Add observation error to the simulation reflecting an abrupt/ strong shift in detectability, as might occur during a large gear change or change in taxonomic identification.
For taxonomy, could have species go from 0% detectable to w/e their baseline detectability is. So it would be some binary coefficient that was initially 0 for some species and 1 for most other (maybe 20-40% undetectable at start), and that coefficient is multiplied by the baseline detectability to yield net detectability.
Gear would be similar, except it might not be so binary. Taxonomy might not be binary (if, e.g., they hired taxonomists and a species was no longer frequently confused for its more common cousin), and a large gear change could yield a binary (100%) change in detectability. So maybe a gear change and a taxonomy change could be lumped together.
Add observation error to the simulation reflecting an abrupt/ strong shift in detectability, as might occur during a large gear change or change in taxonomic identification.
For taxonomy, could have species go from 0% detectable to w/e their baseline detectability is. So it would be some binary coefficient that was initially 0 for some species and 1 for most other (maybe 20-40% undetectable at start), and that coefficient is multiplied by the baseline detectability to yield net detectability.
Gear would be similar, except it might not be so binary. Taxonomy might not be binary (if, e.g., they hired taxonomists and a species was no longer frequently confused for its more common cousin), and a large gear change could yield a binary (100%) change in detectability. So maybe a gear change and a taxonomy change could be lumped together.