Model Description

Our project models a forest fire scenario with real life variables such as the saturation factor of the trees involved. The red trees are burning, the grey trees have been burnt, the green trees are dry, and the blue trees are heavily saturated from close proximity to bodies of water. Finally, the orange trees are redwoods that have thick bark to protect themselves from nearby fires. The forest fire will always find it much harder to get through the wet part of the forest as the heavy water content will most likely deter it.

Nearest Neighbors

The Nearest Neighbors problem is a scenario in which a tree can only catch on fire if one of its nearest neighbors (the tree directly above, below, or to either side) is burning. To make this occur, we had to create two separate processes. The first process, the check, instructs all the trees to check whether or not one of its nearest neighbors is on fire. The second process, the change, instructs all trees that responded to the check positively to, with some percent chance, be set ablaze.

What We Learned

From this model, we learned that saturation rates and inherent protection are very important factors when it comes to forest fires. We also learned that forest fires are very unpredictable. In terms of programming and modeling, we both learned that the two step process of check and change can improve the "real life" modeling quality of a model.

Agents and Shapes

Websites/Software Used

For this project, we used an AgentCubes model and a JavaScript model based off of it.

Sources