Optimal sampling strategy for probability estimation: An application to the Agricultural Quarantine Inspection...

Area of Science:

• Agricultural Science
• Operations Research
• Risk Management

Background:

• Imported agricultural pests pose significant threats to agriculture, food security, and ecosystems.
• Accurate risk assessment is crucial for effective agricultural quarantine and inspection policies.
• Current inspection protocols often rely on container-level heuristics, potentially limiting accuracy.

Purpose of the Study:

• To develop an optimized sampling strategy for pest detection at ports of entry.
• To introduce a pathway-level analysis for improved risk estimation aligned with the Agricultural Quarantine Inspection Monitoring (AQIM) program's goals.
• To address the tradeoff between sampling containers versus boxes per container under resource constraints.

Main Methods:

• Formulation of an optimization model to minimize the mean squared error of pest probability estimates.
• Derivation of an analytical solution for the optimal sampling strategy using approximations.
• Application of the model to a case study of maritime cargo sampling at the Port of Long Beach.

Main Results:

• The optimal strategy consistently favors sampling more containers and fewer boxes per container compared to the current AQIM protocol.
• The accuracy improvement of the proposed strategy is amplified when pest statuses within the same container are highly correlated.
• The study quantifies the benefits of the new approach across various parameter settings.

Conclusions:

• A pathway-level sampling approach offers superior pest risk assessment accuracy.
• Recommends shifting from a container-centric to a more granular sampling strategy.
• Suggests collecting box-level inspection data to refine pest correlation estimates and improve future AQIM operations.