ASTROPHYZIX NEO PHA DUAL-WATCH
ASTEROID TRACKING & SCOUT SYSTEM V8.0

Sources. Confirmed close approaches: NASA JPL SBDB Close-Approach Data (CAD) API, ssd-api.jpl.nasa.gov/cad.api. Unconfirmed NEOCP objects: NASA CNEOS Scout API, ssd-api.jpl.nasa.gov/scout.api. JPL SSD policy prohibits direct website embedding, so the two requests are relayed (serially, one at a time, per the SSD fair-use policy) through a fallback chain: direct, allorigins/raw, allorigins/get, thingproxy, corsproxy. The active relay, API versions, fetch time, and HTTP status are recorded in the provenance line and exported with the CSV.

Time scale. CAD close-approach epochs (cd, jd) are reported by JPL in TDB (Barycentric Dynamical Time). They are displayed as-is and labelled TDB. The observatory clock is UTC. TDB - UTC is presently about 69.2 s; this offset is not applied to approach epochs as it is far below CAD time resolution for monitoring use. SOFA-grade time-scale conversion is not performed because this is a display console, not an ephemeris-generation tool.

Distance. CAD nominal approach distance dist and the dist_min/dist_max 3-sigma band are in AU and converted with 1 AU = 149,597,870.7 km, 1 LD = 384,400 km.

Diameter. When CAD returns a measured diameter (km) it is used directly with diameter_sigma as the band; size source = JPL. Otherwise diameter is estimated from absolute magnitude H using D(km) = 1329 / sqrt(pV) * 10^(-0.2 H), with nominal albedo pV = 0.14 and a band from pV = 0.25 (bright/smaller) to pV = 0.05 (dark/larger); size source = H-derived. Scout objects always use the H-derived estimate.

PHA-class (derived). CAD does not return the official MPC PHA flag or the orbit MOID. A window-local PHA-class proxy is used: H <= 22.0 AND minimum approach in window <= 0.05 AU. This is a heuristic, not the official PHA designation, and can differ from the MPC list. Scout objects use the Scout phaScore (PHA-class true if phaScore >= 50).

Orbit class / period. The CAD and Scout summary endpoints do not return osculating elements, so orbit-class and orbital-period fields and the Apollo/Aten filters from V6 have been removed rather than fabricated. Source, confirmation status, close-approach time uncertainty (t_sigma_f), and the 3-sigma distance band are shown instead.

Validation. On each load a benchmark sanity check confirms 1 LD parses to 384,400 km and Apophis-scale inputs (about 0.0003 AU, 7.4 km/s) yield ARI in the expected high band; failures are surfaced in the provenance line.

Engineering. Numeric columns are held in Float64Array structure-of-arrays state vectors. Parsed datasets are cached in IndexedDB (scientific store, 10 min TTL) with an in-memory fallback. The sync pipeline is a deterministic finite state machine (IDLE -> CACHE -> CAD -> SCOUT -> MERGE -> RENDER / ERROR). Heavier compute elements (N-body, symplectic/RK integrators, WebGPU/WASM SIMD, FLRW, particle-mesh) are intentionally not used: this console displays JPL's authoritative orbit solutions and does not integrate trajectories, so re-deriving them in-browser would reduce accuracy.

ARI scores are recalculated on each load and are not stored or published as impact predictions. ARI does not incorporate orbital covariance, observation-arc uncertainty, or n-body perturbation effects. Astrophyzix CNEO Observatory is a public-facing scientific division of Astrophyzix.org.