DescriptionTransportation systems are at the epicenter of attention when a disaster happens due to their importance in evacuating victims in the pre-disaster phase, as well as providing supplies to the survivors in the aftermath of any major disaster. When disasters happen, transportation systems are degraded by either exogenous risks, such as flooding or earthquake, or endogenous risks, such as accidents and disabled vehicles. Modeling evacuation traffic in degradable transportation networks is thus critical for public officials in developing effective hazard mitigation plans. Over the past decade, a number of emergency evacuation models have been developed. However, these models are generally developed for so-called “expected” conditions (e.g., clear weather and few accidents) without considering disruptions in transportation systems. Moreover, these models were originally developed for different emergency scenarios with specific algorithms and software tools, which limit specialized analysis of emergency scenarios. This dissertation aims to provide a new framework and methodology to model evacuation traffic by extending the use of existing regional transportation planning tools. In order to capture the occurrence of congestion during the evacuation process, a pseudo-dynamic procedure is developed and employed. Moreover, in order to evaluate evacuation planning considering the transportation system variability (e.g., incidents, accidents, and extreme weather conditions), we propose an analytical methodology and solution procedure that employs a sampling technique that randomly selects subsets of the uncertainty set to obtain an approximate solution. The proposed analytical framework and solution procedure are applied to evaluate critical transportation infrastructures in day-to-day degradable transportation networks. Moreover, we also apply the proposed analytical framework and solution procedure to evaluate the impact of endogenously determined risks in order to develop reliable emergency evacuation plans. In addition to evacuation modeling, recently Hurricane Irene (2011) made landfall in New Jersey. We have a special chapter that analyzes the empirical evacuation behavior and constructs an evacuation response curve based on traffic data collected during Hurricane Irene (2011) in Cape May County, New Jersey.