V10 Tacker & ARI V3 Documentation

Astrophyzix Observatory · Technical Documentation

Astrophyzix CAD/Scout/Sentry/SBDB V10 Tracking System.

Full Methodology, Data Provenance & Defensibility Statement

This document provides a complete, transparent, and reproducible description of the Live PHA/NEO Tracker module (Astrophyzix CNEO Observatory V10.0). Every claim about data origin, processing, scoring, and limitations is stated explicitly so that independent verification is possible.

Primary Disclaimer: The Astrophyzix Risk Index (ARI) is a non-official educational heuristic developed solely by Astrophyzix. It is not the Palermo Scale, Torino Scale, or any NASA/ESA formal impact-risk metric. Official impact-risk assessments appear exclusively in the Sentry Watchlist section of the tracker, which displays NASA CNEOS Sentry-II data unchanged.

1. Purpose and Scope

The module is a browser-based, client-side visualisation tool that aggregates publicly available near-Earth object (NEO) and potentially hazardous asteroid (PHA) close-approach data from NASA/JPL and CNEOS APIs. It presents:

• Confirmed close approaches (JPL CAD)
• Unconfirmed NEOCP candidates (CNEOS Scout)
• Official impact-risk objects (CNEOS Sentry)
• On-demand orbital elements and physical parameters (JPL SBDB)
• A schematic radar, distribution charts, and the proprietary ARI visual score

It does not generate predictions, does not alter source data, and does not replace any official NASA product.

2. Official Data Sources

All data are fetched live (or from short-TTL caches) from the following public NASA endpoints. No proprietary or third-party commercial datasets are used.

Source	Endpoint	Content	Official Documentation
CAD Confirmed approaches	https://ssd-api.jpl.nasa.gov/cad.api	Close-approach data for catalogued NEOs/PHAs within a user-selected date window and distance limit. Includes distance, velocity, absolute magnitude, and (when available) diameter.	JPL CAD API docs
Scout Unconfirmed NEOCP	https://ssd-api.jpl.nasa.gov/scout.api	Current Near-Earth Object Confirmation Page (NEOCP) candidates. caDist is expressed in lunar distances (LD). Earth MOID is used when no close-approach distance is supplied.	CNEOS Scout API docs
Sentry Impact-risk list	https://ssd-api.jpl.nasa.gov/sentry.api	Official Sentry-II summary: cumulative and maximum Palermo scale, maximum Torino scale, impact probability, diameter, and potential impact years. Displayed verbatim.	CNEOS Sentry API docs
SBDB Orbit & physical	https://ssd-api.jpl.nasa.gov/sbdb.api	Osculating orbital elements, orbit class, Earth MOID, observation arc, and physical parameters (diameter, albedo, rotation period, spectral type). Fetched on demand per object.	JPL SBDB API docs

Relay architecture: JPL SSD endpoints do not permit unrestricted browser CORS access. The client therefore uses a deterministic fallback chain of public CORS proxies (direct attempt first, then allorigins, thingproxy, corsproxy). Each endpoint maintains its own preferred relay. A circuit breaker cools down a failing proxy for 5 minutes after two consecutive failures. All attempts are timed and logged. This is an engineering necessity, not a data-altering step; the JSON payloads remain identical to the official responses.

3. Processing Pipeline (How the Module Works)

1. Window definition. User selects a date range (default 7 days) and maximum approach distance (default 80 LD). These become date-min, date-max, and dist-max parameters for the CAD request.
2. Stale-while-revalidate cache. IndexedDB stores prior responses with namespace-specific TTLs (CAD 10 min, Scout 5 min, Sentry 30 min, SBDB 6 h). Any existing cache is painted immediately so the UI is never blank.
3. Live CAD fetch. The primary request retrieves all approaches in the window. A second request with pha=true obtains the official PHA list for the same window; designations are matched to set the official PHA flag. If the second request fails, a documented proxy is used (H ≤ 22.0 and minimum approach ≤ 0.05 AU).
4. Scout fetch. The current NEOCP summary is retrieved independently. Objects with a caDist value are treated as approaches; objects lacking caDist are labelled “Earth MOID” and excluded from approach statistics, radar, and charts.
5. Sentry fetch. The Sentry summary is loaded on an independent failure domain so that a Sentry outage never affects the CAD/Scout view. Objects are cross-referenced by designation for visual badges only.
6. On-demand SBDB. Expanding an object card or using the Orbital Viewer triggers a per-object SBDB request. Results are cached for the session and in IndexedDB.
7. Scoring & rendering. ARI is computed client-side from the parsed fields (see Section 5). Radar, charts, and cards are drawn from the in-memory object list. No server-side processing occurs.

4. Traceability and Verifiability

The module is designed so that every displayed value can be traced back to a public API response.

