Identifying and Tracking Solar Magnetic Flux Elements with Deep Learning

Deep learning has drawn a lot of interest in recent years due to its effectiveness in processing big and complex observational data gathered from diverse instruments. Here we propose a new deep learning method, called SolarUnet, to identify and track solar magnetic flux elements or features in observed vector magnetograms based on the Southwest Automatic Magnetic Identification Suite (SWAMIS). Our method consists of a data pre-processing component that prepares training data from the SWAMIS tool, a deep learning model implemented as a U-shaped convolutional neural network for fast and accurate image segmentation, and a post-processing component that prepares tracking results. SolarUnet is applied to data from the 1.6 meter Goode Solar Telescope at the Big Bear Solar Observatory. When compared to the widely used SWAMIS tool, SolarUnet is faster while agreeing mostly with SWAMIS on feature size and flux distributions, and complementing SWAMIS in tracking long-lifetime features. Thus, the proposed physics-guided deep learning-based tool can be considered as an alternative method for solar magnetic tracking.

[1]  Chang Liu,et al.  Predicting Solar Flares Using a Long Short-term Memory Network , 2019, The Astrophysical Journal.

[2]  K. Singleton,et al.  An omnibus test for the two-sample problem using the empirical characteristic function , 1986 .

[3]  C. Zwaan,et al.  Elements and Patterns in the Solar Magnetic Field , 1987 .

[4]  F. Berrilli,et al.  Occurrence and persistence of magnetic elements in the quiet Sun , 2018, 1801.03871.

[5]  Philip R. Goode,et al.  HIGHEST RESOLUTION OBSERVATIONS OF THE QUIETEST SUN , 2010 .

[6]  Philip R. Goode,et al.  The 1.6 m off-axis New Solar Telescope (NST) in Big Bear , 2012, Other Conferences.

[7]  Florian Metze,et al.  Extracting deep bottleneck features using stacked auto-encoders , 2013, 2013 IEEE International Conference on Acoustics, Speech and Signal Processing.

[8]  Yoram Singer,et al.  Reducing Multiclass to Binary: A Unifying Approach for Margin Classifiers , 2000, J. Mach. Learn. Res..

[9]  Kesheng Wu,et al.  Fast connected-component labeling , 2009, Pattern Recognit..

[10]  Jan Swevers,et al.  Ground-based and airborne instrumentation for astronomy , 2010 .

[11]  Jo Bovy,et al.  Deep learning of multi-element abundances from high-resolution spectroscopic data , 2018, Monthly Notices of the Royal Astronomical Society.

[12]  Joel Nothman,et al.  SciPy 1.0-Fundamental Algorithms for Scientific Computing in Python , 2019, ArXiv.

[13]  L. R. V. D. Voort,et al.  ON FIBRILS AND FIELD LINES: THE NATURE OF Hα FIBRILS IN THE SOLAR CHROMOSPHERE , 2015, 1502.00295.

[14]  John F. Wu,et al.  Using convolutional neural networks to predict galaxy metallicity from three-colour images , 2018, Monthly Notices of the Royal Astronomical Society.

[15]  S. Shumko,et al.  Control and operation of the 1.6 m New Solar Telescope in Big Bear , 2014, Astronomical Telescopes and Instrumentation.

[16]  Geoffrey E. Hinton,et al.  Deep Learning , 2015, Nature.

[17]  E. Parker.  Exploring the innermost solar atmosphere , 2020, Nature Astronomy.

[18]  R. Erdélyi,et al.  Generation of solar spicules and subsequent atmospheric heating , 2019, Science.

[19]  C. Schrijver,et al.  The Properties of Small Magnetic Regions on the Solar Surface and the Implications for the Solar Dynamo(s) , 2002 .

[20]  Prasanth H. Nair,et al.  Astropy: A community Python package for astronomy , 2013, 1307.6212.

[21]  Mats Carlsson,et al.  Determination of the Acoustic Wave Flux in the Lower Solar Chromosphere , 2006 .

[22]  Clare E. Parnell,et al.  Nature of the magnetic carpet – I. Distribution of magnetic fluxes , 2002 .

[23]  Sung-Ho Bae,et al.  Solar farside magnetograms from deep learning analysis of STEREO/EUVI data , 2019, Nature Astronomy.

[24]  Trevor Darrell,et al.  Fully Convolutional Networks for Semantic Segmentation , 2017, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[25]  Clare E. Parnell,et al.  SOLAR MAGNETIC TRACKING. IV. THE DEATH OF MAGNETIC FEATURES , 2013, 1307.4019.

[26]  Clare E. Parnell,et al.  Solar Magnetic Tracking. II. The Apparent Unipolar Origin of Quiet-Sun Flux , 2008 .

[27]  B. Altieri,et al.  Detecting Solar system objects with convolutional neural networks , 2018, Monthly Notices of the Royal Astronomical Society.

[28]  Thomas Brox,et al.  U-Net: deep learning for cell counting, detection, and morphometry , 2018, Nature Methods.

[29]  Chang Liu,et al.  Evolution of Photospheric Flow and Magnetic Fields Associated with the 2015 June 22 M6.5 Flare , 2018, 1801.03486.

[30]  S. Freeland,et al.  Data Analysis with the SolarSoft System , 1998 .

[31]  M. Huertas-Company,et al.  Deep Learning Identifies High-z Galaxies in a Central Blue Nugget Phase in a Characteristic Mass Range , 2018, 1804.07307.

[32]  John D. Hunter,et al.  Matplotlib: A 2D Graphics Environment , 2007, Computing in Science & Engineering.

[33]  V. Hansteen,et al.  Small-scale Magnetic Flux Emergence in the Quiet Sun , 2018, The Astrophysical Journal.

[34]  B. T. Welsch,et al.  Solar Magnetic Tracking. I. Software Comparison and Recommended Practices , 2007, 0704.2921.

[35]  Clare E. Parnell,et al.  Solar Magnetic Tracking. III. Apparent Unipolar Flux Emergence in High-resolution Observations , 2010 .

[36]  Maya R. Gupta,et al.  Training highly multiclass classifiers , 2014, J. Mach. Learn. Res..

[37]  Sebastian J. Goerg,et al.  Nonparametric Testing of Distributions—the Epps–Singleton Two-Sample Test using the Empirical Characteristic Function , 2009 .

[38]  Philip R. Goode,et al.  Scientific instrumentation for the 1.6 m New Solar Telescope in Big Bear , 2010 .