Characterising solar surface convection using Doppler measurements

Abstract

The Helioseismic Magnetic Imager (HMI) onboard the Solar Dynamics Observatory (SDO) records line-of-sight Dopplergram images of convective flows on the surface. These images are used to obtain the multi-scale convective spectrum. We design a pipeline to process the raw images to remove large-scale features like differential rotation, meridional circulation, limb shift and imaging artefacts. The Hierarchical Equal Area Pixelization scheme (HEALPix) is used to perform spherical harmonic transforms on the cleaned image. Because we only have access to line-of-sight velocities on half the solar surface, we define a "mixing matrix" to relate the observed and true spectra. This enables the inference of poloidal and toroidal flow spectra in a single step through the inversion of the mixing matrix. Performing inversions on a number of flow profiles, we find that the poloidal flow recovery is most reliable among all the components. We also find that the poloidal spectrum is in qualitative agreement with inferences from Local Correlation Tracking (LCT) of granules. The fraction of power in vertical motions increases as a function of wavenumber and is at 8% level for $\ell = 1500$. In contrast to seismic results and LCT, the flows show nearly no temporal-frequency dependence. Poloidal flow power peaks in the range of $\ell - |m| \approx 150-250$ which may potentially hint at a latitudinal preference for convective flows.