Simulation of a forest fire spreading
- Terrain with semi-random tree density generated using Perlin Noise
- Forest fire spreading = BFS algorithm, allowing nodes to be re-visited a certain number of times
- Each cell catches on fire with a probability based on its density and the number of its neighbor cells that are on fire. (Beta distribution: highest probability around 70% density, goes to to 0 at 0% density and goes to approx half of peak probability at 100% density)
- See heatMap100.jpeg for an example of randomly generated coherent tree density map, and BurnMap.jpeg for heatmap of of the fire's course, overlayed with contour lines representing the original tree density
To implement:
- Wind vector: neighbor's contribution to the probability of the cell catching fire should be multiplied by the dot-product of the vector from the neighbor to the cell and the wind vector. If the dot product is negative, the contribution is simply zero. (ie: with wind blowing towards the left, the neighbor on the left's contribution to a cell catching fire is 0)
- Burn map at different time stamps to visualize the fire propagating