Weather or terrorism-related disasters at seaports can lead to significant economic losses from vessel and cargo delays. In times of such disasters, port capacities and optimal shipment routings would change rapidly, requiring near real-time analyses for planning response operations. To address this challenge, we offer a decision support system to help port networks analyze disaster response scenarios. As part of the model, an algorithm routes arriving ships to ports to optimize the use of network capacity with respect to ocean and inland transportation, port and inventory capital costs. To reflect changing port congestion conditions, port capacities are dynamically updated in the algorithm by regression-based parametric metamodels derived from the results of a simulation model. Demonstrating application through case studies of the North American container port network, we illustrate the magnitude of cost increases from disaster-based port shutdowns and highlight the practical need for coordinated, network-wide planning for disaster response.