Abstract
We report the discovery of an active asteroid, 2016 UU121, for the first time via artificial intelligence-enhanced classification, informed by our NASA Partner program Active Asteroids, a Citizen Science project hosted on the Zooniverse platform. The early version of our deep neural network, TailNet, identified potential activity associated with 2016 UU121 in 40 Dark Energy Camera (DECam) images from UT 2021 September 10 to 11. The discovery was vetted and confirmed by our Active Asteroids core science team. In total, 66 DECam images of this object showed clear activity in the form of a tail. 2016 UU121 has a Tisserand parameter with respect to Jupiter of 3.161, thus we classify the object as an active asteroid. Moreover, the activity occurred near perihelion, so 2016 UU121 is also a candidate Main-belt comet.
Export citation and abstract BibTeX RIS
Original content from this work may be used under the terms of the Creative Commons Attribution 4.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
1. Introduction
Active asteroids are minor planets that display comet-like tails and/or comae, despite being on orbits usually associated with asteroids, such as Main-belt asteroids. Some active asteroids, such as Main-belt comets (MBCs; Hsieh et al. 2015), are active because of volatile sublimation. These objects help us trace the location of volatiles in the solar system, and teach us about the primordial solar system. However, active asteroids and MBCs are rare, with roughly 60 and 20 identified to date, respectively. The paucity is due to, in large part, difficulty in locating these objects as they are rare (roughly 1 in 10,000 objects; Hsieh et al. 2015; Jewitt et al. 2015; Chandler et al. 2018; Chandler et al. 2024a, in press).
2. Methods
We set out to find more active minor planets with the help of the public by creating a Citizen Science project, Active Asteroids, 17 a NASA Partner program hosted on Zooniverse (Chandler 2022). We show images of known minor planets that we extract from the Dark Energy Camera public archive (Chandler et al. 2018) and ask project participants if they see evidence of comet-like activity in the form of a tail or coma. As of 2023 December 12, over 8800 volunteers have carried out 7.3 million classifications spanning some 482,000 unique images, resulting in over 20 discoveries thus far (Chandler et al. 2024a, in press).
To enhance the Active Asteroids user experience, we filter out images unlikely to show activity, such as those displaying an object too faint to reasonably be seen (Chandler 2022). Nonetheless, the number of images needing classification is still overwhelming, so we set out to further improve our screening with the help of artificial intelligence (AI). Our past work has shown that Convolutional Neural Networks (CNNs) (Krizhevsky et al. 2017) can not only surpass the performance of our classic methods of detection and characterization of transient objects (Sedaghat & Mahabal 2018), but also can reveal scientific insights into large astronomical data (Sedaghat et al. 2021, 2023). For the current work, we chose to use a simple CNN-based binary image classifier as an AI assistant, hereafter TailNet, with the intent of bootstrapping it on Active Asteroids labels, and iteratively improving the system step by step.
3. Results
Unexpectedly, 2016 UU121 (Figure 1) was flagged as highly unlikely to be inactive by a very preliminary version of our classifier, TailNet, which had only been trained on a limited number of project classifications. Further investigation by our team identified a total of 66 images of 2016 UU121 showing activity in the form of a tail oriented in the anti-motion direction (Figure 1). The images spanned two dates, UT 2021 September 9 and 10, when 2016 UU121 was at a heliocentric distance of rH = 1.938 au and outbound from perihelion (true anomaly angle f = 16°).
Figure 1. 2016 UU121 (at center) of the 63'' × 63'' (north up, east left) FOV with a tail oriented in the anti-motion (red-bordered black arrow) directions as projected on sky. Also indicated is the anti-solar direction (yellow arrow). All images were 120 s VR-band exposures acquired with the Dark Energy Camera (DECam) on the Blanco 4 m telescope at Cerro-Tololo Inter-American Observatory, Chile (Prop. ID 2019A-0337, PI Trilling). (a) UT 2021 September 10 (observer Edward Lin). (b) UT 2021 September 11 (observer David Gerdes).
Download figure:
Standard image High-resolution image
We classify 2016 UU121 (semimajor axis a = 2.971 au, eccentricity e = 0.354, inclination i = 4
344, perihelion distance q = 1.920 au, aphelion distance Q = 4.023 au, Tisserand parameter with respect to Jupiter TJ = 3.161; retrieved UT 2023 December 13 from JPL Horizons; Giorgini et al. 1996) as an active asteroid as it orbits entirely with the main asteroid belt. Furthermore, as the activity we identified occurred near perihelion, 2016 UU121 is also a Main-belt comet candidate. The simultaneous discovery of two active asteroids with AI and Citizen Science, 2016 UU121 and 2008 GB140 (Chandler et al. 2024), is unprecedented.
Acknowledgments
Many thanks to Arthur and Jeanie Chandler, and Elahe Eslami, for their ongoing support. We thank Elizabeth Baeten (Belgium) for moderating the Active Asteroids forums. A special thanks to our Superclassifiers: Angelina A. Reese (Sequim, USA), Antonio Pasqua (Catanzaro, Italy), Carl L. King (Ithaca, USA), Dan Crowson (Dardenne Prairie, USA), @EEZuidema (Driezum, Netherlands), Eric Fabrigat (Velaux, France), @graham_d (Hemel Hempstead, UK), Henryk Krawczyk (Czeladż Poland), Marvin W. Huddleston (Mesquite, USA), Robert Zach Moseley (Worcester, USA), Thorsten Eschweiler (Übach-Palenberg, Germany), and Washington Kryzanowski (Montevideo, Uruguay). Thanks to Cliff Johnson (Zooniverse), Chris Lintott (Oxford), and Marc Kuchner (NASA) for ongoing Citizen Science guidance.
This material is based upon work supported by the NSF GRFP under grant No. 2018258765 and grant No. 2020303693. C.O.C., H.H.H., and C.A.T. acknowledge support from NASA grant 80NSSC19K0869. W.J.O. and C.A.T. acknowledge support from NASA grant 80NSSC21K0114. This research received support through Schmidt Sciences. Chandler and Sedaghat acknowledge support from the DiRAC Institute in the Department of Astronomy at the University of Washington. The DiRAC Institute is supported through generous gifts from the Charles and Lisa Simonyi Fund for Arts and Sciences, and the Washington Research Foundation.
This project used data obtained with the Dark Energy Camera, which was constructed by the Dark Energy Survey collaboration. This research uses services or data provided by the Astro Data Archive at NSF's NOIRLab. Based on observations at Cerro Tololo Inter-American Observatory, NSF's NOIRLab (NOIRLab Prop. ID 2019A-0337, PI Trilling), and the CADC Solar System Object Information Search (Gwyn et al. 2012).
Facility: CTIO:4m (DECam) - .
Software: astrometry.net (Lang et al. 2010), PyTorch (Paszke et al. 2019).