Balancing health and economic outcomes has been central to pandemic policy debates, yet differentiating the impacts of various intervention intensities remains challenging. To manage COVID-19 outbreaks, China implemented an area-based risk-rating system that classified small areas within cities as low-, medium- or high-risk based on local case counts. This paper examines the effects of medium- versus high-risk designations on travel behaviour during the Zero-COVID period, using a dynamic difference-in-differences approach. It further analyses the weekly travel trajectories of 368 Chinese cities during both pandemic and reopening periods within a non-linear time-varying latent factor framework. Leveraging the latest Baidu mobility data and national risk level information, our findings demonstrate that high-risk designations significantly drive shifts toward work-related trips, and the impact tends to extend beyond the affected localities. People began reacting to high-risk eight days before extensive lockdowns, and the effects persisted for approximately four weeks afterward, longer than the typical lockdown duration. Additionally, our results reveal stratified mobility dynamics, highlighting that (a) cities in western China exhibited lower resilience during the Zero-COVID period and (b) regional disparities widened during this period but gradually diminished with reopening. This study offers transferable lessons for future pandemic preparedness, demonstrating how tiered risk designations can generate responses that extend beyond targeted zones and how regional disparities in mobility evolve through crisis and recovery periods.