Mimicking spectropolarimetric inversions using convolutional neural networks

Abstract

Interpreting spectropolarimetric observations of the solar atmosphere takes much longer than the acquiring the data. The most important reason for this is that the model fitting, or "inversion", used to infer physical quantities from the observations is extremely slow, because the underlying models are numerically demanding. We aim to improve the speed of the inference by using a neural network that relates input polarized spectra to the output physical parameters. We first select a subset of the data to be interpreted and infer physical quantities from corresponding spectra using a standard minimization-based inversion code. Taking these results as reliable and representative of the whole data set, we train a convolutional neural network to connect the input polarized spectra to the output physical parameters (nodes, in context of spectropolarimetric inversion). We then apply the neural network to the various other data, previously unseen to the network. As a check, we apply the referent inversion code to the unseen data and compare the fit quality and the maps of the inferred parameters between the two inversions. The physical parameters inferred by the neural network show excellent agreement with the results from the inversion, and are obtained in a factor of $10^5$ less time. Additionally, substituting the results of the neural network back in the forward model, shows excellent agreement between inferred and original spectra. The method we present here is very simple for implementation and extremely fast. It only requires a training data set, which can be obtained by inverting a representative subset of the observed data. Applying these (and similar) machine learning techniques will yield orders of magnitude acceleration in the routine interpretation of spectropolarimetric data.

arXiv comment12 pages, 7 figures, accepted for publication in A&A

References in corpus (18)

• VFISV: Very Fast Inversion of the Stokes Vector for the Helioseismic and Magnetic Imager
• Numerical simulations of quiet Sun magnetism: On the contribution from a small-scale dynamo
• A Unified Computational Model for Solar and Stellar Flares
• An Open Source, Massively Parallel Code for Non-LTE Synthesis and Inversion of Spectral Lines and Zeeman-induced Stokes Profiles
• Inversion of the radiative transfer equation for polarized light
• Non-LTE inversions of the Mg II h&k and UV triplet lines
• DeepVel: deep learning for the estimation of horizontal velocities at the solar surface
• Deep Learning of Quasar Spectra to Discover and Characterize Damped Lya Systems
• Spectropolarimetric NLTE inversion code SNAPI
• The usefulness of analytic response functions
• Recovering Thermodynamics from Spectral Profiles observed by IRIS: A Machine and Deep Learning Approach
• Enhancing SDO/HMI images using deep learning
• Stokes Inversion based on Convolutional Neural Networks
• Comparison of inversion codes for polarized line formation in MHD simulations. I. Milne-Eddington codes
• Real-time multiframe blind deconvolution of solar images
• A new MHD-assisted Stokes inversion technique
• RADYNVERSION: Learning to Invert a Solar Flare Atmosphere with Invertible Neural Networks
• Using the infrared iron lines to probe solar subsurface convection

Cited by in corpus (12)

• Waves in the lower solar atmosphere: the dawn of next-generation solar telescopes
• Inference of the chromospheric magnetic field configuration of solar plage using the Ca II 8542 Å line
• Machine learning initialization to accelerate Stokes profile inversions
• SunnyNet: A neural network approach to 3D non-LTE radiative transfer
• Spectropolarimetric Inversion in Four Dimensions with Deep Learning (SPIn4D): I. Overview, Magnetohydrodynamic Modeling, and Stokes Profile Synthesis
• Inferring Line-of-Sight Velocities and Doppler Widths from Stokes Profiles of GST/NIRIS Using Stacked Deep Neural Networks
• A nonlinear solar magnetic field calibration method for the filter-based magnetograph by the residual network
• Convolutional Neural Networks and Stokes Response Functions
• Mapping the Sun's upper photosphere with artificial neural networks
• The MODEST catalog of depth-dependent spatially coupled inversions of sunspots observed by Hinode/SOT-SP
• Neural translation for Stokes inversion and synthesis
• Solar activity classification based on Mg II spectra: towards classification on compressed data