Abstract: Objective To investigate the utility of the Moving Epidemic Method(MEM) in establishing an epidemic warning model for various influenza virus subtypes in Yunfu, thereby providing an evidence-based reference for influenza prevention and control strategies. Methods Weekly virological surveillance data for influenza were collected from three national influenza sentinel hospitals in Yunfu, spanning from week 36 of 2014 to week 35 of 2023. The MEM was employed to construct models based on the nucleic acid detection positivity rates for influenza A(H1N1)pdm09 subtype, A(H3N2) subtype, and B(Victoria) lineage, respectively. Epidemic and intensity thresholds were subsequently established for each virus subtype. The resulting models were then utilized to assess the epidemic activity of these influenza subtypes during the 2023/2024 influenza season. Results The MEM-derived epidemic thresholds for the nucleic acid detection positivity rates of influenza A(H1N1)pdm09, A(H3N2), and B(Victoria) were determined to be 5.87%, 14.58%, and 6.67%, respectively. The A(H3N2) subtype exhibited the highest intensity thresholds across all levels, followed by the B(Victoria) lineage, whereas the A(H1N1)pdm09 subtype demonstrated the lowest intensity thresholds. Assessment of the 2023/2024 influenza season using these models indicated that the A(H1N1)pdm09 subtype reached a high level of epidemic activity, the A(H3N2) subtype remained at a low level, and the B(Victoria) lineage achieved a moderate level of epidemic activity. Conclusion The MEM is a viable tool for the epidemic surveillance and early warning of different influenza virus subtypes in Yunfu

Key words: Influenza, Moving epidemic method, Epidemic intensity, Early warning