powdered clay will be randomly applied to the surface of the sand to an average of 1 cm thickness. An additional layer of sand will then be randomly applied on top of the clay layer. A surface flow will be initiated to create natural waterways. Path and channel depth will be documented. An artesian flow will then be created from below. Channel path and depth will then be documented at an even interval and compared to the non-artesian state.
• Sand and dry clay powder will be placed in the box in such a way as to simulate a dual aquifer system separated by a 1 cm thick confining bed. A surface flow will then be created to allow a natural waterway to form. Artesian flow will then be applied to the system. Flow rates along the waterway will then be documented at even spacing to locate any high discharge areas.
• A plastic sheet will be placed in the box covering half the screen to block flow laterally from the artesian source. An even layer of sand, 4 cm thick will be placed in the box. An overland flow component will be created in the box by means of tube supplying constant water volume to the surface of the sand. Natural waterways will be allowed to form. The artesian source will then be initiated. Only one lateral half of the box will receive ground water discharge. The channels will then be allowed to change positions naturally and pre-artesian river position will be compared to artesian positioning.

Expected Results

• The box will form a meandering river system of low sinuosity. Cannel width will be approximately 1 cm and channel depth will be .2 cm. When groundwater flow is added, sinuosity will increase and channel depth will become less uniform and from several low spots where groundwater discharge is randomly higher.
• It is expected that the dye travel time will be faster with the artesian component added.
• Water channels will migrate towards random thin spots in the clay layering. This will create a more sinuous river with a more random meander.
• Flow volume increase will be highest at the point where the two aquifers converge. It is possible that the river may try to follow the clay confining layer outcrop for this reason.
• The river will migrate to the groundwater discharging half of the sand box. Given enough time, the entirety of the river channel will be in the artesian half of the box.

Computer Model Methodology: An appropriate model must be able to accurately simulate groundwater discharge in normal river conditions. The program Visual Python will be for its versatility and ease of programming. A two dimensional model will be created to represent a lateral cross section of river, including floodplain and subsurface geology. Properties such as critical shear stress, flow volume and velocity, water density, sediment load, and bed and bank cohesiveness will be taken into account. The initial profile will represent the profile of a stream immediately following channelization. Groundwater discharge areas will then be added to various locations along the profile. The profile will then be allowed to entrain, erode, and deposit sediment based on the governing equations of river flow dynamics. Channel location will be monitored as the time step progresses. Sub layers of clay of varying thickness can then be added below the riverbed. Bank stability can also be made to simulate riprap, concrete encasement, or other forms of bank stabilization. The model will also be able to calculate suspended load, bed load, and flow velocity. Results of the model will be verified with field measurements of similar setting.

Scientific Importance of Study

The straightening of rivers has caused a host of problems, many of which we may not be aware of yet. The ability to predict a rivers future path would allow planning engineers to produce a more environmentally sound, cost effective, and structurally stable solution. Knowledge of the dynamic processes at work in such exchange systems would also aid in contaminant transport prediction in groundwater systems. The possibility also exists for measuring river flow volume by satellite if a true temperature gradient in peak discharge locations can be identified. Geomorphic surface forms may allow interpretations of subsurface geology. Applications of this would have an impact on hydrology, the petroleum industry, and even extra-planetary exploration.

Work Schedule

Field Data Acquisition

May 13 Scout the river • Find suitable boat landing points • Locate apparent meander cut banks and point bars • Locate suitable study area • Record depths at various distances downstream to get an idea of general depth range • Create a mid-water downstream temperature profile o Purpose is to test equipment o Generates a “general� idea of the effectiveness of the equipment o Identify unforeseen needs for taking temperature samples o Is the stream temperature significantly different than the groundwater Temp?

