Multi-Path Validation Dataset for ERA5-Derived Atmospheric Refraction Coefficient (κ): Path 2 — Pic de Finestrelles–Pic Gaspard

Overview This deposit is the first entry in an expanding multi-path validation panel supporting Real-Time Atmospheric Refraction Coefficient Determination Using ERA5 Reanalysis Data for Line-of-Sight Visibility and Geodetic Applications (Alkalali, K.Z.A., under revision, Geodesy and Geodynamics, manuscript GEOG-D-26-00098R1), created in response to reviewer requests for validation across multiple paths, seasons, and climatic regions. It is separate from the Zenodo deposit for the original Sağrak–Shkhara path (DOI: 10.5281/zenodo.21413841) because this path uses a fundamentally different validation approach — ground truth from a documented, photographically-confirmed historical observation, rather than independent radiosonde bracketing. Path 2: Pic de Finestrelles → Pic Gaspard (443 km, 16 July 2016) On 16 July 2016, Marc Bret photographed Pic Gaspard (Massif des Écrins, French Alps) from Pic de Finestrelles (Pyrenees), a 443 km line of sight that held the Guinness World Record for longest photographed line of sight on Earth at the time. This is a positive ground-truth case: unlike the original Sağrak–Shkhara path (blocked under the resolved December 2024 conditions), visibility here is a known, documented fact, so a correctly functioning model must predict κ ≥ required κ for this specific path, date, and time. Observer: Pic de Finestrelles, Pyrenees (42.4145°N, 2.1335°E, 2820 m). Target: Pic Gaspard, Écrins Alps (44.9983°N, 6.3308°E, 3867 m). Great-circle distance: 443 km. Date/time: 2016-07-16, approximately 04:00 UTC. Result: κ = 0.1458, required κ = 0.1373 — clear, consistent with the documented observation. Contents One JSON file containing the full path definition, κ result, atmospheric layer data, modified-refractivity M-profile, sensitivity analysis, and provenance/verification notes; and a README documenting method and data source provenance. Method Identical to Path 1: refractivity N = 77.6·P/T + 3.73×10⁵·e/T² (Smith–Weintraub approximation to ITU-R P.453-14); vapour pressure e derived from ERA5 specific humidity; ordinary least-squares regression of N against geometric altitude over the observer–target slice; κ = −R_Earth·(dN/dz)·10⁻⁶, R_Earth = 6,371,000 m. Full method detail in the main manuscript, §3.2–3.5. Provenance and verification Coordinates and elevations were cross-checked against multiple independent sources (Wikipedia, PeakVisor, and an unaffiliated third-party geometric re-analysis) before adopting the primary source's own stated elevations (2820 m / 3867 m, from the photographer's original account of the observation) as canonical, since these are the figures that best reproduce the specific historical observation being validated against. The observation time (04:00 UTC) is an approximation reconstructed from the source's own description combined with astronomical sunrise calculation for this date and location; it is not an independently confirmed exact timestamp. This dataset reflects a corrected computation after identifying and fixing a display bug in the KAPPA frontend's modified-refractivity table; the bug affected only that table's display of N, not the reported κ, N_avg, or dN/dh values. Notes This deposit will be extended with additional paths via Zenodo versioning as the validation panel grows, per reviewer requests for a broader panel spanning baseline lengths, climatic zones, and seasons.