Doug Radford
dougradford.bsky.social
Doug Radford
@dougradford.bsky.social
4/4 - If you're wondering how we evaluated so many 1,000s of different fuel treatment plans and their effect on reducing burn probability across a landscape, then you will be interested in reading the 2nd paper of my PhD, which explored "metamodelling" of burn probability.

doi.org/10.1016/j.je...
Redirecting
doi.org
October 31, 2025 at 12:19 AM
3/4 - A key message is that what makes a “good” decision will depend highly on (1) how we define our objectives, and (2) how we navigate trade-offs between objs. Optimisation can help to explore decisions, but the process of defining what “good” looks like is values-based and can't be out-sourced.
October 31, 2025 at 12:15 AM
2/4 - A great feature of MOEAs is that they are very flexible. We can incorporate a range of simulation models to represent different objectives. We optimised fuel treatment plans to reduce burn probability, exploring many thousands of different plans to find those that best achieved the objective.
October 31, 2025 at 12:11 AM