• Provenance display. The System Health panel shows last-successful relay, endpoint latency, success/failure counts, and a rolling diagnostics log of every request (timestamp, endpoint, relay used, HTTP result, milliseconds).
• Raw field tab. Every object card contains a “Raw Fields” tab that dumps the exact parsed object used for rendering and scoring.
• CSV export. The export includes the exact window, relay identifiers, API version strings returned by JPL, validation status, and every numeric field used in ARI. The file header states the sources and the distinction between close-approach and Earth-MOID rows.
• Reproducibility. Any user can issue the identical CAD URL (visible in browser developer tools or reconstructible from the window parameters) and compare the JSON directly with the values shown. Scout and Sentry endpoints are parameter-free summaries.
• No hidden transformations. Distances are converted only by the published constants 1 AU = 149,597,870.7 km and 1 LD = 384,400 km. Scout caDist is used as lunar distances exactly as documented. Velocity prefers v_rel (CAD) or vInf (Scout). Diameter uses the JPL-supplied value when present; otherwise the standard H-magnitude formula with pV = 0.14 is applied and labelled “H-derived”.
• PHA flag transparency. Each card states the exact basis used (“JPL CAD pha=true filter” or “derived proxy” or “Scout phaScore ≥ 50”).

5. Astrophyzix Risk Index (ARI) V3 — Exact Methodology

ARI is a heuristic visualisation aid only. It is deliberately capped at 98 so that it can never be confused with a 100-point official scale. The published Zenodo record for the earlier formulation is available at https://doi.org/10.5281/zenodo.22240223. The implementation below is the version used in Observatory V10.0.

5.1 Formula

ARI = min(98, round( D + V + S + P ))

where the four additive components are:

Component	Maximum	Exact expression	Input source
Distance (D)	40	40 · e(−0.138629 · missLD) (half-life of 5 LD; ARI_K = ln(2)/5)	CAD reported approach distance converted to LD, or Scout caDist (LD). If Scout supplies no caDist, this term is forced to 0 (object is treated as MOID, not an approach).
Velocity (V)	20	min(20, 10 · log10(max(1, speed_km/s)))	CAD: v_rel if present, else v_inf. Scout: vInf if supplied; missing velocity scores 0 and the card displays an en-dash.
Size (S)	20	min(20, diameter_m × 0.05)	JPL diameter (CAD or SBDB) when available; otherwise H-derived nominal diameter using D(km) = 1329 / √pV · 10(−0.2H) with pV = 0.14. The nominal value is used.
PHA flag (P)	18	+18 if PHA-class flag is true, else 0	Official JPL CAD pha=true list for the window (preferred); fallback derived proxy H ≤ 22.0 and min approach ≤ 0.05 AU; or Scout phaScore ≥ 50.

5.2 Design Rationale (for defensibility)

• The exponential distance term concentrates score on the closest objects while remaining continuous and free of arbitrary thresholds.
• Logarithmic velocity prevents a single extreme-speed object from dominating.
• Size contribution is linear and capped so that a large but distant object cannot outscore a small but very close object.
• The PHA term is a fixed bonus that recognises the official (or Scout-derived) designation without double-counting distance or size.
• The hard cap at 98 and the explicit “ARI / 98” labelling on the dial prevent any visual implication of a percentage or official scale.

5.3 Validation checks performed on every render

• Constant verification: 1 LD must equal 384,400 km.
• Apophis-scale benchmark: inputs approximating the well-known 2029 approach of (99942) Apophis must produce a high ARI (≥ 70).
• Far-object benchmark: a 60 LD, low-speed, small, non-PHA object must score ≤ 10.

Failures are recorded in the provenance line of the CSV export and in the diagnostics log. These checks do not re-validate the Scout unit conversion (which is fixed in the parser) or replace official Sentry metrics.

6. Explicit Limitations

• ARI is not an impact probability, energy estimate, or formal risk assessment.
• Scout objects have short observational arcs; their miss-distance and velocity values are subject to rapid revision. The module surfaces positional uncertainty and, on demand, Scout’s Monte-Carlo orbit cloud so that the spread is visible.
• CAD times are reported in TDB; the observatory clock is UTC. The ≈ 69 s difference is not applied.
• Radar bearings are rank-ordered for visual separation only; they are not true sky positions (CAD/Scout summaries do not supply RA/Dec at closest approach).
• Diameter estimates derived from H-magnitude assume a fixed albedo (0.14) and are labelled as such.
• Public CORS relays are third-party services and can be rate-limited; the circuit-breaker and multi-proxy design mitigate but do not eliminate transient failures. Cached data remain visible when live fetches fail.
• The module is a visualisation layer. All authoritative numbers remain those published by NASA/JPL/CNEOS.

7. Caching, Persistence, and Privacy

Only IndexedDB is used (astrophyzix-neo-v10). No localStorage or sessionStorage is written. Cache keys are namespaced by endpoint and window. Users can clear all caches from the System Health panel. No user-identifying data are collected or transmitted.

8. Conclusion — Defensibility Summary

• Every displayed datum originates from a documented, publicly accessible NASA/JPL or CNEOS API.
• Every numeric transformation is a published constant or an explicitly stated formula.
• PHA status is either the official CAD filter result or a clearly labelled fallback.
• ARI is mathematically defined, capped, and repeatedly distinguished from Palermo/Torino scales.
• Raw fields, diagnostics logs, and CSV exports allow any third party to reproduce or audit the exact values shown.
• Failure modes (relay outages, missing Scout caDist, missing diameter) are handled transparently and labelled rather than silently imputed.

No claim is made beyond what the source data and the published ARI formula support.

Astrophyzix CNEO Observatory V10.0 Technical Documentation
Data sources: NASA Jet Propulsion Laboratory Solar System Dynamics Group and Center for Near-Earth Object Studies.
ARI is an independent heuristic. Not affiliated with, endorsed by, or sponsored by NASA or JPL.
© Astrophyzix 2026. This documentation may be freely cited for verification purposes.