Hiroyuki Masaki | Searcharxiv

arXiv2024

Detection of the solar internal flows with numerical simulation and machine learning

The solar interior is filled with turbulent thermal convection, which plays a key role in the energy and momentum transport and the generation of the magnetic field. The turbulent flows in the solar interior cannot be optically detected due to its significant optical depth. Currently, helioseismology is the only way to detect the internal dynamics of the Sun. However, long-duration data with a high cadence is required and only the temporal average can be inferred. To address these issues effectively, in this study, we develop a novel method to infer the solar internal flows using a combination of radiation magnetohydrodynamic numerical simulations and machine/deep learning. With the application of our new method, we can evaluate the large-scale flow at 10 Mm depth from the solar surface with three snapshots separated by an hour. We also apply it to the observational data. Our method is highly consistent with the helioseismology, whereas the amount of input data is significantly reduced.

astro-ph.SR↗

arXiv2023

Solar horizontal flow evaluation using neural network and numerical simulation with snapshot data

We suggest a method that evaluates the horizontal velocity in the solar photosphere with easily observable values using a combination of neural network and radiative magnetohydrodynamics simulations. All three-component velocities of thermal convection on the solar surface have important roles in generating waves in the upper atmosphere. However, the velocity perpendicular to the line of sight (LoS) is difficult to observe. To deal with this problem, the local correlation tracking (LCT) method, which employs the difference between two images, has been widely used, but LCT has several disadvantages. We develop a method that evaluates the horizontal velocity from a snapshot of the intensity and the LoS velocity with a neural network. We use data from numerical simulations for training the neural network. While two consecutive intensity images are required for LCT, our network needs just one intensity image at only a specific moment for input. From these input array, our network outputs a same-size array of two-component velocity field. With only the intensity data, the network achieves a high correlation coefficient between the simulated and evaluated velocities of 0.83. In addition, the network performance can be improved when we add LoS velocity for input, enabling achieving a correlation coefficient of 0.90. Our method is also applied to observed data.

astro-ph.SR↗