Modeling in Scientific Research

In order to provide realistic output, scientific models

• a.incorporate all known variables in the real world.
• b.must be accurate physical representations of a system.
• c.use equations to describe every process within a system.
• d.simplify the real world where appropriate.

Vilhelm Bjerknes saw "no intractable mathematical difficulties" with predicting the weather, and many numerical models have been built since he wrote those words in 1904. Even with the best of those models, our weather predictions are still frequently incorrect. Why?

• a.Bjerknes was wrong: The math is too complicated.
• b.Not all aspects of weather can be described mathematically.
• c.We don't have fast enough computers to do the calculations in time.
• d.The weather system is very complex: Small changes in variables can have big effects.

Why is climate particularly well suited to research using modeling rather than other methods like experimentation?

• a.We know all of the equations describing air circulation in the atmosphere.
• b.It is impossible to do experiments in climate science.
• c.Modeling allows us to study the effect of changes in a complex system.
• d.We don't have another system with which to compare Earth's climate.

Hydrologists have developed a computer-based model to simulate flow of a contaminant through a groundwater reservoir. In order to develop their model, the hydrologists most likely

• a.collected data from the reservoir they are modeling.
• b.took an existing groundwater model and modified it to suit their needs.
• c.simplified the shape of the reservoir in the model.
• d.did all of the things listed as answers.

Physical "ball-and-stick" models are often used to represent molecules, even though scientists agree that atoms are not like billiard balls and bonds are not like sticks holding them together. Why do you think these are still used?

• a.Scientists haven't been able to come up with a better model.
• b.For some purposes, such as visualizing spatial relationships, the model is useful.
• c.Better models are available, but ball and stick models are easy to create and explain.
• d.All of these answers are correct.

The first step a scientist takes in developing a model – conceptual, numerical, or physical – is to define the system that the model is meant to represent. Which of the following statements best describes a system?

• a.A system has boundaries defined by a researcher.
• b.A system in a model should not simplify a real system.
• c.A system includes many variables, all of which are equally important.
• d.A system is a large entity with many variables and processes.

Computer models are superior to physical models because they are more quantitative.

• a.true
• b.false