Help the Government of Canada organize its website!

Complete an anonymous 5-minute questionnaire. Start now.

Skip booklet index and go to page content

Additional Technical Guidance - How to Conduct Effluent Plume Delineation

Field Tracer Study

Field work involves tracking the dispersion of effluent using a tracer that is either naturally present in the effluent or is added. The purpose of this study is to obtain sufficient field data on a "snap shot" of effluent dispersion such that numerical modeling can then be used to estimate the maximum extent of the ≥1% effluent plume, and the long term average extent of the ≥1% and the ≥0.1% effluent plumes.


Study Team Personnel

Table 3.1 outlines the roles and responsibilities of study team personnel for conducting a plume delineation study, including a field tracer study.

3.1 outlines the roles and responsibilities of study team personnel for conducting a plume delineation study, including a field tracer study
RoleResponsibility
Team LeaderScientist or engineer who directs and represents the study team and is the liaison with local authorities and the pulp and paper mill.
Field SupervisorScientist, engineer or technician with substantial practical field experience who directs and conducts the field work.
Tracer Injection SupervisorScientist, engineer or technician who supervises injection of the tracer in to the effluent stream. This person may be accompanied by a technical assistant.
Boat Crews (one or two)At least one boat crew is required to conduct the field work. A second boat is useful to assist with drogue tracking, if conducted at the same time as the tracer study, and as a support to the main boat during the tracer study. The field supervisor may be a crew member.
Numerical modelerScientist or engineer with experience in numerical modeling of plume behaviour.

Communication

It is important to contact local authorities to notify them of the planned activities and the possible visible presence of dye in the water. Authorities to contact will depend upon the location, but may include one or several of the following: local port authority or harbour master, nearest offices of Environment Canada and Fisheries and Oceans, nearest office for the provincial departments of environment and/or natural resources, local municipal town hall or city hall, local fishing groups, non-governmental environmental groups, and possibly the local radio station. During the course of the field work, continuous communication between the team leader, the dye injection crew and the boat crew(s) is important.


Boats

The boat hull design and the propulsion unit should minimize disturbance/mixing to the plume as the vessel moves through it. While speed is generally very desirable, it will require judgement as to how much it may compromise some of the other requirements. A hull mounted recording depth sonar, combined with a GPS unit, is very desirable. Radar is also useful in coastal and marine areas.

If a dye tracer is to be used, the boat should include a firmly braced outrigger off the bow (with a restraining wire to the bow) to position the sampling intake or head of the fluorometer at a predetermined depth in the receiving water. The intake or head should be positioned such that it is clear of bow waves and is easy to detach and bring on board, or to reposition to a different water depth. Use a depressor if the intake is towed at depths greater than 2 m. In general, a system towed astern is not recommended because sampling is disturbed by the wake of the vessel; in addition, its positioning depth is very sensitive to the boat speed and length of the tow rope.

If high capacity 12/24V batteries are required for operating equipment, it is recommended to use batteries that are independent of vessel batteries, although the same charger may be used. It may be useful to have a second boat for handling drogues and for collecting bottled grab samples (if required).


Positioning (GPS)

Positioning of the sample stations, drogues and the boat track in relation to the discharge is important. Use of a Global Positioning System (GPS) unit may be the most convenient and accurate method to obtain and record positioning information. The accuracy of the GPS unit should be better than 2.5m, but this accuracy is dependent on a number of factors, including the availability of navigational satellites (maintained by the U.S. Department of Defense), environmental conditions that may result in "shading" of satellite signals, and the differential correction of signals. Differential correction on the GPS (DGPS) is provided by fixed receiver stations located at known positions, maintained by the Canadian Coast Guard for Canadian Atlantic and Pacific coastal waters, as well as the St. Lawrence Seaway. On the Great Lakes, DGPS is provided by Canadian and American receiver stations. Additional accuracy can be obtained using averaging of two or more receiver results. At some sites, more traditional survey methods, such as triangulation, may be equally effective, as long as the target accuracy is obtained.


Tracer Selection

Ideal plume tracers have the following characteristics:

  • not harmful to the environment (dye tracer)
  • near-zero background level
  • very slow decay rate (conservative substance) during field work
  • mixes freely into the effluent and receiving water
  • readily measured in the field at low concentrations
  • released at a rate proportional to the effluent discharge rate

Two types of tracers may be used:

  1. Tracers that occur normally in the effluent at known and relatively constant concentrations
  2. Tracers that are added to the effluent for the duration of the test.

