Abstract
Tropical cyclones reduce port capacity and leave heterogeneous congestion. Yet empirical measurements of port resilience are rarely connected to operational strategy evaluation on the same observed event baseline. This study develops a dual-layer framework that measures event-level operational resilience under prevailing practice and estimates modelled marginal improvements from three operational levers. Aggregate baseline measures calibrate multi-server queues for 28 ports without event-period tuning, and discrete-event simulation replays 178 port-event trajectories. The calibrated representation places heterogeneous ports on a common queueing scale. Measured states are benchmarked against a dynamic four-hour business-as-usual (BAU) baseline that retains normal temporal variation. During exposure, median service-capacity loss reaches 0.627, and sustained operational-capacity restoration is confirmed after a median 44 h. Service-focused recovery (SES), coordinated recovery (CRS), and two-stage proactive–reactive response (TPRS) are compared after matching costs within each event. CRS gives the largest mean reduction in cumulative queue burden under the base-case cost coefficients and ranks first in 113 of 128 cost-sensitivity scenarios. SES leads when capacity coordination becomes sufficiently expensive, while TPRS becomes more competitive under prolonged, high-loss exposure. The framework supports strategy assessment according to event state, operational feasibility, and implementation cost.