Multi-Scale Deep Learning for Estimating Horizontal Velocity Fields on the Solar Surface

Abstract

The dynamics in the photosphere is governed by the multi-scale turbulent convection termed as granulation and supergranulation. It is important to derive 3-dimensional velocity vectors to understand the nature of the turbulent convection. However, it is difficult to obtain the velocity component perpendicular to the line-of-sight, which corresponds to the horizontal velocity in disk center observations. We developed a convolutional neural network model with a multi-scale deep learning architecture. The method consists of multiple convolutional kernels with various sizes of the receptive fields, and it performs convolution for spatial and temporal axes. The network is trained with data from three different numerical simulations of turbulent convection, and we introduced a coherence spectrum to assess the horizontal velocity fields that were derived at each spatial scale. The multi-scale deep learning method successfully predicts the horizontal velocities for each convection simulation in terms of the global-correlation-coefficient, which is often used for evaluating the prediction accuracy of the methods. The coherence spectrum reveals the strong dependence of the correlation coefficients on the spatial scales. Although coherence spectra are higher than 0.9 for large-scale structures, they drastically decrease to less than 0.3 for small-scale structures wherein the global-correlation-coefficient indicates a high value of approximately 0.95. We determined that this decrease in the coherence spectrum occurs around the energy injection scales. The accuracy for the small-scale structures is not guaranteed solely by the global-correlation-coefficient. To improve the accuracy in small-scales, it is important to improve the loss function for enhancing the small-scale structures and to utilize other physical quantities related to the non-linear cascade of convective eddies as input data.

arXiv comment9 pages, 5 figures, accepted for publication in A&A

References in corpus (29)

• Adam: A Method for Stochastic Optimization
• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• Deep Residual Learning for Image Recognition
• Squeeze-and-Excitation Networks
• Going Deeper with Convolutions
• Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification
• Chromospheric Anemone Jets as Evidence of Ubiquitous Reconnection
• Heating of the Solar Chromosphere and Corona by Alfven Wave Turbulence
• Convectively driven vortex flows in the Sun
• Ubiquitous torsional motions in type II spicules
• Vorticity in the solar photosphere
• Critical Science Plan for the Daniel K. Inouye Solar Telescope (DKIST)
• The Solar Internetwork. I. Contribution to the Network Magnetic Flux
• Three-dimensional Magnetohydrodynamic Simulation of the Formation of Solar Chromospheric Jets with Twisted Magnetic Field Lines
• DeepVel: deep learning for the estimation of horizontal velocities at the solar surface
• Vortices in simulations of solar surface convection
• Comparison of solar photospheric bright points between SUNRISE observations and MHD simulations
• Temporal power spectra of the horizontal velocity of the solar photosphere
• Dynamics of isolated magnetic bright points derived from Hinode/SOT G-band observations
• Evaluating local correlation tracking using CO5BOLD simulations of solar granulation
• Various Local Heating Events in the Earliest Phase of Flux Emergence
• Supergranulation as the largest buoyantly driven convective scale of the Sun
• Small-scale dynamo simulations: Magnetic field amplification in exploding granules and the role of deep and shallow recirculation
• Persistent magnetic vortex flow at a supergranular vertex
• Power spectra of velocities and magnetic fields on the solar surface and their dependence on the unsigned magnetic flux density
• Spontaneous Formation of Surface Magnetic Structure from Large-scale Dynamo in Strongly-stratified Convection
• Analysing the effects of apodizing windows on local correlation tracking using Nirvana simulations of convection
• Average radial structures of gas convection in the solar granulation
• Dynamics of Trees of Fragmenting Granules in the Quiet Sun: Hinode/SOT Observations Compared to Numerical Simulation

Cited by in corpus (2)

• Synthesizing impurity clustering in the edge plasma of tokamaks using neural networks
• New approach for analysing dynamical processes on the surface of photospheric vortex tubes