INTEGRATING FIELD OBSERVATIONS WITH PHYSICAL AND COMPUTER MODELS IN AN INTRODUCTORY ENVIRONMENTAL GEOLOGY COURSE.

[Journal of Geological Education, 1994, v.42, p. 264-271]

Andrew P. de Wet
Department of Earth and Environment, Franklin and Marshall College, Lancaster, PA 17604-3003

 ABSTRACT
I have developed an exercise in which retention/detention ponds are used to illustrate how computer models can be developed from field observations and conceptual models. In addition, the advantages and disadvantages of conceptual, physical, and computer models can be examined. Retention/detention ponds were chosen because they are relevant to introductory geology courses (flooding, soil erosion, sediment pollution), they are relatively simple systems, and they are ubiquitious.

The exercise consists of five parts: [1] field observations (this could be replaced by a video or slides), [2] the development of a conceptual model, [3] the use of a simple physical model, [4] modeling the system using system dynamics software (eg. STELLA (R) II), and [5] a report documenting the observations and results of the exercise.

The exercise combines careful observations with analytical thinking and group problem solving. It introduces system thinking and applies the scientific method to a dynamic geological process. It has proven to be an effective pedagogical approach which goes beyond standard textbook and lecture teaching.

Keywords: Apparatus; education - computer assisted; education - undergraduate; engineering and environmental geology; geology teaching and curriculum.

About the author
Dr. Andrew P. de Wet is an Assistant Professor in the Department of Earth and Environment at Franklin and Marshall College. He has a B.Sc. in geology from Natal University, South Africa and a Ph.D. from Cambridge University, England. He teaches introductory environmental geology and a senior seminar examining environmental problems using a multidisciplinary group problem solving approach. His principal research interests include regional tectonics of Greece and the Piedmont, and the application of computer technology, such as geographic information systems, to geological mapping and environmental issues.

INTRODUCTION
The understanding and use of computer models of dynamic systems is an increasingly important part of the earth and environmental sciences (Sheridan and Wohletz, 1992; Phillips, 1993). For example, global climate models are frequently referred to when climate change and the human impact on the environment are discussed (Graedal and Crutzen, 1993; Trenberth, 1992). However, students, especially at the introductory level, often do not understand the connections between the real world and these computer models. The examination of a relatively simple dynamic system, such as retention/detention ponds, can be used to bridge this gap in understanding.

This exercise is a further development in the introductory environmental geology course taught at Franklin and Marshall College over the last few years (see Merritts and Shane, 1992). It is designed to introduce students to system thinking and modeling while examining a system relevant to an environmental geology course - retention/detention ponds. A retention/detention pond system was chosen for several reasons: a) they are relevant to environmental geology as they assist in controlling flash flooding, soil erosion and soil pollution, b) they are ubiquitious and can thus be easily observed in the field, c) They are relatively simple systems that can be easily modeled with physical and computer models, and d) the material that flows through the system (water) is familiar and easily observed.

The central theme of this exercise is the use of modeling to understand a physical system. There are several reasons why models of these natural systems can usefully supplement or replace observations of a natural system:
a) Complexity: because natural systems are frequently very complex and thus difficult to understand, models may be able to simplify processes and facilitate understanding.
b) Temporal scale: many phenomena may occur either to rapidly, too slowly or too sporadically to be adequately observed in the field.
c) Spatial scale: many phenomena may occur on too small or large a scale to be adequately observed in the field.
d) Prediction: models can be used to test hypotheses and to make predictions.

In general there are several types of models including conceptual models, physical models and numerical/computer models.
Conceptual models attempt to illustrate a system, concept or phenomenon. They may be in the form of a diagram that is a stylised version of the actual system. They are generally descriptive and not quantitative. They may simplify a complex system or explain parts of a system that are not easily observed in the field.
Physical models attempt to physically reproduce a particular system or phenomenon. As the name suggests this type of model is a physical replica of the natural system but is usually reduced in size and complexity.
Numerical/computer models can describe a system/concept quantitatively and can make quantitative predictions about the system. In addition in the study of complex systems, numerical models must be used in order to understand system behaviour as a whole. This is especially true in the environmental sciences where many of the systems are dynamic and extremely complex. Once confidence in a numerical model has been acquired, the model can then be used to test hypotheses and to predict future conditions arising from changes in relevant processes: natural or anthropogenic. This ability for testing and prediction is one of the most powerful tools of models.

Why use computers?: Complex numerical models (and even fairly simple ones) would be extremely tedious to develop and run without the aid of computers. A number of software programs are available that make numerical modeling fairly straightforward and manageable (e.g. STELLA II).

Retention/detention ponds may be used to illustrate how computer models can be developed from field observations and conceptual models. In addition, the advantages and disadvantages of conceptual, physical, and computer models can be examined.

