A Multi-Axis Anomaly Survey of the Jupiter Trojan Population
A seven-axis anomaly detection pipeline applied to 1,848 Jupiter Trojans, cross-correlating NEOWISE thermal properties, JPL SBDB orbital elements, and per-epoch photometric variability. The corrected pipeline identifies several objects warranting targeted follow-up — but no object survives as a robust multi-axis anomaly once photometric quality is strictly enforced.
Premise
The Jupiter Trojan clouds are dynamically stable for gigayears without active station-keeping, rich in volatile resources on surrounding bodies, and harbor a population of roughly ten thousand catalogued objects. These properties make them an attractive natural laboratory for anomaly searches: any object with unusual surface, thermal, or orbital properties would stand out against a well-characterised population baseline.
The multi-axis approach is motivated by a simple recognition. Single-parameter outliers are common in large asteroid populations — measurement error, unusual but natural compositions, observation-geometry effects. Objects that are simultaneously anomalous on multiple independent axes are far less likely to arise from measurement artifacts alone. That is the screen the pipeline applies.
Method
Seven independent anomaly axes were computed using robust statistics (median and median absolute deviation, to resist outlier contamination):
albedo (pV) · beaming parameter (η) · diameter/H consistency · orbit quality · eccentricity · inclination · photometric variability (W1/W2 excess scatter)
From an initial catalogue of fifteen thousand known Trojans, 1,848 objects had sufficient thermal and catalogue data for full scoring across all axes. A composite score combined per-axis z-score with outlier and extreme-outlier counts. Three scoring tiers were reported: raw (all axes), conservative (photometric quality enforced), and no-variability (scatter axis excluded).
An earlier pipeline version contained a catalogue-thermal join bug that produced NaN-to-NaN merge corruption (notably affecting 588 Achilles). The remediated pipeline strips and parses asteroid numbers correctly, avoids NaN merge keys, and yields the 1,848 matched objects reported here.
Multi-axis outlier counts
Three scoring tiers separate genuine multi-axis anomalies from photometry-driven artifacts:
Raw (all axes): 5 objects anomalous on 3+ axes.
Conservative (quality cuts): 3 objects.
No variability (scatter axis excluded): 0 objects.
The drop from five to zero when variability is excluded shows that photometric scatter is the sole driver of the 3+ axis signal in this sample. The multi-axis signal is real but fragile — it depends entirely on the scatter term, which is itself vulnerable to background-source confusion in the NEOWISE single-exposure pipeline. No Trojan in this sample is simultaneously anomalous on 3+ non-variability axes.
269620 (2010 UJ58) — primary object
Of the five raw 3+ axis candidates, 269620 remains the most interesting individual object. Its beaming parameter η = 0.628 is genuinely anomalous against the Trojan median of 0.88 (z = −2.2), indicating thermal emission more strongly concentrated toward the sub-solar point than is typical for Trojan regoliths. Its eccentricity (0.158) sits at the 96th percentile of the population. And critically, its NEOWISE field is the cleanest of any candidate: no VLASS radio detection in two epochs of imaging; no Gaia or 2MASS contaminant at either W1 or W2 outlier position; zero coincident sources across fourteen archival survey-date checks spanning gamma-ray to meter-wave radio.
What 269620 is not: a conservative 3-axis anomaly. Only three valid W2 epochs survive quality cuts, which is insufficient to validate the photometric variability axis cleanly. Under conservative scoring it retains two outlier axes (beaming + eccentricity), not three. The object is interesting; the claim is bounded.
30806 (1989 UP5) — a new target from the correction
The most notable new finding from the pipeline correction is 30806. It was not present in the pre-correction anomaly list and emerged only after the catalogue-thermal join fix resolved the NaN merge corruption. Its conservative scatter z-score of 9.03 is the highest in the sample. It has not yet undergone Gaia/2MASS contamination vetting and should not be overinterpreted until that step is complete.
Archival exhaustion
Twenty-four surveys from gamma-ray to meter-wave radio were checked for 269620, with zero coincident detections across the fourteen successfully cached survey-date checks. Two VLASS radio epochs yielded flux measurements consistent with noise (SNR ≈ 0.0 and +0.3; five-sigma upper limit ~660 μJy/beam at 2–4 GHz). Nearby-field source counts at the Gaia DR3 (2016) and 2MASS (1999) epochs are low. Further progress on 269620 is not an archival question: it is an observational one.
Recommended follow-up
The discriminating observation is near-infrared reflectance spectroscopy of 269620 across 0.8–2.5 μm. A D/P-type spectrum returns the object to the Trojan fold; metallic features or sharp spectral edges would be diagnostic of something more unusual. Roughly one hour on an eight-metre-class aperture at V~20.5 suffices — Gemini/GNIRS (Fast Turnaround) or Keck/NIRES. Optical spectroscopy at 0.4–1.0 μm provides a secondary taxonomic constraint; contamination vetting of 30806 is a zero-cost archival step that should precede any interpretation of it.
The working console
This report is the bound publication; Artifact-SETI maintains a live console with radar profiles for the top candidates, the seven-axis scoring distribution across 1,848 objects, the multi-wavelength archival sweep of 269620, and the recommended follow-up queue.