Launching the VASCO citizen science project

Brief

VASCO cross-matched the USNO catalog against Pan-STARRS to produce 150,000 candidate pairs spanning a 70-year temporal baseline, then cut the working sample to 68,632 (45%) by removing diffraction spike artifacts and unconfirmed POSS sources. Citizen scientists completed 15,593 classifications (~10% of candidates), returning 798 objects labeled 'vanished' and 359 flagged for deeper inspection via a 10-band review tool. A prior sweep of ~15% of the sample using SDSS imaging found ~100 red point sources visible in only one POSS-I epoch; shape and timescale analysis ruled out solar system objects, variable stars, low-redshift supernovae, and AGN, and none appeared in iPTF, Gaia, or the Catalina Sky Survey. The classification pipeline runs three parallel modes: hypothesis-driven vanishing-star identification, AI-assisted image-matching ranked by a lowest-match-index heuristic, and open-ended multi-band exploratory inspection.

Metadata

Category

Search

Venue

preprint (arXiv astro-ph)

Type

Preprint

Year

2020

Authors

Beatriz Villarroel, Lars Mattsson, Hichem Guergouri, Enrique Solano, Stefan Geier, Onyeuwaoma Nnaemeka Dom, Martin J. Ward

arXiv

2009.10813

Access

Open access

Length

21.0 M

Programs

VASCO, Galileo Project, Breakthrough Listen, Zooniverse, Galaxy Zoo

Instruments

POSS-I, Pan-STARRS, SDSS, Gaia, iPTF, Catalina Sky Survey, WISE/NEOWISE

Data sources

USNO catalog, Pan-STARRS DR1, POSS survey plates, Gaia catalog, iPTF, Catalina Sky Survey

Tags

SETI, technosignature, transients, citizen science, photographic surveys, archival astronomy

Key points

• Initial cross-match of USNO and Pan-STARRS catalogs using a 30-arcsecond radius yielded 150,000 candidates; artifact filtering reduced the active sample to 68,632 objects (45% of original).p.5
• At time of writing, 15,593 image pairs (~10% of candidates) had been reviewed by citizen scientists, producing 798 'vanished' classifications and 359 additional 'Inspect'-flagged objects.p.11
• A prior 15% sweep using SDSS found ~100 red point sources visible only in POSS-I single epochs; none matched detections in iPTF, Gaia, or the Catalina Sky Survey.p.3
• Shape and timescale analysis of the ~100 red transients explicitly excluded solar system objects, variable stars, low-redshift supernovae, and AGN as explanations.p.3
• The AI component ranks candidates by a 'matching index' measuring central-field alignment between epochs; images with the lowest index are prioritized for follow-up.p.8
• Nigeria's CBSS group, funded by an IAU/Office of Astronomy for Development grant, fielded 30 participants from Nigeria and Cameroon and analyzed ~3,000 images.p.12
• IES Tartessos (Sevilla, Spain) enrolled 22 students aged 12–13, who reviewed 700 candidates during monthly physics classes using school tablets.p.12
• Follow-up for the most interesting candidates is planned with optical, infrared, and radio telescopes described as 'the most potent radio telescopes.'p.2

Verbatim

• “The Vanishing & Appearing Sources during a Century of Observations (VASCO) project investigates astronomical surveys spanning a time interval of 70 years, searching for unusual and exotic transients.”

  p.2

• “We found about ∼ 100 red point sources where nearly all were visible in only one epoch and in the red images of the POSS-I survey.”

  p.3

• “The shapes and time scales involved rule out solar system objects, variable stars, low-redshift supernova, and AGN.”

  p.3

• “The fact that these point sources have escaped all transient surveys so far suggests they were not a phenomena with repetition time scales less than a few years.”

  p.4

• “The number of cases labeled as "vanished" by the citizens is 798.”

  p.11

• “The members in the Sirius association report that the interest to connect with VASCO is due to its intriguing aspect with non-conventional astronomy and even searches for extraterrestrial intelligence.”

  p.12

Most interesting

• Figure 1 shows two sources that both disappear between two POSS-I E (red-band) images taken only six days apart in early spring 1950, the shortest plausible natural-transient window the project has documented.
• The citizen science web interface includes a competitive game mechanic: users try to beat the AI's image-matching accuracy score, displayed side-by-side in green (human) vs. purple (AI) progress bars.
• The ~100 single-epoch red transients have escaped detection by every automated transient survey, iPTF, Gaia, and Catalina Sky Survey, despite those programs collectively cataloguing tens of thousands of flaring stars, supernovae, and AGN.
• Participants in the Nigeria/Cameroon CBSS cohort worked entirely from home during the COVID-19 pandemic; IAU grant funds were used specifically to purchase WiFi access for the 30 volunteers.
• Algeria's Sirius Astronomy Association joined VASCO partly because of its SETI dimension, a rare case of a non-Anglophone amateur association explicitly motivated by extraterrestrial intelligence searches.
• The Zooniverse platform has logged over half a billion total classifications across all projects; VASCO deliberately built its own separate interface to allow longer, multi-step missions and a higher 'playability' factor than Zooniverse's standard workflow.