Damien FOURNIER

Noise model for time-distance helioseismology

The basic input for time-distance helioseismology are time-series of the line-of-sight Doppler velocity. By cross-correlating these observations at different locations, we can obtain wave travel times that are the basic input data for helioseismic inversions. Due to the random excitation of the wave by convection, the data are extremely noisy. A noise model for travel times and products of travel times have been proposed and successfully tested on observations [2]. The noise covariance matrix is extremely important to interpret properly the data and to perform an inversion.

[2] Damien Fournier, Laurent Gizon, Thorsten Hohage, and Aaron C. Birch. "Generalization of the noise model for time-distance helioseismology." Astronomy & Astrophysics 567 (2014): A137.
Adaptive mesh
Noise covariance matrix between East-West and North-South travel times for a pair of points separated by a distance d. The structure of this covariance matrix is important to understand the data and to perform a reliable inversion.