The currently preferred added tracer is Rhodamine WT, a fluorescent dye that is most often used for EEM studies. It fulfils the characteristics of an ideal tracer. This dye has been shown to be non-carcinogenic and has low potential for toxicity and adverse effects in the aquatic environment (Parker 1973). It is safe when handled with care, generally available and can be readily measured in the field at concentrations less than 1 µg/L. For practical reasons, it should be obtained in liquid form. Rhodamine WT is considered conservative in most cases and typically has a near zero background level. Fluorescent tracers such as Rhodamine WT can be affected by some types of solids and chemical agents (e.g., bleaches, sulphides, sunlight, and microorganisms). Chlorine in its elemental form rapidly destroys the fluorescence of Rhodamine WT. This effect is particularly noticeable in sea water due to the supporting effects of bromine. Fortunately, elemental chlorine exists only transiently in solution. Chlorine found as NaCl in sea water does not affect the fluorescence. Preliminary tests are recommended of dye-effluent interaction to describe the stability of the tracer and to determine any loss coefficient that should be used with the tracer.

An advantage of using tracers already present in the effluent is that they have an established equilibrium in the receiving environment. Effluents from most mills contain a variety of constituents that could potentially be used as tracers for delineating the zone of effluent mixing, such as effluent colour, sodium, chloride, magnesuim, tannin-lignins, conductivity and chloroform. An evaluation of effluent constituents as tracers should consider the following: detectability, ability to measure in real time, decay rate, variability in concentration in the effluent, and variability in background concentrations in the receiving water.

Additional resources for guidance on tracer selection and use include: Feunstein (1963), Ferrier et al. (1993), Kilpatrick and Cobb (1985), and Wright and Collings (1964).


Tracer Injection (Added Tracer)

The effluent system should be inspected and the dye injection point selected. Key factors for the selection of the injection point should include the following:

  • an adequate mixing length (at least 40 times the diameter of the discharge pipe) before the final discharge point;
  • no additional discharges after the injection point; and
  • a sample access valve from which the fully mixed tracer can be sampled prior to final discharge.

The dye injection pump should be set up in the laboratory to confirm that the desired volumetric dosage rate is obtained. It is also important to determine the total time from introducing the tracer to the suction tube of the pump until it reaches the discharge. This includes the time to prime the injection system and the time for the dye to reach the final discharge point from the dosage point.

A continuous flow-rate injection system is preferred to simulate the operation of a discharge with flow proportional, continuous discharge loading. This type of injection system makes field measurements more reliable. The discharge rate of the pump versus battery charge should be monitored as well as the effect of cold temperature on battery voltage (if relevant).

For pulp mills discharging effluent in batches, mix the dye into the batch prior to release and allow for sufficient time for complete mixing. Sample the discharge pipe at regular intervals during the discharge period.

For an added tracer with zero background levels, the required tracer quantity can be calculated as follows:

M = Cx X qeff X T X %eff X 3,600 seconds/hour, where:

  • M = amount of tracer required for the test (kg)
  • Cx = tracer detection limit concentration (e.g., 1x10-9 kg/L or 1 ppb)
  • qeff = effluent flow rate (e.g., 1,000 L/sec)
  • T = duration of the test (e.g., 12 hours)
  • %eff = dilution limit of the plume in % effluent concentration (e.g., for 1:100, use 100)

The injection rate (kg/hr) is obtained by dividing the amount of tracer required by the duration of the test. The concentration of the tracer in the injection mixture (typically 20% by weight) does not have to be considered in the calculation, since the detection limit is based on the diluted initial mixture. Dilution standards are typically prepared based on weight. Should the standards be prepared by volume, the correct specific gravity of the tracer should be applied. The specific gravity of Rhodamine WT is in the range of 1.15 to 1.2, and typically 1.19.

Duration of Dye Injection

Dye must be injected over a sufficient time period to establish an equilibrium concentration in the receiving water and to give sufficient time for the field team to complete the sampling. The duration of dye injection is site-specific. As a minimum in unidirectional flow, dye injection should continue until the plume has been delineated in the field. The more dynamic receiving environments require longer injection times, particularly if the plume is found to be unstable. In lakes and rivers, injection may need to continue for several hours. In estuaries, injection should continue through at least one tidal cycle from low water, to high water, and back to low water (normally 13 hours). In coastal marine environments and fjords where already polluted water may be re-circulated back into the plume, dye may have to be injected over several tidal cycles. A judgement will need to be made on whether the time and effort is best spent continuing dye injection, or using the predictive strength of numerical modeling.

