So close, yet so far. That's the best description of the current state of the Plate Tectonics code. I've got a decent feel for the parameters now, and will be exposing them through the UI so different effects can be achieved. However, there's two different artifacts caused by the erosion period parameter. Run erosion too frequently, and the mountains erode down from something like the Rockies, Alps, and Himalayas, to something more like the lesser portions of the Appalachians or Catskills. Run it too infrequently, and you get long "fuzzy" islands or island chains that remind me of caterpillars. Hours of running simulations show it is almost impossible to find a "happy medium" by simply adjusting the erosion period. I think the erosion algorithm needs to be tweaked to increase the level of erosion for sea-exposed cells and/or decrease it for land-bound cells. Perhaps a scaling factor calculated from the percentage of neighbor cells that are below sea level?
 |
| An example terrain mapped onto a globe |
 |
| Examples of "caterpillar" islands arising from too little erosion |
 |
| Examples of heavily eroded mountains, from too much erosion |
No comments:
Post a Comment