May 20 Measurement • From up stream end of study area, measure flow volume at beginning, middle, and end of study area • Create a sub-streambed temperature cross sectional profile along bend, riffle zone, and the up and downstream transition zones from bend to riffle • Create an in stream temperature profile. Record temperature and depth just above streambed, mid-depth, and at surface across the river in 1 meter intervals

May 27 Downstream Profile • Record sub-streambed temperature at 2 second intervals along the highest discharge line found on day 2 • Record river temperature 10 cm above streambed at 2 second intervals along the highest discharge line found on day 2

Sand Box Modeling

March – July March – Initial Setup, testing of materials Run model with no groundwater element

April – Test groundwater element Tests involving partial lateral groundwater discharge

May – June Tests involving various layers or layer pinch-outs

July – Analysis of sandbox data

Computer Model

April – July April – compile needed equations for model Initial design and begin writing model program

May – Finish writing model program Initial test runs

June – run model changing various elements and locations of groundwater

July – Compile results

Budget

Item means of acquisition Source of Funding Amount needed GPS Unit use GWI owned unit (3) waterproof temperature data loggers use GWI units Boat use DES or other 100 ft of polybraided rope purchase personal purchase $15 15 ft, 1 in diameter PVC pipe purchase ? $5 thermocouple data logger purchase ? $100 data logger interface dock purchase ? $30 J type thermocouple, 10 cm shielded purchase ? $30 30 feet, type J thermocouple wire, vinyl coated purchase ? $10 Type J thermocouple coupling purchase ? $5 Sand Box supplies 8 X 10 foot plywood purchase grant $30 water pump purchase grant $35 tubing purchase grant $15 fiberglass resin purchase grant $25 plumbing fittings purchase grant $20 screws purchase grant $5 reservoir bucket purchase grant $10 Total $335 (optional) Infrared Camera Rental rental ? $400 Total $735

Sources:

• Becker, M. W., Georgian, T., Ambrose, H., Siniscalchi, J., Fredrick, K., 2004. Estimating flow and flux of ground water discharge using water temperature and velocity, Journal of Hydrology 296, 221 – 233. • Cleland, Carol E., 2001. Historical science, experimental science, and the scientific method, Geology 29, 987 – 990. • Hudson, Paul F., Kesel, Richard H., 2000. Channel Migration and meander bend curvature in the lower Mississippi River prior to human modification, Geology 28, 531 – 534. • Jackson II, Roscoe G., 1975. Velocity-bed-form-texture patterns of meander bends in the lower Wabash River of Illinois and Indiana, Geologic Society of America Bulletin 86, 1511 – 1522. • Knighton, David, 1998. Fluvial Forms and Processes, Oxford University Press, New York. • LaFleur, Robert G., 1999. Geomorphic aspects of groundwater flow, Hydrology Journal 7, 78 – 93. • Larkin, Randell G., Sharp, John M., 1992. On the relationship between river-basin geomorphology, aquifer hydraulics, and ground water flow direction in alluvial aquifers, Geologic Society of America Bulletin 104, 1608 – 1620. • Lewin, John, 1976. Initiation of bed forms and meanders in coarse-grained sediment, Geologic Society of America Bulletin 87, 281 – 285. • Malin, Micheal C., Carr, Micheal H., 1999. Groundwater formation of martian valleys, Nature 397, 589 – 591. • Pederson, Darryll T., 2001. Stream Piracy Revisited: A Ground water Sapping Solution, GSA Today September 2001, 4 – 10. • Scott A. Wells, Robert L. Annear, Report for 2002OR2B: Temperature Effects of Streambed Heating, Water Resources Research institute Annual Technical Report, 2002, pgs 3 – 10. • Scott A. Wells, Robert L. Annear,2002. Report for 2002OR2B: Temperature Effects of Streambed Heating, Water Resources Research institute Annual Technical Report 2002, pgs 3 – 10. • Simon, Andrew, Curini, Andrea, Darby, Stephen E., Langendoen, Eddy J., 2000. Bank and near-bank processes in and incised channel, Geomophology 35, 193-217. • Smith, Charles E, 1998. Modeling High sinuosity meanders in a small flume, Geomoprphology 25, 19 - 30. • Torgersen, Christian E., Faux, Russel N., McIntosh, Bruce A., Poage, Nathan J., Norton, Douglas J., 2001. Airborne thermal remote sensing for water temperature assessment in rivers and streams, Remote Sensing of Environment 76, 386 – 398. • Toth, Jozsef, 1999. Groundwater as a geologic agent: An overview of the causes, processes, and manifestations, Hydrogeology Journal 7, 1 – 14. • Van Arsdale, Roy, Waldron, Brian, Ramsey, Natasha, Parrish, Shane, Yates, Rhonda, 2003. Impact of River Chennelization on Siesmic Risk: Shelby County, Tennessee, Natural Hazards Review, February 2003, 2 – 11.