The term "slug release" is used when a known and generally small volume of dye, possibly diluted with receiving water, is introduced into the water column at the level of the anticipated plume with as little disturbance as possible. Great care must be taken to ensure that the dye release liquid has the same density as the receiving water into which it is being released. The movement and subsequent dispersion of this dye patch is monitored in a similar manner as a plume and dispersion coefficients can be computed. In this case, a secondary objective is to determine the extent of, and dye concentrations within, the dye patch at regular timed intervals. As a check on the quality of delineation of the dye patch, the quantity of dye calculated to be present at each time interval should approximate the quantity released. Slug tests will not provide an adequate description of effluent dispersion. However, they may provide useful information on localized dispersion characteristics that can then be used for numerical modeling of effluent behaviour.


Water Quality Meters

Water quality parameters and tracer concentrations should be measured in situ with the probe immersed in the water. Water quality parameters to be measured include fluorescence (dye tracers), temperature, and salinity (estuarine and marine waters). Sample location, time and immersed depth of the sampling probe must be recorded, along with visual observations, if applicable. It is recommended that as many of the desired parameters as possible be measured simultaneously.

Fluorometer

The fluorometer measures fluorescence of injected dyes, and must be in clean and reliable condition. Since the fluorometer readings must be converted to effluent concentrations, a site-specific calibration curve is required, and this will be generated in the laboratory. A typical calibration curve is illustrated in Figure 3.1.

Figure 3.1: Typical calibration curve relating fluorescence to tracer dye concentration

Typical calibration curve relating fluorescence to tracer dye concentration

The relationship between dye concentration and corresponding effluent concentration must be established prior to initiating field measurement, in order to be able to determine, in the field, at what fluorescence readings the target effluent concentrations have been found. In addition, the dye detection limit of the fluorometer in the receiving water will be determined. This concentration is required for the computation of the dye injection rate (see equation in Section 3.6).

To prepare the calibration curve, a dilution series for the dye should be established for the expected ranges of concentrations. Separate dilution series must be established for dye mixtures in the receiving water, pulp mill effluent and clean water. For saline receiving waters, at least a maximum and minimum salinity should be used. Any variations between the types of water used should be recorded and considered in the interpretation of field results.

Dilutions typically range from 0.1 x 10-9 kg/L (0.1 ppb) to 10-6 kg/L (1 ppm). The fluorescence measurements should show a linear correlation with dye concentration over the range of interest. A regression analysis will provide the mathematical expression required for converting fluorescence readings to effluent concentrations.


Equipment Used to Track Currents and Effluent Movement

Drogues and current meters are the basic types of equipment that may be used to track the movement of currents and thereby also the movement of an effluent. A description of each type is provided below.

Drogues

Drogues are used to determine movement of water in the effluent plume and also other current speeds and directions in the field. Drogues may also be used to assist in determining where to sample for tracers, particularly for sub-surface plumes. Drogues released near the discharge will drift with the current, indicating where the plume is being transported, provided it stays within the same water mass. Should the plume lie beneath the surface water mass, the drogue must be designed (weighted) to allow it to stay within the appropriate water mass. Examples of drogues are shown in Figure 3.2.

Figure 3.2: A variety of drogue types used for plume delineation (adapted from the Canadian Tidal Manual)

Variety of drogue types used for plume delineation
Drogues that are used to follow surface waters should be a relatively simple design, such as the "cross drogue" (Figure 3.2) made of two sheets of plywood weighted to keep the upper edge of plywood just below the water surface (50 cm or less). The dimensions of each vane should be not less than about 30 cm but otherwise can be constructed to a scale appropriate to conditions. Visual location of the drogue is typically marked by a flagged rod sticking out of the water and/or by a surface float, although acoustic markers may also be used (see below). The use of a surface float with some reserve buoyancy makes sure the drogue always remains at the correct depth in the water column. It is important that the wind resistance of any marker or the drag of the surface float will be minimal compared to that of the drogue itself. If there is considerable wave action or turbulence, then the distance between the surface float and drogue should be increased (see "cross drogue" with longer shaft in Figure 3.2) to ensure that the upper edge of the drogue does not flip out of the water and become influenced by wind forces. To check that the drogue is remaining within the same water mass as the plume, measurement of a simple parameter (e.g., temperature, salinity) alongside the drogue may be used.

