F. Benvenuto et al 2020 ApJL 904 L7

When applied to the prediction of the events associated with the 2017 September solar storm, a machine-learning algorithm was able to sound alerts every day a flare actually occurred and identify the corresponding flare class.

David Jewitt and Yoonyoung Kim 2020 Planet. Sci. J. 1 77

Optical observations of the asteroid P/2010 H2 (Vales) across 4 months in 2010 allow analysis of its outburst, which is consistent with the delayed activation of buried supervolatiles after past encounters with Jupiter.

Takashi Horiuchi et al 2020 ApJ 905 3

New analysis of Starlink satellites show that the darkened satellites (Darksat) have comparable magnitude or are brighter than the ordinary Starlink satellites in the Sloan Digital Sky Survey g’ band.

Meiji M. Nguyen et al 2021 AJ 161 22

Hubble Space Telescope images spanning a 14 yr period have provided the first detection of orbital motion of HD 106906 b, a wide-separation (737 au) planetary-mass (~11 M_Jup) companion to a spectroscopic binary.

Krzysztof Belczynski 2020 ApJL 905 L15

New calculations show that extremely massive black hole–black hole mergers (like GW190521) could, in fact, be formed through the classical isolated binary evolution, without the need to invoke exotic physics.

Hunter Goddard and Lior Shamir 2020 ApJS 251 28

Convolutional neural network classification is used to produce a catalog containing morphology labels for 1,662,190 galaxies with ~95% accuracy using the data of Pan-STARRS DR1.