<p>According to the 14<sup>th</sup> Annual Road Safety Performance Index Report by the European Transport Safety Council, annually&#8239;more than 100,000 accidents occur on European roads, of which 22,660 people lost&#8239;their lives&#8239;in 2019.&#8239;The factors contributing to road traffic accidents are commonly grouped into three categories:&#8239;environment,&#8239;vehicle&#8239;or driver.&#8239;The European accident research and safety report 2013 by Volvo states in about 30% of&#8239;accidents contributing factors could be attributed to weather and environment leading for example to unexpected changes in road friction, such as&#8239;black ice.&#8239;In this work,&#8239;we are developing a solution&#8239;to forecast road conditions in Norway&#8239;by applying the&#8239;<em>Model of the Environment and Temperature of Roads &#8211;&#8239;METRo</em>, which&#8239;is a surface energy balance model&#8239;to predict the road surface temperature. In addition, METRo includes modules for water accumulation&#8239;at the surface&#8239;(liquid and frozen) and vertical heat dissipation&#8239;(Crevier and Delage, 2001). The road condition is forecasted for a&#8239;given&#8239;pair of latitude, longitude&#8239;and&#8239;desired&#8239;forecast time.&#8239;Data from the closest road weather station&#8239;and postprocessed weather forecast&#8239;are used to initialize METRo and provide boundary conditions to the road weather forecast. The weather forecasts are&#8239;obtained from the&#8239;THREDDS service and the road weather station data from the FROST service, both provided by MET Norway.&#8239;We develop algorithms to obtain the data from these services, process them to match the METRo model input requirements and send them to METRo&#8217;s&#8239;pre-processing&#8239;algorithms, which combine observations and forecast data to initialize the model. In a case study, we will compare short-term METRo forecasts with observations obtained by road weather stations and with observations retrieved by car-mounted environmental sensors (e.g., road surface temperature).&#160;This work is part of the project <em>AutonoWeather - Enabling autonomous driving in winter conditions through optimized road weather interpretation and forecast</em> financed by the Research Council of Norway in 2020.&#160;</p>
Development of Road Surface Temperature Prediction Model using the Unified Model output (UM-Road)
