Adverse weather conditions severely affect transportation networks. Decades of research have been dedicated to analyzing these impacts and developing countermeasures to reduce their negative effects on travelers and infrastructure. Recent developments in technology have enabled the introduction of intelligent transportation system applications used for network planning, safety assessments, countermeasure evaluation, and roadway operations. One such application is microsimulation modeling, which is a powerful tool used to emulate traffic flow. Agencies are interested in using microsimulation to forecast the effects on safety and mobility of adverse weather conditions; however, there is limited knowledge on how to calibrate the model to reflect different weather conditions. This paper contributes a methodology for calibrating car-following behavior required for successful development of microsimulation models. This research was completed using SHRP2 Naturalistic Driving Study (NDS) data to capture realistic driving behavior in a variety of weather conditions. This study has two primary objectives. First, calibrate the Wiedemann 1999 car-following model for a subset of NDS trips, cluster trips with similar weather conditions, and identify an optimal parameter set to represent that condition. Second, apply the optimal model parameters in a realistic microsimulation network to assess the predicted traffic flow in each weather condition. Findings support the hypothesis that the calibration of driving models for use in microsimulation results in more realistic estimations of traffic flow. Moreover, this research illustrates that the use of high resolution trajectory-level data can successfully capture weather-dependent driving behaviors.
Ship Handling in Unprotected Waters: A Review of New Technologies in Escort Tugs to Improve Safety
