This study analyzes spatio-temporal evacuation traffic patterns based on traffic data collectedduring Hurricanes Irene and Sandy in New Jersey. The temporal analysis shows that the totalnumber of evacuees in Sandy was less than 30 percent of those in Irene. This is possibly due tothe late-summer season, differences in background traffic, as well as so called “crying wolf”effect. However, similar to the evacuation departure behavior during Hurricane Irene, peoplewho evacuated during Hurricane Sandy responded very quickly to the mandatory evacuationorder. Traffic volumes increased significantly when the official mandatory evacuation order wasissued. The spatial analysis shows that the most significant evacuation movements were locatedin the southern portion of the State, closest to the shore areas. The vast majority of thisevacuation traffic moved westward, instead of traveling northbound along the shore areas madevulnerable by the storm. These evacuation traffic patterns are similar to the typical outboundtraffic patterns seen in the shore areas at the end of a summer weekend but with departure timesthat were three to four hours earlier than usual. Moreover, in examining the evacuation trafficpatterns and collected travel time data, this study was also able to identify traffic bottlenecks thatarose during Hurricane Irene and Sandy evacuations. The results show that the bottlenecks weregenerally located at merging areas in the vicinity of on/off ramps or interchanges. Theseempirical observations can be used to improve real-world emergency response operations andevacuation models.
展开▼
