VASCO: Vanishing & Appearing Sources in Astrophysics

• VASCO is an astrophysical project that leverages archival and modern sky surveys to detect sources that vanish or appear over decades.
• It employs a multifaceted methodology combining catalog cross-matching, machine learning, and citizen science to minimize false positives and uncover rare transients.
• The project provides constraints on astrophysical variability, failed supernova rates, and technosignature candidates with advanced detection techniques.

The Vanishing & Appearing Sources during a Century of Observations (VASCO) project is an astrophysical initiative designed to systematically search for astrophysical sources that either disappear or appear across a baseline of approximately 70 years. By leveraging historical photographic sky surveys—primarily the Palomar Observatory Sky Survey (POSS) and USNO-B1.0 catalogs—and modern digital datasets such as Pan-STARRS, VASCO aims to identify rare transients, highly unusual variable stars, and potentially exotic or techno-signature phenomena. The project combines archival analysis, machine learning, Virtual Observatory services, and distributed citizen science efforts to address both the technical and data volume challenges posed by such a search for outlier events.

1. Scientific Motivation and Objectives

VASCO is motivated by the prospect of probing astrophysical variability and transients on decadal to centennial timescales, which lie beyond the reach of traditional time-domain surveys focused on events lasting from seconds to a few years. The main scientific drivers include:

• Search for unrecognized or rare astrophysical phenomena: This includes failed supernovae, extreme variable stars, red transients, direct black-hole formation, and events with little to no precedent in the literature.
• Set empirical constraints on the rate of ‘vanishing’ (or appearing) stars in the Milky Way: The project seeks to quantify how often genuinely luminous objects disappear or emerge over 70-year baselines.
• Technosignature searches: The project investigates the possibility that vanishing sources may serve as empirical signatures of advanced extraterrestrial technology, such as energy-harvesting megastructures or directed energy emissions, that evade standard transient detection pipelines.
• Expansion of time-domain astronomy to multi-decade epochs: While large-scale synoptic surveys (e.g., ZTF, CRTS, DASCH) generally cover timescales up to ~14 years, VASCO extends this to nearly a century with limiting magnitudes in the r-band reaching ~23.4 (Villarroel et al., 2019, Villarroel et al., 2020).

2. Data Sets, Surveys, and Temporal Baseline

VASCO integrates multiple historical photographic and modern digital sky surveys:

• POSS I/II (Palomar Observatory Sky Surveys):
  • Epochs: ~1950s (POSS I), ~1980s–1990s (POSS II)
  • Bands: O (blue), E (red) with E ≤ 20 mag, O ≤ 21 mag
  • Sky coverage: ~18,000 deg² (northern sky), plate scale ~67″ mm⁻¹
• USNO-B1.0:
  • Contains ~1 billion objects, each detected at least twice across five bands (O, E, J, F, N)
  • Includes data from POSS and other large-plate surveys
• Pan-STARRS1 (PS1):
  • Epochs: 2010–2014
  • Bands: g, r, i, z, y; r ≤ 23.3 mag (5σ depth), ~30,000 deg² sky coverage
• Virtual Observatory (VO):
  • Facilitates standardized cross-matching and artifact rejection
• Supplementary plate archives: Hamburg Observatory Großer Schmidtspiegel plates (1954–1957, < 0.91″/px) provide independent temporal and spatial coverage for cross-confirmation (Busko, 20 Mar 2026).

Typical time baseline between paired images is 50–70 years, with color coverage from the visual to near-infrared. The temporal extension enables detection of objects or phenomena varying/vanishing on decade-long timescales.

3. Methodology: Detection, Classification, and Validation

The VASCO methodology integrates catalog cross-matching, artifact rejection, multi-modal candidate evaluation, and human-computer collaboration:

• Catalog Cross-Matching:
  • The principal workflow contrasts USNO-B1.0 (historic) and Pan-STARRS1 (modern) positions and photometry. A matching radius of 30″ is used to account for up to μ < 0.43″ yr⁻¹ proper motions over 70 years (Villarroel et al., 2019).
  • Further filtering removes high proper-motion contaminants (Gaia DR2), known asteroids, and catalog blends.
• Artifact and Defect Filtering:
  • Diffraction spike filter: eliminates USNO sources within δ of a bright star (m ≤ 12.4, m ≤ –0.09δ + 15.3).
  • SExtractor applied to POSS I red plates with S/N ≥ 10; confirm USNO–POSS separation ≤ 3.5″.
  • After artifact filtering, 68,632 candidates remain from an initial pool of 150,000 (Villarroel et al., 2020).
• Three-pronged Candidate Evaluation:
  • Measures correlation via $$f(I_1, I_2) = 1 - \mathrm{Corr}[\mathrm{aligned}(I_1, I_2)]$$ over central 30×30 px.
  • Candidates with high f (low correlation) prioritized for review.
  • System calibrates against human alignments using user scores $u \in [0,1]$.
  • 3. Exploratory:
  • Direct inspection of up to ten bands across epochs for non-standard or peculiar features, with an "Other" flag for unanticipated anomalies.
• False/True Positive Control: Preliminary analyses suggest FPR ≲ 5% for candidates passing both AI ranking and at least two independent “vanished” votes (Villarroel et al., 2020).
• Spatial, Morphological, and Photometric Assessment: FWHM and ellipticity compared with field stars; plate defects are systematically excluded via profile morphology (see Section 5).

