Table 2. Upper and lower bounds on the objective function value for differing numbers of spread scenarios.

The mean values, $\overline{U}$and $\overline{L}$, of the objective function over a specified number of scenarios S, as well as their 95% confidence intervals, were calculated for a base case with the desired probability of successful eradication d = 0.5 in all scenarios (p = 1), the detection rate γ = 0.7 and the proportion of the site covered by surveys β = 1. The objective function minimized the expected program cost (i.e., F = 1) and was computed with sets of 20 independent replicates with increasing numbers of scenarios, S = 400, 800, 1200, 1800, 2400 and 3000.

Number of scenarios, S	Lower bound ($\overline{L}$), 95% confidence interval	Upper bound ($\overline{U}$), 95% confidence interval	Optimality gap*
400	16670611 ± 86508.3	18013682 ± 75871.6	8.06%
800	19307048 ± 89452.9	20317904 ± 88144.9	5.24%
1200	20952310 ± 89724.6	21672209 ± 48226.4	3.44%
1800	22385237 ± 90386.8	22787222 ± 30790.8	1.80%
2400	23333257 ± 91223.4	23629602 ± 55014.0	1.27%
3000	23864635 ± 92225.0	24119056 ± 48244.8	1.07%

*The optimality gap is $(\overline{U}-\overline{L})/\overline{U}$ [74].