Some drogues should be placed in the water to determine surface water movement and a few should be placed at greater depth, particularly if the plume may plunge or be entrapped at a lower layer. The drogue tracks will give visual confirmation of the likely plume track. In estuarine and marine conditions, the drogues need to be dropped in the discharge area at high water, low water and both mid tides. Since it is desirable that some of the earlier drogues are allowed to run the full course, a significant number of drogues may be required. On board plots on charts can be used to track drogue movement, which should greatly assist in recovery of the drogues.

There may be situations in which the plume may remain below the surface or may plunge later. In both cases, the plume will be moving below a less dense water mass at the surface. Some guidance can still be obtained in these cases using the same basic drogue design but taking care to weight the drogue so that it represents the same density as the displaced water of the plume. A long, slack line to the surface buoy of the drogue will enable the drogue location to be approximated. Alternatively, an acoustic locator can be attached to the drogue. These "acoustic drogues" are used to track movement of submerged plumes in the open ocean and may also be used in shallower water to estimate drift (below the threshold of most current meters) of the lower layers in a fjord or a large lake. Depending on the mixing in the plume this will suffice only for a few kilometres because the density in the plume will be constantly changing as it mixes with lighter surface and denser deeper water on its upper and lower boundaries. These types of plume behaviour are discussed in more detail in Section 5.

Current Meters

Current meters may be used to describe the hydrodynamics of receiving waters, and in particular the spatial variability in currents for use in a numerical model. Most current meters can record temperature and salinity as well, providing additional information about the types of water masses that are moving. However, current meters are typically more expensive to use than drogues in plume delineation studies.

Current meters are particularly useful for large lakes and marine waters where currents are rotary, or where wind and wave action may be the dominant current inducer. In these environments, current meters may be deployed at near-surface and mid-depth during selected seasons of the year (minimum 30 day deployment) to describe the currents. A good description of modern current meters is given in IEEE (2003).

The Acoustic Doppler current meter is a more sophisticated type of current meter that is versatile and has been used in EEM studies. There are limitations for use of this meter type for plume delineation, including the following: it is expensive; interpretation of the results requires an experienced hydrographer; poor resolution of data obtained close to the water-air and water-bottom substrate interfaces when using the full depth range (this is a problem when tracking buoyant plumes that tend to be initially in the upper meter of water).


Tracking Effluent Dispersion

Standard guidance is provided in this section for tracking effluent in a simple unidirectional flow receiving environment, using Rhodamine WT as a tracer. Additional guidance is provided in Section 5 for specific types of receiving environments: rivers, small lakes and impoundments, large lakes, estuaries and fjords and coastal marine. Use of an alternative tracer will require appropriate modifications to the recommended procedure. Details on the use of Rhodamine WT are available from fluorometer manufacturers and the U.S. Geological Survey (1996).

For a mill with multiple discharges, each discharge should be traced separately at different times to determine the configuration of each effluent plume. In most cases, the cumulative plume may be evaluated using numerical models. However, most of the mills in Canada have consolidated their effluent discharges so that only one point of discharge is typically relevant.

Sampling in the Initial Dilution Zone

Sampling in the rising plume is difficult and unproductive for the purpose of plume delineation. Instead, sampling should be focused in the area where the plume breaks the surface or is arrested in its vertical ascent. This point may be several tens of meters down drift of the discharge point. Concentrations or dilutions may vary as much as 50% around the point of emergence. Sampling should be undertaken to confirm the variability of effluent concentration at right angles to the flow of the receiving water and also parallel to the flow.

Additional sampling should be undertaken at right angles to the flow approximately 50 to 100 m down-flow from the surface plume break-out to determine the plume width, thickness and depth. From this point on, further mixing would be considered to be beyond the initial dilution zone and sampling should be as recommended below.

Subsequent Dispersion

Beyond the initial dilution zone, the effluent plume typically moves horizontally, borne by the velocity of the receiving waters. At this point, drogues released within the surface plume will provide guidance on plume location. The following description of subsequent dispersion refers to surface plumes; specific guidance for plunging or trapped plumes is provided in Section 5.4.

