Resilience metric for post-flood road network

In the present work, forty-six urban centers in North America with flood events in the past twenty years were selected for development of post-flood resilience metrics.Public domain high-resolution satellite data from pre-flood event, one year after the flood, and five years after the flood were collected and compared.Public domain machine learning algorithms for object identification were utilized to identify the post-flood pavement condition and functionality of bridges.A set of network connectivity indices were calculated for the pre-flood and post-flood road networks based on the functional status of pavement and other transportation elements.The Resilience index was derived from the change in network indices between post-flood and pre-flood events.The current work provides an approach for estimation of resilience metrics for data scarce locations.