3877:13224Share edittextuser=1087 post_id=13224 initial_post_id=0 thread_id=3877
« Climate Models Discussion
Best ways to introduce students to models
My concern is that climate models get complex very quickly. Still, at the core of every model, including GCM's, are a few basic equations. I think students should not be given a climate model until they understand those basic equations. This may limit the use of these models, at least as far as I am concerned, to more advanced classes.
I guess that I am struggling with complex climate models being offered up as "black boxes", as V. Ramanathan once lamented, without the student really knowing what is going on inside. However, Bob Mackay's approach of using those feedback loop diagrams could offer a solution. Building a model, one part at a time, with students developing a conceptual model of what each component is doing, offers me some hope.
However, I also don't want to teach my entire course on this one topic, so how do I do all of this in a few weeks of instruction?
I probably didn't answer the question posed here, but students are going to resist using something they don't understand, and will not have any faith in the predictive powers of the model until they are comfortable with it and embrace it.
3877:13229Share edittextuser=4071 post_id=13229 initial_post_id=0 thread_id=3877
To answer the question: I am not deriving a lot of equations in class (except for the planetary energy balance and the importance of the atmosphere) and students are actually excited about using the models. The resistance is when higher math is involved.
3877:13231Share edittextuser=1892 post_id=13231 initial_post_id=0 thread_id=3877
I agree that students, at least my students, would be more resistant to learning about models the more math that is involved. I also agree that it's best for students to understand the math that goes into these models, but that is not something that I will achieve with most of my students so I am considering ways around the complicated math for now. The level of detail about the models will obviously be tied to the lesson goals – with my introductory students, I would be more likely to introduce model outputs and have them reflect on what those tell us. My science majors (most of them) would be able to consider some of the math, but it's more important for my course goals that they understand the processes that go into the model (e.g., feedbacks) rather than the math behind it. (I personally love math, but my students do not share my passion for it.) The interactive model that Bob MacKay uses would keep my students interested – it seems that most students become engaged when we do computer-based labs, and this would certainly fit. But I can also see them becoming overwhelmed with the model at first, so I might first introduce students to the "very, very simple climate model", which I have used before – this one has students input a CO2 emission rate and then step forward to 2100. It's very basic but a good place to start because it would allow me to gauge whether the students can interpret this basic output based on what they put in. I would like to use the MacKay model in my majors Climatology course, and the way he has students make predictions prior to running each model, then testing their predictions, would likely keep them engaged. If students change one parameter at a time, that should keep them from getting too overwhelmed. I think models lend themselves to hands-on activities and inquiry-based lessons, but I envision having my students complete these activities with heavy teacher involvement so that frustration levels may be monitored during the lesson. But I do think that allowing students to interact with a computer based model is a good way to keep their interest and give them an understanding of how complicated models can quickly become – but that they still provide useful outputs that students themselves are able to interpret.
3877:13232Share edittextuser=4072 post_id=13232 initial_post_id=0 thread_id=3877
i also wonder if this could be an opportunity for jigsaw, where students focus on one aspect/element of a simple climate model with peers (understanding the impact of a variable in real terms, natural variability, reasonable estimates of past/current/future values), then reorganize to evaluate a climate model from soup to nuts. or at least from soup to salad. this also mixes it up with more and less math-oriented or geo-oriented students having something to add to the small group conversations and helping catch others up.
3877:13237Share edittextuser=15868 post_id=13237 initial_post_id=0 thread_id=3877
I started by teaching them systems thinking concepts for problem framing to chose our model topic. The next week I taught them about parameters, decision variables, and feedback, and causal loops. They dove into the modeling with a tool similar to Stella. I heard a lot of grumbling the first week and confusion. I dedicated a full 3-hour session to trouble-shooting live with the class on their model the next week and put them back on task to improve the model. This 3rd week they seemed proud of their progress and are asking good questions about how to move the model onward.
Can this be called experiential or collaborative learning in the classroom?
3877:13240Share edittextuser=4082 post_id=13240 initial_post_id=0 thread_id=3877
I like Liz's description of having the students working with a model change one parameter at a time. I imagine that the model responses would be easier to interpret, and probably be fairly intuitive too.
3877:13247Share edittextuser=3761 post_id=13247 initial_post_id=0 thread_id=3877
Afterwards, we discussed the model and the results. This is a different approach from covering the model details first, I let them work with data output first then we discussed the details of modeling. Many of the students were able to question the results based on what they had learn earlier in the course. I wanted to show them that models were not "scary" but quite useful tools.
I agree that blackboxes can be dangerous, but they are still useful. We use "blackboxes" everyday without fully comprehending the details inside. For example, we drive cars, fly in airplanes, use computers and the internet, but we do not completely understand everything inside these devices that make them work!
3877:13254Share edittextuser=4073 post_id=13254 initial_post_id=0 thread_id=3877
3877:13256Share edittextuser=1087 post_id=13256 initial_post_id=0 thread_id=3877
I would like to know more about the "models" used in classes to which people are referring. Are there others besides stella and excel?
3877:13259Share edittextuser=4086 post_id=13259 initial_post_id=0 thread_id=3877
(I also linked it to the Resources page with a link to a summary of the Forecast Factory) - look under "Other resources" -> Articles and Books
3877:13283Share edittextuser=537 post_id=13283 initial_post_id=0 thread_id=3877