The exercise consists of five parts: [I] field observations (this could be substituted by a video or slides), [II] the development of a conceptual model, [III] the construction and use of a simple physical model, [IV] modeling the system using system dynamics software (eg. STELLA II), and [V] a report documenting the observations and results of the exercise.

The exercise has the following objectives:


a) Careful field observations combined with sketches and descriptions.

b) Analytical thinking - the students are required to develop a conceptual model of the system and then participate in the design of both a physical and a computer model of the system.

c) Application of the scientific method - the models can be used to develop and test hypotheses.

d) Understanding the dynamic nature of natural systems and how human intervention effects these systems. Demonstrations of physical processes in the classroom is a very effective teaching tool (see Cawthorn, 1991).

e) Introduction to systems thinking and modeling - an increasingly important approach to earth and environmental sciences.

f) Understanding the development of and the advantages and disadvantages of conceptual, physical and computer models.

g) Group problem solving and interactive learning - both useful pedagogical approaches that enhances the students learning environment (see D'Allura, 1991; Shea, 1991).

Part I: FIELD OBSERVATIONS
One of the critical aspects of this exercise is that the students see the connection between the real/physical world and models. The best way to achieve this is through a fieldtrip to a local site with a retention/detention pond system. An alternative is to show a video or slides of these features. Obviously, the best time to see a retention/detention pond system is during a rainfall event, as the students can then observe the water flow through the system first hand. This however is usually not possible (and frequently not desirable from the students point of view!) and the physical model does a good job of illustrating the flow in the laboratory.

I have not attempted to collect data on retention pond capacities and inflow and outflow rates during a rainfall event, infiltration rates and stream flow rates in the field. It is possible to do this but logistically difficult and not essential to this exercise. I have rather concentrated on getting the students to understand the basic layout of a retention/detention pond system and to explain what happens during a flood event. Issues such as flash floods, soil erosion and sediment pollution are discussed in the field by incorporating a stop at a retention/detention pond into a fieldtrip that examines broader issues of landuse and their impact on the environment. The students are required to spend some time in sketching the retention/detention pond system, a generic example of which is shown in figure 1.

Part II: CONCEPTUAL MODELS
In a follow-up laboratory the students are divided into groups and are required to describe their field observations and sketch a conceptual model to illustrate the retention/detention pond system (figure 2. ). They must divide the system into reservoirs and flows (the general concepts of reservoirs, flows and stocks (accumulations) are introduced in the first few lectures in the course). Each group then presents its model to the class. In the sample conceptual model the precipitation and the impermeable surface are combined into a reservoir (upper reservoir). The flow out of this reservoir (runoff) can then be varied to simulate different rainfall events. The retention/detention pond and the aquifer are also represented by reservoirs, as is the stream, for if the latter were conceptualised as a flow it would be much more difficult to model physically. A general discussion of the problem of where to place the systems boundaries is appropriate at this time. The class then discusses how the system could be physically modeled.

Part III: PHYSICAL MODELS
The physical model is an essential component of the exercise because it links the field observations and conceptual model with the often more difficult concept of a computer model. It is here that the students get to see how the system works. They are able to observe the change in the water levels in the different reservoirs, collect and present data and gain a much more detailed and analytical understanding of the system. My observations indicate that the students experience a quantum leap in their understanding largely due to the fact that the demonstration is dynamic, interactive and concrete.

The physical model comprises a series of clear plastic containers (reservoirs ) linked with plastic pipes (flows) and looks very similar to the conceptual model (compare figure 2. with figure 3 ). You may require the students to build a physical model in the laboratory (with your assistance); alternatively the instructor can build the model prior to the laboratory and use it to demonstrate the system. I recommend this course, as the instructor can eliminate the bugs before the laboratory period. It is however important to get the students to think about how to build a physical model because it forces then to think analytically about the layout and the parameters of the system including flow paths, flow rates, reservoir capacities and the height of the raised pipe in the retention/detention pond.

In constructing the physical model I used manageable stocks and flow rates (see Table 1. These stocks and flow rates are not based on field measurements but rather represent an attempt to simulate a generic system in a reasonable time (5 - 20 minutes) with a reasonably sized model. It is possible to convert these values to make them appear more realistic. For example, the time for the physical model run (seconds) could represent minutes and the stocks (measured in liters) and flows (measured in liters/second) could be scaled up by a factor of 103 - 104. But the absolute values are much less important than the relative values and relative changes in the stocks.

