A Comparison of Change Point Detection Methods for Pest Outbreak Detection

This study evaluates the application of change point detection (CPD) methodologies in the Agricultural Quarantine Inspection Monitoring (AQIM) program to identify significant shifts in pest arrival rates. Unlike prior work that focused on optimizing sampling strategies, this paper emphasizes the numerical evaluation of CPD algorithms for detecting pest outbreaks using AQIM data. The objective is to assess whether these methods can reliably detect and signal outbreaks when pest arrival rates exceed critical thresholds. Through extensive computational experiments, we compare the accuracy, sensitivity, and robustness of CPD methods under various outbreak scenarios and data conditions. The results demonstrate that CPD techniques, particularly cumulative sum (CUSUM) and exponential moving average (EMA), achieve high detection rates and low false alarm rates, even in challenging settings such as small outbreaks. These findings highlight the feasibility of integrating CPD algorithms into AQIM operations, providing a scalable and practical approach to enhancing outbreak detection and strengthening agricultural biosecurity in the context of global trade and evolving pest risks.