Sampling traverses should be conducted at right angles to the flow of the plume and at intervals of approximately 5 times the last plume width. The fluorometer or hose intake should be held at a constant depth while the boat traverses the plume. The recommended sampling depth is 1 m in homogeneous water, possibly less in stratified flow, and possibly deeper in homogeneous water if it has been determined that the plume is well mixed with depth. The depth of maximum tracer concentration is used for profiling the tracer concentration.

It is particularly important to locate both edges of the plume and to determine if the edge of the plume touches a shoreline. In the first few traverses, the boat should return to the center of the plume and the fluorometer or hose intake should be lowered to determine the vertical extent of the plume, and this should be compared with the expectation from the conceptualization of dispersion. If necessary, the boat should then return along the traverse with the fluorometer of hose intake at a deeper depth (e.g., 2 to 3m) to better delineate the lower surface concentrations.

Sampling should continue until the 1:1000 dilution (i.e., 0.1% effluent concentration) limit is determined. It should be recognized that at the far edges of the plume, sections of the plume may become separated from the main plume and form independent patches of effluent that float with the current.

Under no circumstances should a flow-through fluorometer be placed within 2 m of the bottom substrate, as this may result in equipment damage or failure. If it is suspected that the effluent plume is in or near contact with the bottom substrate, then bottled samples should be collected for subsequent analysis (see paragraph below on "grab" samples). To confirm that this is the case, at least five samples per cross section should be taken.

Any unusual characteristics in terms of plume location or concentration should be noted. Examples could include high concentrations observed at places beyond where the effluent concentration has dropped below the specified concentration (e.g., accumulation in an inlet or bay), or where an undertow moves effluent downstream or off-shore such that effluent resurfaces elsewhere.

Bathymetry should be measured and recorded at tracer profiling locations. Sonar techniques are generally adequate. Where a detailed hydrographic chart is already available, a survey of bathymetry may not be necessary. It is recommended that bathymetric data for the receiving water be presented on a map of the exposure area.

Measurements of fresh water flow and tidal water level changes should be recorded for at least 24 hours prior to and during the tracer release. At a minimum, these measurements should be taken hourly, but continuous recording is recommended.

Grab samples are recommended only if the plume is not directly accessible by the fluorometer or hose intake. The grab sampling procedure is slower, has poor spatial resolution, and does not provide a continuous profile of the tracer concentrations. Consequently, it is difficult to carry out a mass balance of the tracer. Nevertheless, it is sometimes necessary to collect grab samples. Grab samples should be collected using a water pump or by lowering sample bottles that are under vacuum pressure. Use of a Niskin or similar sample collection bottle for multiple-use is not recommended because of the risk of contamination from previous sampling. Grab samples should be stored in the dark under refrigeration, and be analyzed within 24 to 48 hours. If samples are to be collected only by this method, then at least 12 samples at each transect are recommended from within the plume to adequately define the plume configuration. It is expected that at least 10 cross sections would be sampled for any of the plumes resulting from a discharge.


Data Quality

Using the site-specific calibration curve (Figure 3.1), fluorescence readings are converted from unit- less values to dye concentrations in µg/L (i.e., ppb), and then to percent effluent. A similar unit conversion is necessary for naturally occurring tracers. Compensation for temperature changes and varying effluent discharge quantities during the field test may be required. The results should be displayed in tabular format listing time (for marine conditions, use time relative to high or low tide), water level, position, salinity (if applicable), immersion depth of the sonde or sampling tube, local water depth (optional), dye concentration and calculated effluent concentration.

A discussion on the confidence limits of the results should consider the effects of such factors as environmental conditions during the test, method of testing to measure fluoresence (e.g., grab samples, pump-through, immersed fluorometer), accuracy of the positioning data, variation in effluent discharge, and confidence in the calibration curve.


Numerical Modeling

Numerical modeling provides a means of extrapolating from the plume measurements in the field to simulate effluent dispersion over a much wider range of environmental conditions at the site. Numerical models have been developed that superimpose water quality computations onto hydrodynamic processes. The models allow for a qualitative and graphic representation of the transport and dispersion of the effluent (tracer) in time and space.