In the example simulation I used an initial stock of 9.5 liters in the upper reservoir with all other reservoirs empty. The approximate reservoir capacities and flow rates are shown in Table 1. For each simulation I use 8 students to collect the data (two students per reservoir). One student reads out the stock in liters in the reservoir while the other student operates a stopwatch and records the time and stock on a datasheet. The data are then plotted manually or with the use of graphing software on a computer; I use both methods in order to expose the students to the advantages and disadvantages of each. Data collected from an example simulation is shown in figure 4. The flow rates can be determined by direct measurement or from the slope of the graphs. This is a useful exercise for the students to undertake. They will soon realise that using the slope method is difficult when more than one inflow and/or outflow is affecting the stock in a reservoir. The result for a simulation without a retention/detention pond system is shown for comparison. Determining and comparing the flow rates into the lower reservoir (stream) can then be used to discuss the effectivenes of a retention/detention pond system.

It should be noted that the flow rates in the physical model are influenced by the head in each reservoir and are thus not linear (figure 4 ). In addition, the exact positioning of the outlet pipe in the reservoir influences the flow rate when the stock of water is nearly drained.

The flow rates and stocks can be varied in the physical model to simulate different rainfall events, infiltration rates and so on. The height of the raised pipe can also be varied and the optimum retention/detention pond capacity can be determined for different rainfall events. Beware, however, of flooding your retention pond! Using the physical model to simulate different rainfall events and to determine the optimum retention/detention pond capacity quickly becomes tedious. This is the time to suggest that perhaps a computer can help...!

Part IV: COMPUTER MODELING
Using field observations, the conceptual model and the physical model it is easy for the students to understand the concepts of stocks and flows and how flows and flow rates affect stocks. It would be ideal to get the students to develop their own computer models using system dynamics software. Usually time and financial constraints dictate that the process of building a model must be undertaken by the instructor. However, students can be actively engaged in the process even if the instructor is the one using the software.

I have used the STELLA II system dynamics modeling software running on a Macintosh computer to fulfill this task. The crucial aspect here is the process of development of the model and how it relates to the real world. There are many sophisticated computer models available that better describe the geomorphological and hydrological processes addressed here (see Harbor and McClintock, 1993) but they are not as useful in the demonstration of the process of model development.

Figure 5 illustrates the basic layout of the STELLA II computer model. The boxes represent reservoirs and the pipes with balloons are flows.The similarities between field sketch, conceptual model, physical model and computer model are evident. The flows are determined partly by the stocks in the reservoirs, hence the connectors between the flows and the reservoirs. The flows are quantified by simple equations that are determined by the model builder and are immediately recognizable to most people with some experience in programming (figure 6). However even those with little or no programming experience (or interest) can master this software. All other equations are generated automatically by the program. A detailed description of the software is beyond the scope of this paper and the reader is urged to contact the vendor for more information.

The results of a computer model which simulates the physical model are shown in table 3 and figure 7 . Approximate flow rates were calculated from the physical model data. Compare figure 7 with figure 4 to see the similarity between the data generated by the physical and computer model. In contrast to the physical model the flow rates in the computer model are constant and thus linear (figure 7 ). The results are not appreciably effected, but the details of the flow rates and stocks are. This is an excellent opportunity to discuss the advantages and disadvantages of different types of models.

It is surprising to most of the students that we can produce results so similar to the physical model. This is very empowering to the students. They begin to feel "we can do this, this is fun and exciting". Once you have the students attention and they find the process of learning fun and exciting then the sky's the limit! Very quickly it becomes evident that the computer model is a much more desirable tool to use to test different hypotheses, but only once the computer model and its limitations are fully understood.

Part V: REPORT
The final report gives the students practice in scientific writing, a skill that is sorely lacking in many students (see Niemitz and Potter, 1991). The report should document the field observations, development of the models, the advantages and disadvantages of the different models, and the results of the simulations. This process of writing provides the students with the opportunity to cement their understanding of the exercise and should not be foregone.

ADVANTAGES AND DISADVANTAGES OF THE DIFFERENT MODELS
Each type of model has advantages and disadvantages and this exercise is extremely effective in illustrating these aspects.
Conceptual models: Advantages: a) assist in understanding a complex concept and/or system, b) can convey the essence of a system observed in the field to someone who has not had the opportunity to make those observations, c) can clarify a complex concept/system (it may include parts of a system not easily observed).

Disadvantages: a) may tend to be too simple, b) is generally not quantitative, c) may have limited prediction capabilities.
Physical models: Advantages: a) visual, can be easily understood, b) systems that are poorly understood or extremely complex can be modeled (e.g.. river system or coastal system), and c) can lead to the development of numerical models.

Disadvantages: a) expense - building a scale model of a complex system can be extremely expensive, b) flexibility: most models do allow for some flexibility in the design, parameters, dimensions, flow rate etc. but it is frequently difficult or impossible to make major changes to a physical model, c) scale: it is usually desirable to make a reduced scale physical model of the system to be modeled, however this may fundamentally alter the processes being modeled, d) simplification: it is usually necessary to simplify the system to be modeled. This may affect the processes being modeled. Usually it is only possible to physically model a part of a complex system or individual system interactions.
Computer models: Advantages: a) cost effective: once you have the software they are easy and cheap to create, b) quantitative: parameters can be changed, experiments conducted, quantitative data produced.

