Schematics

These are general overviews of the processes we use to simulate erosion and to determine the most vulnerable areas of the coastline. We use a 2D array to represent plots of land.

To simulate erosion, we do the following:

• Iterate down each column.
• If the value is greater than 0...
  • Check each of its neighboring coasts.
  • Subtract some constant $a$ if there is water in the row above the cell.
  • Subtract some smaller constant $b$ for each neighboring column containing water.

Here is how one timestep would look on this table, setting $a = 0.2$ and $b = 0.1$.

Note that our erosion constants $a$ and $b$ are reduced if there is a tree in neighboring cell.

For tree placement, we do the following:

• Loop over each cell in the array.
• If the value = 0...
  • Check each of its neighboring coasts.
  • If all four coasts contain land, place a tree.
  • Subtract 1 from the number of remaining trees.
• Loop over each cell in the array again, adjusting our parameters.
  • If three coasts contain land, place a tree.
  • Subtract 1 from the number of remaining trees.
• Loop again...
  • If two coasts contain land, place a tree.
  • Subtract 1 from the number of remaining trees.
• Loop for the last time!
  • If any coast contains land, place a tree.
  • Subtract 1 from the number of remaining trees.

Here is where our algorithm would plant three trees on this coastline array.

!jupyter nbconvert --to html Schematic.ipynb

[NbConvertApp] Converting notebook Schematic.ipynb to html