4. Citizen Science Platform and International Network

VASCO's validation and candidate discovery leverage distributed visual inspection through a custom citizen science interface:

• Platform Workflow:
  • Tutorial onboarding, then random epoch-pair display (old/new images, blinking, manual alignments).
  • Decision options: "Still there," "Defect," "Moved," "Vanished," "Other."
  • "Inspect" mode provides access to a ten-panel mosaic (five POSS bands, five Pan-STARRS), plus text annotation.
• International Network:
  • Amateur and educational communities (Algeria, Nigeria/Cameroon, Spain, Sweden, Puerto Rico, France) form the core volunteer base.
  • Teams are organized for distributed image checking, with instructional support and periodic group vetting.
• Throughput and Statistics: As of October 2022, 15,593 image pairs (~10% of filtered sample) manually classified, yielding 798 vanished objects (5.12% ± 0.18%) and 359 anomalous "inspect" flags (Villarroel et al., 2020).

Metric	Value	Notes
Initial cross-matched candidates	150,000	USNO × PS1
Surviving after artifact filtering	68,632	m ≤ –0.09δ+15.3; S/N ≥ 10; sep ≤ 3.5″
Image-pairs examined by volunteers	15,593	≃10% of sample; avg. 3 classifications per pair
"Vanished" votes (≥2× agreement)	798	5.12% ± 0.18%
"Inspect"-mode flags	359	2.30%
Volunteer communities	6	Algeria, Nigeria/Cameroon, Spain, etc.

5. Candidate Properties and Interpretation

Systematic follow-up and quantitative image analysis inform candidate vetting:

• Morphological Characterization:
  • The most robust transients, particularly single-epoch, red-only detections, are compared against the morphological point spread function (PSF) of local field stars.
  • VASCO transients exhibit systematically narrower FWHM (by ~30%) and lower ellipticity (by ~15%) relative to field stars (FWHM ≈ 2.3″ vs. 3.1″; ε ≈ 1.06 vs. 1.23), with Gaussian profile fits (Villarroel et al., 21 Jul 2025).
• Physical Interpretation:
  • Sub-second optical flashes, such as sunlight glints from rotating artificial objects in low Earth orbit, are consistent with the narrow, circular PSFs, lacking the smearing present in stellar images from long-exposure averages of atmospheric seeing, wind-shake, and tracking (Villarroel et al., 21 Jul 2025).
  • Emulsion flaws are ruled out for uniform, Gaussian, round profiles.
  • Some candidates’ lightcurves and color-magnitude properties suggest rare red flares (M dwarf), high-z SNe, or other unknown variables.
• Statistical Limits:
  • Across ~90 million stars surveyed, no unambiguous vanished main-sequence objects were found, setting an upper limit of P(vanishing star) < 1/90 million per 70 yr in the Milky Way (Villarroel et al., 2019).
  • Among 298,165 POSS I red-only sources, no objects were detected as vanishing in both POSS I and II, constraining the failed supernovae rate to <1 per 10⁹ stars per 70 yr (Solano et al., 2022).

6. Multiwavelength Follow-up and Future Directions

VASCO integrates optical, infrared, and radio follow-up strategies to confirm and characterize high-priority vanished candidate events:

• Optical/Infrared:
  • Devasthal Optical Telescope (DOT) targets r ∼24 mag; near-IR campaigns (UKIRT, VLT) aim for J, H, K down to ~22–23 mag to test for dust or late-type object counterparts.
• Radio:
  • FAST and MeerKAT observe at L-band and UHF for synchrotron or technosignature candidates (sensitivity to 50 μJy in 1 hr at 1.4 GHz, probing down to L ∼ 10²⁰ W Hz⁻¹ at D = 10 kpc) (Villarroel et al., 2020).
• Cross-archive Consistency:
  • Independent searches of Hamburg Schmidt plates yield confirmatory events with narrow FWHM profiles and similar rate estimates (~1 per 100 deg² per night), supporting the real transient interpretation (Busko, 20 Mar 2026).
• Scaling and Prospects:
  • Ongoing expansion to all ~70,000 candidates; machine learning to improve classifier efficacy; expected completion of citizen science classification and vetting by 2023.
  • Statistical non-detections will sharpen upper limits on vanishing phenomena, with implications for astrophysics and the search for advanced civilizations.

7. Controversies and Interpretive Challenges

• PSF Uniqueness and Physical Plausibility: There remains debate about whether the observed morphologies (narrow, round, Gaussian PSFs) could be artifacts of plate emulsion or spurious detection. Quantitative modeling of atmospheric and tracking effects supports the transient hypothesis over the emulsion flaw hypothesis for uniform, sub-seeing-width events (Villarroel et al., 21 Jul 2025).
• Contamination and False Positives: Ongoing efforts address photometric calibration errors, proper-motion confusion, and plate defects. Independent multi-instrument confirmation remains essential for establishing astrophysical origin.
• Sample Completeness and Systematics: The completeness of the cross-matching pipeline is influenced by proper motion corrections, matching radii, and artifact rejection. Future algorithmic refinements are expected to improve reliability (Villarroel et al., 2019).

VASCO thus provides a novel, statistically robust, and multi-pronged approach to surveying decadal baseline transients, offering constraints on both astrophysical rarity and the presence of exotic phenomena in the Galactic sky (Villarroel et al., 2020, Villarroel et al., 21 Jul 2025, Busko, 20 Mar 2026, Villarroel et al., 2019).