The validation and improvement of route-based road weather forecasts

This thesis aims to develop the foundations for a new validation strategy for route-based road weather forecasts that will enable validation of route-based models at a vastly improved spatial and temporal resolution, and in doing so provide a tool for rapid appraisal of new model parameterisations. A validation strategy that uses clustering techniques to create clusters of forecast points with similar geographical and infrastructure characteristics is presented, as well as two methodologies for de-parameterising key geographical and infrastructure parameters in the ENTICE route-based model that are currently not measured at the spatial scale demanded by a route-based forecast. The proposed validation strategy facilitates the analysis of forecast statistics at the cluster level, which is shown to provide a more representative measure of the model’s spatial forecasting ability. The majority of thermal variations around the study route are well represented by the clustering solutions, presenting the opportunity for new sampling strategies with the potential to validate forecasts at a vastly improved spatial and temporal resolution. De-parameterisation of the road construction and surface roughness parameters within the ENTICE model using Ground Penetrating Radar and airborne LIDAR data has been shown to significantly improve the spatial forecasting ability of ENTICE, with the model changes leading to refinement of the clustering solution which enables it to better capture the physical relationship between road surface temperature and the geographical and infrastructure parameters around the study route. Suggestions for future research are provided along with a blueprint for the future of route-based road weather forecasts.