Computer modelling in science
Dr Peyman Zawar-Reza, Geography Department at the University of Canterbury, uses modelling to predict the flow of particulates in the air over Christchurch. This video explains the development and use of computer modelling using past and present data to predict what the real world might be like in the future.
Transcript
DR PEYMAN ZAWAR-REZA Models give us predictive abilities, and that is really important because a lot of environmental problems right now need some sort of a prediction. If I want to see what a situation is at the moment I can just go and measure it. I can take a temperature sensor with me there, or some sort of device that measures air quality, and go and measure this in real time at the moment. Scientists also have also come up with a lot of really smart ways in trying to find out how the environment was in the past, for example, studying tree rings, by taking ice cores in Antarctica and analysing air bubbles. So we have the past and the present. How are we going to know how things happen in the future? Because it hasn't happened yet, obviously. That is where computer models come in because they can predict for us how things are going to evolve into the future. They are really good teaching tools because they are really dynamic. Students can usually see movement, and they can study environments really effectively with the models. This is not something you can do with gathering data, for example, with a thermometer if you go outside and measure something. There is only a certain amount of the real world you can explain with measured data. After that, to complete your information, you need models. One of the neat things about using models, it allows you to test your hypothesis and your knowledge because you can just act like God with your model and change anything you like. You can remove the mountains in New Zealand and see what effect that has on the climate. You can pump in artificial carbon dioxide into the atmosphere, see what effect that will have on global warming, and things of that nature. So there's a … I guess there is a lot of switches that you can turn on and off with your model to try to define your problem and see what effects they have.
Acknowledgments: Dr Katja Riedel, NIWA Cath Samson David Etheridge, CSIRO
