TABLE 1. Questions explored using the simulation framework and the parameters relevant to each experiment. Results of each experiment are summarized with reference to key figures. The ‘*’ symbol signifies parameters that are fixed across all simulations.
| Question | Fixed parameters | Varied parameters | Result |
| Do simulations produce similar assemblages as found in the empirical fossil assemblages? | Number of specimens* Species abundance distributions along the gradient* |
DCA value of the simulated assemblage | DCA scores are accurately replicated, richness and evenness are comparable to empirical assemblages, but overestimated |
| How does excursion magnitude and resolution potential affect event detection? | No bioturbation 0.001x Transition 100% Completeness |
Excursion magnitude Resolution potential Background value |
Increased resolution potential and increased excursion magnitude increase accurate event detection |
| How does the background value alter event detection? | Background value does not affect event detection at similar relative magnitudes | ||
| How does the abruptness of the transition between background and event affect event detection? | -0.5 Background No bioturbation 100% Completeness |
Transition duration | Increased transition duration increases accurate detection of events |
| How does sampling completeness affect event detection? | -0.5 Background | Sampling completeness | Event detection is reduced with decreased sampling completeness |
| How does bioturbation affect event detection? | -0.5 Background 0.001x Transition 100% Completeness |
Complete mixing of assemblages over 10 cm mixing depth | Bioturbation decreases event detection at low resolution potentials, but only affects recovered excursion magnitude at high generating event magnitude |
| How does sample mixing affect the observed event value? | -0.5 Background | Proportion of event assemblage Event value | DCA-1 values approach the true event value as the contribution of the event assemblage increases |
| Does a small pilot data set produce the same estimates as the full data set? | -0.5 Background 0.001x Transition 25% Completeness |
Sample thickness Event duration | Yes, simulations based on the smaller dataset produce similar estimates of event detection accuracy |
| What is the expected ability to resolve 100-year-long events over the Gulf of Alaska record? | 3 cm sample thickness -0.5 Background 25% Completeness 0.001x Transition |
Sedimentation rates | 100-year-long events are accurately detected when sedimentation rates are high, but excursion magnitudes are underestimated when sedimentation rates are low |
