The Earth's thermal evolution is largely determined by mantle convection, of which large mantle plumes, like the one found under French Polynesia, are an important element. The details of the mantle structure beneath French Polynesia contain important information about the Pacific super-plume. I propose to make a high resolution velocity, attenuation and anisotropy model of the mantle under French Polynesia. My research will be directly related to the recently funded multi-disciplinary PLUME project, which includes the deployment of 10 three-component broad-band seismometers for a period of two years in French Polynesia. My proposed project will involve both data analysis and improvement of the existing analysis techniques. I will start by using existing data from broad band digital seismographs in the Pacific region and the automated multimode waveform inversion technique of Debayle (1999) to obtain a more detailed model for the whole Pacific than currently exists. I will also integrate two-station multimode differential seismogram techniques into the surface wave tomography methods, which will further improve model resolution under the target area. I will then use these improved tomographic methods and the new data from PLUME to build a high resolution seismic model of the mantle beneath French Polynesia. This model will be a crucial input for the other geophysical and geological studies within the PLUME project.