Disadvantages: a) errors in simplification and parameters, b) need to understand the system fairly well before is can be numerically modeled - flow rates, stocks etc. need to be known, c) real danger of believing the results of a numerical model even if it is completely incorrect. It is essential to check the numerical model against observations wherever possible.

DISCUSSION
The exercise was developed for, and used in, a class with 20 - 24 students. The minimum requirements include a fieldtrip to examine retention/detention ponds, the physical model, a Macintosh computer with the STELLA II software and a projection pad. The exercise can however easily be adapted for use in a larger lecture situation. The fieldtrip could be substituted by a video or slides and the computer/projection pad combination is very effective even in large lecture halls (Gunter 1991, 1993). The physical model could be viewed by a video camera and projected to TV screens in the lecture hall.

The whole exercise has several potential achievements; because the students need to come up with a design, if not actually build a physical model of the retention/detention pond system, they must first understand and then think analytically about the system. They are able to see how a computer model can be developed from field observations and conceptual/physical models.

The physical and computer models are dynamic. The students can see what happens when a rainfall event occurs and understand the role of retention/detention ponds in reducing flash flooding and sediment pollution. It is rare that dynamic phenomena can be seen in the field. Video and slides frequently do not convey the dynamics of natural and anthropogenic processes, in part because they are not interactive. This exercise, however, is interactive and has evoked excitement and enthusiasm from the students as well as being an effective teaching tool.

The flow of water between stocks/reservoirs is easy to observe and the effect of inflows and outflows on stocks is easily understood. This can then be used as a building block to understand the flows of other materials that cannot be easily observed between reservoirs. For example, the flow of CO2 in the carbon biogeochemical cycle and how the buildup of CO2 in the atmosphere is a result of increased fossil fuel combustion and deforestation is readily understood by the students following completion of this exercise. Both the physical and the computer models allow the students to practice the scientific method in a interactive way. They have the opportunity to ask the question : What if.....?, create a hypothesis and test it using the models.

The development and use of this exercise was very rewarding because the response from the students was so enthusiastic. The exercise combines careful observations with analytical thinking and group problem solving. It introduces system thinking and applies the scientific method to a dynamic geological process. It has proven to be an effective pedagogical approach which goes beyond standard textbook and lecture teaching.

REFERENCES
Cawthorn, Richard Grant, 1991, Simulation of Layered Magma Chambers: Journal of Geological Education, v. 39, p. 44-47.

D'Allura, Jad Alan, 1991, Interactive Education in an Introductory Environmental-Geology Course: Journal of Geological Education, v.39, p. 279-283.

Graedel, T.E. and Crutzen, Paul J., 1993, Atmospheric Change: An Earth System Perspective: W.H. Freeman and Company, New York, 446 p.

Gunter, Mickey Eugene, 1991, In-Class Computer Demonstrations For Physical Geology: Journal of Geological Education, v. 39, p. 373-375.

Gunter, Mickey Eugene, 1993, Some Thoughts about Teaching Introductory Geology: Journal of Geological Education, v. 41, p. 133-139.

Harbor, Jonathan Martin and McClintock, Keith Alan, 1993, Teaching Applied Geomorphology with an Exercise In Urban Storm-Water Management and Erosion Control: Journal of Geological Education, v. 41, p. 38-42.

Merritts, Dorothy Jane and Shane, Edward Brendan, 1992, Effective Use of Hands-On Activities, State-of-the-Art Technology, and Computers In Introductory Environmental Geology: Journal of Geological Education, v. 40, p. 272-278.

Niemitz, Jeffrey William and Potter, Jr, Noel, 1991, The Scientific Method and Writing in Introductory Landscape-Development Laboratories: Journal of Geological Education, v. 39, p. 190-195

Phillips, Danny C., 1993, 3-D Geological Modeling: No Longer Just a Research Tool: Geotimes, v. 38, n. 7, p. 14-16.

Shea, James Herbert, 1991, A Computer-Based Subduction-Zone-Earthquake Exercise for Introductory-Geology Classes: Journal of Geological Education, v. 39, p. 388-392.

Sheridan, Michael F. and Wohletz, Kenneth H., 1992, Animated Computer Models of Volcanic Eruptions: Geotimes, v. 37, n. 7, p. 15-17.

Trenberth, Kevin E., Ed., 1992, Climate System Modeling: Cambridge University Press, Cambridge, 788 p.

Computer Software Information:
STELLA II is a computer modeling application that uses systems thinking; available from:

|High Performance Systems Inc, 45 Lyme Road, Hanover NH 03755,
Tel: (603) 643-9636, FAX: (603) 643-9502.