| description abstract | When crashes occur, roadways often need to be closed for emergency response, which would greatly interrupt traffic operations. Many studies have discussed the road closure issue caused by crashes, but few of them have focused on road closure events on bridges. However, bridges are critical components of roadways. Their unique geometry determines whether road closures on bridges could impair traffic operations more seriously. Using traffic crash data from 2001 to 2020 in Pennsylvania, this study aims to identify the characteristics and trends of roadway closure events due to bridge traffic crashes. First, bridge traffic crashes and nonbridge crashes were compared in terms of collision, environment, driver, and vehicle features. Bridge traffic crashes are confirmed to be more likely to result in lane closures than nonbridge ones, and they also have a longer average road closure time. Compared with nonbridge crashes, bridge traffic crashes occur more in winter. In this study, a zero-inflated Poisson regression model is built to identify the factors influencing the road closure time. The modeling results show that bridge traffic crashes occurring on wet pavements, urban areas, and interstate or local roads have a significantly shorter lane closure time, whereas bridge traffic crashes involving hit-fixed-object collisions, driving under the influence, speeding, and heavy trucks would result in a longer lane closure time. The findings are expected to provide new insights for agencies to develop effective measures to reduce durations of road closures caused by bridge traffic crashes. | |
