Computer Science in Science PD: Agent Based Modeling of Complex Adaptive Systems - Discussion


The CAS I chose was the solar system. The agents would be all of the astronomical bodies - moons, planets, asteroids, etc. The environment would be space. The interactions would come from the roles that gravity would be playing in the system, objects getting close together or farther apart or perhaps colliding with each other.


This would be interesting to see in an agent based model as students could really explore gases that their own consumption produces as well as the effects of global temp. changes which I think are hard for many students - the polar ice caps are pretty far away. Rachael


The phenomenon I would use is population growth. The agents would be people, animals and land. The environment would be an area near the ocean. The animals would have a food source and as the population of people grew, the animal and food sources would also be monitored.


Iā€™d examine interactions within different ecosystems. Iā€™d want students to examine predator-prey interactions involving energy transfer but then be able to alter the model to examine the how abiotic factors influence agent populations.


Given the drought in California and the dry conditions in the rest of the west, it would be interesting to develop a model around hydrology. Ideally this could look at this on a small and lagers scale. The agents would be the members of the community, the agriculture, and other consumers of water as well as the sources of water. Water sources would also be agents ā€“ Rainfall, snow pack, aquifer systems, etc. I think the ideal environment would be the community the student lives in.


The model I used was to show how predator and prey relationships effect the size of the population. The agents would be the predator and the prey. The environment is the habitat where the organisms live. The interactions would be that the prey die when eaten by the predator. The more interactions between the predator and prey the more the prey would decrease. Fewer interactions would result in a decrease in the predator population.


I would use this for a unit about natural resources.

The agents would be the minerals or fresh water or the biosphere resources and humans.
The environment would be as large as the entire earth or a bit smaller as a continent or even smaller a country. This really depends on the type of information you want.
The interaction would be the rate of human use of particular natural resources and the depletion rate of those particular resources.


During the Ebola scare this past year, the students would be highly interested in what would happen if there was a case of infection at the school, so that would be the phenomenon. The agents would be the uninfected and infected people, teachers/staff, students, visitors. The environment would be the different locations in the school. Interactions between agents and environment would include whether it was the teacher or student infected and how they interact with the other people in the school. (I hope I am on the right track here in thinking).


the phenomenon that i think was a CAS would be how a viral disease like HIV or ebola infection spreads. the agents are infected people or animals. the environment is the family, community, and local region. interactions include individual behaviors and communal actions based on societal factors like laws, customs, and beliefs.


The phenomenon I chose was the Mono Lake ecosystem. The agents are algae, brine shrimp, brine flies, and migrating birds. The environment is Mono Lake, which is a large, shallow saline soda lake. In Mono Lake, the growth of algae is dependent on the salinity of the lake and the weather. The growth of algae affects the populations of brine shrimp and brine flies. The population sizes of the shrimp and flies determines whether the migrating birds will have enough to eat and reproduce. If there are too many migrating birds, they will deplete the populations of shrimp and flies, and the algae will overtake the lake. On the other hand, if there are not enough migrating birds, the shrimp and flies will overpopulate and wipe out the algae.


I love both your ideas. I am obsessed with traffic since I live in Chicago and have to endure traffic that makes no sense to me. I would like to study why jams happen and how to help the flow of traffic during rush hour, rain storms, concerts etc.


In my example of a complex adaptive system I described teaching the different cycles. For this exercise I want to use the water cycle as an example.

The agents in this example could be the three stages of water, the sun and cold. The environment can be a mountain side with cold towards the top and heat from the sun.

The interactions can be Ice at the top of the mountain melting into water running doing the side of the mountain forming a pool which continues to get heated by the sun. The pool water turns into vapor rising to the cool mountain top and turning into snow and ice.


That is a good system!

The phenomenon that I choose would be the energy consumption in a city. The agents would be the the time of day, the seasons or temperature, and the population.
Since this is my first foray into agent based modeling, I would need more interactions with these systems to get a better understanding of it.


To engage my students, I would change my original idea to focus on traffic. We have an excessive amounts of traffic-related deaths in Chicago, so this would spark their interest. The agents would be cars, motorcycles, and trucks. The environment would be the Dan Ryan Expressway (we have tons of highways but this is most familiar to my students). The interactions between the agents would be merging (?), switching lanes, and exiting.


Your idea sounds very interesting. I could see an entire unit based around this topic, pulling in several resources and allowing students to do extensive research.


I thought about having 8th graders use this for thermal energy transfers. The environment could be any medium (air, water, magma) and how energy can be transferred to the agents (molecules).


The idea I suggested in my earlier post was the environment of a Lava Lamp. Agents in the system could include the particles of the wax/oil inside, or just the material itself. Energy input could be manipulated with different wattage bulbs or thermal energy input and the resulting change of volume from expansion and the resulting change of density could be modeled. The cooling and subsequent contraction of the material could also be modeled. The focus could be on the macroscopic level or the particle level. Other variables might also be room temperature, coefficient of expansion, etc.,


That is a great idea!


The parameters would be set by the environmental conditions needed for hurricanes to form.


This lesson reminds me of the pHet simulations that I incorporate into some of my lesons. In particular, I am thinking of the wolf/rabbit population sim that allows students to change time of day, type of food, habitat, and color of fur for the rabbits. Students can manipulate the habitat to see how fur color might affect the survival of a species. When the rabbits fur blends in with the environment his/her survival rate is higher, and how that effects the wolf population. What affect does an increase in sagebrush have on the rabbits survival and how does that affect the number of wolves in the area.