Due to the increasing demand for automation and optimization of the winter service and, more recently, also with regard to autonomous driving, the requirements for small-scale predictions of the road weather, which can represent the route-specific properties, also increase. In project FE FE040279 'Route-related sleet prediction', the task was to determine the current state of the art regarding road weather forecasting and to clarify the extent to which road condition detection by mobile sensors, together with stationary road weather stations, are suitable to map the different route properties and to improve the predictions. For this purpose, methods for the more precise prediction of the winter-conditioned sleet for the individual sections within a road network to be treated were investigated on three selected test routes (A4, A9, St2139) over two winters (2015/16 and 2016/17). The main factors were the road temperature, the water film height and the dew point temperature. For this, (1) weather models, (2) point forecasts, (3) roadway forecasts, (4) route forecasts and (5) temporal variability (gradients) were compared and new concepts were tested in a 5-step process. (Author/publisher)
Abstract