Camera-Trap Occupancy Survey

Ecosystem monitoring rarely means watching every animal directly — instead, ecologists scatter sensors (camera traps, acoustic recorders, eDNA samplers) across a landscape and run repeated "occasions," accepting that any single visit will miss animals that are really there. This simulator lays a grid of habitat cells with a hidden, ground-truth occupancy pattern, deploys a configurable network of camera traps across it, and runs a live single-season occupancy survey: each occasion, occupied camera sites are detected only with probability p, so the naive tally of "sites with at least one hit" consistently under-counts the true occupied fraction. A detection-corrected estimator — the same closed-form correction used in real MacKenzie-style occupancy models — is recomputed after every occasion so you can watch it converge on the truth as survey effort accumulates.