Skip to Main content Skip to Navigation
Theses

Validation d'un modèle physique de prévision de la température de surface du revêtement de la chaussée : intégration de données in situ et de prévisions à moyenne échéance pour l'élaboration d'informations en météorologie routière hivernale

Abstract : Ensuring road safety and maintaining good traffic conditions are road managers' duties. This responsibility is being complicated during winters with harder meteorological conditions especially road-icing. To fight road-icing, it is proper to be able to foresee the temperature of roads surface. Such an approach was the framework of the studies carried out during this thesis and has been set up through a system called 'PREVIROUTE' on a semi-mountains motorway, the A75 motorway in France. The studies were based on two themes. First, the study of the system sensitivity to variables and parameters of the 24 hours term surface temperature forecasting physic model. Second, the study of the system's performances in different conditions in order to know its limits. The study of surface temperature forecasts sensitivity to variables and parameters of the model has started with the constitution of a database containing eight years of measures and meteorological forecast on the motorway A75. Then the study has been carried out in two ways. First, the connection between the forecasted meteorological variables' errors and the errors on surface temperature forecasts has been done. Then voluntary changes on the variables and parameters were made and the influence on surface temperature forecasts observed. This study allowed concluding that among the model parameters; only thermal diffusivity has a significant influence on the model. As for the atmospheric variables, it is air temperature and cloud cover which are the more important with their role in the radiation magnitude's calculation. Moreover, using the database I have evaluated the model performance in various configurations. It has been possible to distinguished three functioning phases for the model. A first diurnal phase, in which the model gives an average difference forecast of 2.5°C±3.2°C; a nocturnal phase with an average difference forecast of -0.1°C±2.1°C. And a second diurnal phase, with an average difference forecast of -1.1°C±3.8°C. I have been able to notice and demonstrate that no matter which month of forecast, road surface conditions or geographical location, these performances remain unchanged. This makes the 'PREVIROUTE' system an efficient tool, especially for the surface temperature forecast at night, a critical phase in road icing.
Complete list of metadatas

Cited literature [34 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-00688806
Contributor : Camille Meyer <>
Submitted on : Wednesday, April 18, 2012 - 2:59:27 PM
Last modification on : Friday, April 5, 2019 - 8:07:52 PM
Document(s) archivé(s) le : Wednesday, December 14, 2016 - 11:10:18 PM

Identifiers

  • HAL Id : tel-00688806, version 1

Collections

Citation

Jamal, Charles Yahia. Validation d'un modèle physique de prévision de la température de surface du revêtement de la chaussée : intégration de données in situ et de prévisions à moyenne échéance pour l'élaboration d'informations en météorologie routière hivernale. Interfaces continentales, environnement. Université Blaise Pascal - Clermont-Ferrand II, 2006. Français. ⟨NNT : 2006CLF21645⟩. ⟨tel-00688806⟩

Share

Metrics

Record views

547

Files downloads

2212