Objective The COVID-19 pandemic has profoundly altered global influenza circulation. This study aimed to investigate the impact of meteorological factors on influenza transmission in Jiuquan, China, during three distinct phases: before, during, and after the COVID-19 pandemic. Methods Weekly influenza surveillance and concurrent weather data were modeled using seasonally adjusted SARIMAX to capture temporal and seasonal structures. An XGBoost surrogate was trained to emulate the SARIMAX outputs, and TreeSHAP was applied to decompose and quantify the relative contributions of the environmental variables. Results 2016-2019: Influenza followed the northern hemisphere winter-spring pattern, with A/H3N2 being predominant. 2020: COVID-19 non-pharmaceutical interventions suppressed transmission; positivity rates decreased sharply, and a 78-week “low circulation” persisted. Post-mitigation relaxation: 2023–24 winter resurgence. Weeks with positivity 10% increased from approximately seven pre-pandemic to approximately 11 post-pandemic. Overall positivity differed across phases (pre:25.29%; pandemic:12.07%; post:12.42%; P0.001). SARIMAX explained moderate variance pre-COVID (R 2 =0.5189), little during/after pandemic (R 2 =0.0373, 0.0671). TreeSHAP surrogate achieved high fidelity (R 2 =0.8121-0.8628). Temperature was dominant across phases: low temperatures were linked to higher positivity, and high temperatures suppressed it. Pandemic phase: Wind speed, humidity, and pressure gained importance. Post-pandemic: temperature differentials strengthened, wind and pressure distributions’ influence “broadened”. Conclusion During COVID-19, NPIs drastically suppressed influenza transmission in Jiuquan, disrupting its winter-spring cycle and causing prolonged “low circulation”. Following relaxation, immunity debt fueled stronger and longer seasonal resurgences, with subtype niche shifts; influenza rebounded but not to pre-pandemic levels after relaxation. Stable temperature effects and shifting roles of wind, pressure, and thermal gradients may reflect environmental-behavioral transmission restructuring, highlighting region-specific, immunity- and seasonality-informed control strategies.
Wang et al. (Wed,) studied this question.