About ab astris

ab astris is a publication-grade discovery platform for variable stars and exoplanet candidates using TESS with multi-mission cross-validation (Kepler/K2 and Gaia DR3).

  • Variable Stars: Lomb-Scargle periodograms, VSX catalog cross-matching, publication-ready exports
  • Transit Detection: BLS transit search, 8-layer false positive filtering, ExoFOP validation
  • Cross-validation: Kepler/K2 + Gaia DR3 independent period confirmation
  • Export formats: VSX submission packages, ExoFOP, LaTeX, CSV, FITS

How It Works

Click any step to view technical details

1. Index
Populate database with TESS targets and cross-validation coverage
2. Explore
Find unanalyzed TESS targets ready for investigation
3. Analyze
Lomb-Scargle periodogram, bootstrap errors, variability scoring
4. Enrich
Full automated pipeline: cache, refine, validate, plots
5. Transit Search
BLS exoplanet candidate detection with 8-layer false positive filtering
6. Submit
Generate VSX submission package for TRUE discoveries

Discovery Scoring

0-100 point system reflecting publication readiness and scientific significance

Score Tiers
90-100Exceptional
75-89High-Quality
60-74Solid
50-59Moderate
40-49Marginal
<40Low
Variability0-40 pts

Measured amplitude: 8 pts per 1% variability, capped at 5%

Period Significance0-25 pts

FAP confidence: 25 pts for <0.0001, 20 for <0.001, 15 for <0.01

Data Quality0-15 pts

Sector count (2 pts each, max 10) + data points bonus (up to 5)

Bonus Multipliers
+20 pts
Ultra-Short PeriodP < 0.1 days (2.4 hours)
+5 pts
High Proper Motion> 50 mas/yr
+5 pts
Multi-Sector ConsistencyPeriod consistent across 3+ sectors

Discovery Categories

Periodic variable stars are classified into four tiers based on their presence in astronomical databases

NovelNot in SIMBAD

Truly unknown to science. No prior astronomical characterization exists. Highest priority for VSX submission.

New VariableIn SIMBAD, not in variable catalogs

Known star (HD, TYC, BD designation) but not previously identified as variable. Still valuable for VSX.

Cross-listedIn other variable catalogs

Already cataloged in ASAS, ZTF, or other surveys but not yet in VSX. May warrant VSX cross-submission.

PublishedAlready in VSX

Well-known variable star. Useful for validation but not a new discovery.

Classification Logic
  • SIMBAD coordinate check (5" radius + 1.5 mag tolerance) determines Novel vs New Variable
  • VizieR queries 7 variable star catalogs (VSX, ASAS-SN, ZTF, ATLAS, etc.)
  • Gaia/2MASS/WISE entries alone do NOT disqualify Novel status (survey detections ≠ characterization)

Feature Roadmap

Click any feature to expand for full details. Ordered by priority.

8features planned
17+weeks total
6/8complete
Filter

BLS Transit Detection Pipeline

Priority 1
Complete
High complexity
3-4 weeks
60% reuse
High impact
100%

VSX Export & VizieR Validation System

Priority 2
Complete
Low complexity
2-3 weeks
100% reuse
High impact
100%

Kepler/K2 Cross-Validation

Priority 3
Complete
Low complexity
3 weeks
95% reuse
High impact
100%

Gaia DR3 Cross-Validation

Priority 4
Complete
Medium complexity
3 weeks
90% reuse
High impact
100%

Transit Candidate Review Console

Priority 5
Complete
Medium complexity
1-2 weeks
85% reuse
High impact
100%

AI Transit Candidate Analysis

Priority 6
Complete
Medium complexity
1 week
70% reuse
High impact
100%

Transit Candidate Multi-Source Enrichment

Priority 7
In Progress
Medium complexity
2-3 weeks
70% reuse
High impact
70%

GNN Stellar Population Graph

Priority 8
Planned
Very High complexity
2-3 weeks
50% reuse
Medium impact
50%
Legend
Complexity:
● Low (2-3 weeks, 90%+ reuse)● Medium (3-4 weeks, 70-90% reuse)● High (2-3 weeks, complex implementation)● Very High (ML/training infrastructure)
User Impact:
● High (immediate tangible value)● Medium (platform expansion)● Low (automation benefits)
Priority Order: Ranked by value/complexity ratio, dependencies, and immediate user value. "Quick Wins" = Low/Medium complexity features with high impact.

Command Palette

Search for a command to run...