Modélisation de cisaillements de vent et assimilation de données dans la couche limite atmosphérique

Abstract : The objective of this thesis is to study the ability of numerical weather prediction model to represent windshears in the lower layer of the atmosphere over an airport area and to determine the impact of high temporal frequency observations on predictions. The study focuses on the international Nice Côte d'Azur airport where horizontal windshears, also called wind reversals, happen regularly. A wind pro ler and three anemometers were installed on the Nice airport runways. At the beginning of 2009, a eld campaign with a wind lidar and a sonic anemometer took place in order to provide additional observations. Both high temporal frequency data and numerical simulations performed with the mesoscale research model Meso-NH supply a general picture of the evolution of the various ows leading to a windshear event over the airport. We also use observational data to validate the model predictions for three different situations of windshear. The simulations reproduce quite well the horizontal and vertical structure of the ow despite a spatio-temporal misplacement. Local ows such as land and valley breezes are important to generate low level conditions for a horizontal windshear event but the front position is mostly in uenced by mesoscale ow (trapped gravity waves or low geopotential). We compared the Meso-NH simulations with the results of the operational model AROME and carried out sensitivity testing against initial and coupling conditions at 2.5 km resolution. We then ran experimental simulations at 500 m resolution, centered on the airport platform, to evaluate the impact of an increased resolution on the windshear predictions. Such a resolution provides improvement of local ow and generates rapid and local wind changes but does not improve the windshear front position compared to a 2.5 km resolution simulation. Moreover the small horizontal grid domain increased the sensitivity to large scale lateral boundary conditions. In order to constrain numerical simulations toward high temporal frequency observations we considered a data assimilation system based on the nudging technique called the 'Back and Forth Nudging' (BFN) technique. We fi rst applied this algorithm to the Lorenz system to compare its behaviour with published results considering other data assimilation techniques. The promising results allowed the implementation of the BFN within the Meso-NH model. We performed assimilation experiments in idealized conditions with high temporal frequency of wind pro les that show a consistent response of the model.
Document type :
Complete list of metadatas

Cited literature [13 references]  Display  Hide  Download
Contributor : Alexandre Boilley <>
Submitted on : Monday, November 7, 2011 - 2:04:59 PM
Last modification on : Friday, April 5, 2019 - 8:14:08 PM
Long-term archiving on : Thursday, November 15, 2012 - 11:25:12 AM


  • HAL Id : tel-00638822, version 1



Alexandre Boilley. Modélisation de cisaillements de vent et assimilation de données dans la couche limite atmosphérique. Météorologie. Université Paul Sabatier - Toulouse III, 2011. Français. ⟨tel-00638822⟩



Record views


Files downloads