CALMET 99 Simulation Software in Meteorological Education

Presentation

A traditional teaching approach of physical processes, based on its mathematical representation, does not necessarily attain successful learning.

Students sit in basic courses of Physics and Mathematics in the Facultad de Ciencias Exactas y Naturales of the Universidad de Buenos Aires. Biology students take the Bioclimatology Course on their fourth university year, and usually by then, they "have forgotten the physical-mathematical reasoning" taught in previous physics and math courses. They do not hold the understanding of physical processes through their abstract representation.

Meteorology and Oceanography students study general concepts of physics, algebra, calculus and physical laws of the atmospheric motions, approximations and simplified models of the weather and climate phenomena during their basic cycle. The complex equations systems and the big quantities of variables make difficult to handle equations through manual calculation. Students act as "formula manipulators" missing physical concepts. Often, they acquire a quantitative knowledge of meteorological concepts without the necessary conceptual comprehension that enables a real understanding.

It is interesting to analize an example given by Gardner ( 1993), about DiSessa's computerized game. The player gives orders to a ¨dinaturtle¨ that moves and has to kick a ball. Students have to apply Newton's laws to win the game, they have to predict how the turtle moves after kicking the ball. Their common belief is that objects always move in the direction of the most recently applied force. Elementary school students and Physics university ones failed to take into account the speed of an object in motion when trying to predict how the application of an impulse force would affect its trajectory. Both groups used the same strategies unrelating the game with the physical laws studied before. Students applied their previous ideas, only after they repeatedly failed to win the game with their misconceptions, they used the physical laws. To remember demonstrations, definitions and equations seems to be enough to assess an exam, but is not sufficient to invoke the physical learning applied to a particular case. Teaching practices should therefore point toward this direction.

Use of the software

The use of visually oriented software for simulating dynamic systems is proposed to provide students with a better comprehension of the concepts. This software supplies a rich set of visual tools for making explicit the structural relationships that reflect the behaviour of a dynamic system.

Martin (1997) defines computer simulation as the imitation of system behaviour through numerical calculations performed by a computer on a system dynamics model that represents the structure of the system.

The computer simulation program provides a framework and an easy-to-understand graphical interface for observing the quantitative interaction of variables within a system. The key idea underlying this approach is the notion that complex phenomena can be better understood by examining the behaviour of the system and how it changes over time. One of the advantages of the modelling dynamic software is the ease with which both the structures of the model and its parameters can be examined and modified by the user.

In a simulation, students specify the initial conditions and the computer shows the behaviour of the different model variables over time. Simulation gives students the power to analyse the behaviour of the system under many different conditions. Using pre-built models, students manipulate individual parameters of the model and then observe subsequent changes among other variables in the system. Building a model, students must decide which variables are important, identify the causal relationships between the variables, quantify those relationships, and test the validity of their model.

By approaching problems with a computational model, the concepts can be introduced qualitatively and decoupled from the mathematical background necessary for the closed analytic form solution. Thus, modelling provides an opportunity for students to express their own conceptual understanding of physical phenomena, using visual representations of both, quantities and their causal relationships.

The power of simulations created for the purpose of facilitating conceptual change is linked to choices made about which objects, objects attributes, and relations of the referent domain will actually be represented on the screen, the way they are represented and controlled in the simulation (Snir, Smith and Grosslight, 1995). In this case the software is controlled by the mathematical relationships embedded in the structure.

Description of the software

The computer simulation program provides a framework and an easy-to-understand graphical interface for observing the quantitative interaction of variables within a system. The interface is a window used for constructing new models, or for modifying, navigating, and simulating existing models. The structural elements of the model are created with four basic building blocks: the stock, the flow, the converter and the connector. The connections among the blocks define the causal relationships and feedback loops among model variables and the initial conditions for the starting state of the model. One of the advantages of the software is the ease with which both the structure of the model and its parameters can be examined and modified by the user.

The stock or level icon is a generic symbol for anything that accumulates or drains. The flow represents the rate at which the stock increases or decreases. Flows fill or drain stocks, the flow can be uni-directional or bi-directional and the flow within the system can be either conserved or non-conserved. Converters are used to make explicit the details of the logic that controls a flow regulator. They can hold constants, define external inputs to the model, and calculate algebraic relationships using built-in functions and graphical functions. The final building block is the connector, which links the components in the model to each other. The symbol "cloud" defines the boundaries of the model. They define what is relevant to consider, and what is not. In this case, we used VensimPLE 32, an academic free version available in Internet.

Multiple representations

The visual interface of the diagram window has particular importance as a "conceptual representation of the model" that supports the expression of students' own ideas and, at the same time, allows them to experiment with those ideas. Students must make explicit their own understanding of the structure of the system and the underlying causal relationships. The underlying activity involved in modelling is to formulate, test and revise hypothesis about relationships within a system.

The software generates multiple representations of the modelled system since the analysis tools review tree-type graphical representation, equations, definitions, and units of measure for the whole model and display the results of a simulation run over time as a graph and a table of values. The multiple-linked representations help: enriching dialogue with self, raising conflict and surprise, affirming (if not paralleling) students' expressed models and the ones they act on (Goldenberg, 1995).

Pedagogical innovations

The software enables a good comprehension of the concepts if it is used to:

confront student's erroneous beliefs about physical phenomena with correct concepts through active student's participation,

apply theoretical concepts learnt in class to problem-solving.

To reach this goals professors have to:

promote the students hands-on involvement that is essential to internalising the ideas and establishing them as mental models,

change from a role of central knowledge authority to one of guide and facilitator,

develop, implement and support new teaching strategies.

Students have to:

have an active attitude,

accept and practice the innovations profesors propose.