Helioseismological determination of the subsurface spatial spectrum of solar convection: Demonstration using numerical simulations

Understanding convection is important in stellar physics, for example as an input in stellar evolution models. Helioseismic estimates of convective flow amplitudes in deeper regions of the solar interior disagree by orders of magnitude among themselves and with simulations. We aim to assess the validity of an existing upper limit of solar convective flow amplitudes at a depth of 0.96 solar radii obtained using time-distance helioseismology and several simplifying assumptions. We generated synthetic observations for convective flow fields from a magnetohydrodynamic simulation (MURaM) using travel-time sensitivity functions and a noise model. We compared the estimates of the flow with the actual values. For the scales of interest ($\ell<100$), we find that the current procedure for obtaining an upper limit gives the correct order of magnitude of the flow for the given flow fields. We also show that this estimate is not an upper limit in a strict sense because it underestimates the flow amplitude at the largest scales by a factor of about two because the scale dependence of the signal-to-noise ratio has to be taken into account. After correcting for this and after taking the dependence of the measurements on direction in Fourier space into account, we show that the obtained estimate is indeed an upper limit. We conclude that time-distance helioseismology is able to correctly estimate the order of magnitude (or an upper limit) of solar convective flows in the deeper interior when the vertical correlation function of the different flow components is known and the scale dependence of the signal-to-noise ratio is taken into account. We suggest that future work should include information from different target depths to better separate the effect of near-surface flows from those at greater depths. The measurements are sensitive to all three flow directions, which should be taken into account.