Depending on the nature of the receiving water, two or three-dimensional models may be used. Principal processes in numerical modeling include model setup, model calibration and verification (both of which use the field tracer study results), and a sensitivity analysis to determine what limitations may have to be placed on input parameters. While model calibration should be carried out against the most recent tracer measurements, verification can be achieved by applying the model to an historical event with different environmental and discharge conditions. After successful calibration, verification and sensitivity analysis, the model is then ready to be applied to a number of environmental conditions that may lead to effluent plume dispersions other than the dispersion observed during the field measurements. A discussion should be provided on the frequency of the various environmental conditions being considered.

As a minimum, the models used should be able to accurately reproduce the hydrodynamic process of the study area and the behaviour of a conservative substance introduced near the discharge. The initial near field dilution may be computed using descriptive models, such as Cormix or Visual Plumes. The tracer concentrations computed for the edge of the near field can then be used as boundary conditions for the far field model. Other typical boundary conditions for far field models are the fresh water input on the upstream end and the water level on the downstream end. In river systems with fast flowing currents, the models have to be able to simulate sub and supercritical flow conditions. In tidal waters, where the shore line near the discharge changes from high to low tide, the model should be able to reproduce wetting and drying of tidal flats.

The near field and two-dimensional far field models are readily available and have been commonly used for effluent plume delineations of pulp mill effluents. The three-dimensional models are very costly to purchase and require a significant effort for data collection, model setup and calibration. Some hydrographic research institutes are well equipped for applying three-dimensional models.

Table 3.2 lists potentially applicable models for a variety of effluent discharge scenarios, and includes models that are commercially available and routinely used. This table is intended as a preliminary guide only because models are constantly being developed and modified, and because there are many models not listed that are developed for in-house use only, and are therefore not commercially available.

Table 3.2: Numerical Models for Describing Effluent Dispersion

Numerical models for describing effluent dispersion
Typical ScenarioInformation Needs for EEMExamples of Commercially Available Numerical Models1
Plume is highly transitory or there is rapid plume dilution in the initial dilution zone to within target level (i.e., 1% effluent concentration)Conceptual spatial delineation of 1% and 0.1% limits of effluent concentration. Whether the 1% concentration limit is reached within 250 m from the discharge.Numerical models such as Cormix and Visual Plumes for the initial dilution assessment only
Effluent is discharged into turbulent, narrow stream; complete mixing is achieved rapidly over a short distanceLength and width distance until plume as dissipated to target levels (i.e., 0.1% and 1% effluent concentration)1D numerical models, such as HEC-5Q, Qual 1E , WASP5/Dynhyd5
Effluent is discharged into uniform, wide body of water. No stratification is observed.Length and width distance until plume as dissipated to target levels (i.e., 0.1% and 1% effluent concentration)Numerical models such as Cormix for initial plume dilution, 2D numerical models, such as RMA 2/RMA 4, Qual 2E, MIKE 21, for subsequent dilution simulations
Effluent is discharged into non-uniform, wide body of water. Stratification is observed as result of thermal or salinity differences in the receiving water or between the effluent and the receiving water. Stratification may be non-uniform and dynamic.Length, widths and depth dilutions until plume as dissipated to target levels (i.e., 0.1% and 1% effluent concentration)Numerical models such as Cormix and Visual Plumes for initial dilution assessments, only. 3D numerical models, such as RMA 10/RMA11, WASP5/Dynhyd5, MIKE 3, TELEMAC, DELFT 3D for far field simulations.

1Sources for these models vary; some may be obtained directly from the model developers while others may be obtained from one or a number of commercial distributors.

The selected numerical model should be used to estimate the desired regions of maximum extent, average conditions, and minimum dilution. A discussion on the confidence limits of the results should be provided (see Section 3.10).

If current meters are being used to measure ambient currents (see Section 3.8.2), the modeled plume configuration may be modified through statistical analysis of the current meter data. For conditions of a long flushing period, the measured dye concentrations may be used to calibrate a numerical transport- diffusion model. The model can then be used to simulate the effluent delineation and characteristics arising from a continuous discharge. The model can be run for a variety of conditions (e.g., seasonal variations of water movements and wind patterns), thereby overcoming the limitations of the particular conditions recorded in a single field study.

Additional resources for guidance on numerical modeling include: Baumgartner et al. (1994), Chung and Roberts (1998), Ettema et al. (2000), Frick et al. (2000), and Sharp (1989).