How Sky Sync decides

A fan of rays is projected along the sun's azimuth at the event, sampling low and mid cloud cover at multiple distances (40, 100, 180, 300, 400 and 450 km) toward the horizon, weighted toward the far samples where sunlight enters under the cloud deck. A clearing gradient toward the sun adds a bonus; a sealed near-horizon hard-caps the overall score, because a sealed horizon makes a burn physically impossible no matter how good the sky overhead looks. Directional optical depth is also checked along the sun's corridor.

National terrain-horizon profiles (Copernicus DEM GLO-30 / USGS 3DEP) gate the light-path score at each bearing: if the sun's apparent altitude at event time sits below the local terrain horizon, that path is scored as blocked. Missing terrain data fails open rather than penalizing the score, so the system never guesses at a horizon it cannot see.

Backglow reuses the horizon-gap and shootability factors from the primary event score rather than recomputing them, since sun-side clearance and wind or precipitation do not change depending on which way the camera is pointed. It adds two checks: cloud cover sampled at 10, 30 and 60 km along the anti-solar bearing, and a terrain-relief lookup against the same 30m DEM terrain-horizon dataset used for the sun-side gate so a mountain range can catch direct alpenglow with no cloud at all. Either source can carry the score on its own; a flat coastal cloud burn and a clear-sky mountain glow are both scored on their own merits, with no bias toward one or the other. If the sun-side horizon is sealed, Backglow caps low regardless because light has to clear the true horizon before it can reach behind you.

Models boundary-layer moisture, 2m dew-point depression, planetary boundary-layer height (HPBL), and 925/850mb thermal inversion caps to badge shallow Radiation Mist and valley Cloud Inversions independently without inflating your sky-color burn score. Within 6 hours of an event (including the T-2h pre-dawn departure check), Sky Sync queries nearby NOAA METAR/ASOS stations to confirm active <3 km surface visibility and compares station elevation against your trailhead to verify when a valley station is socked in below your ridge overlook.

Monitors precipitation recency and cloud clearing trends across a multi-hour lookback window, surfacing an active clearing badge for near-term events and applying a score bonus to the light-path horizon gap evaluation.

Classifies atmospheric optical depth into distinct environmental aerosol regimes (dust, smoke, sulfate, marine, clear/mixed). Dust and smoke follow independent optical transmission curves, accounting for dust maintaining directional light transmission longer before reaching total opacity, with directional corridor sampling checking air clarity along the sun trajectory.

Cloud-screen factor scores base and top altitude directly (distinguishing high ice cirrus from low stratus decks). For night sky modes, a Krisciunas & Schaefer (1991) physical scattered-moonlight brightness model accounts for moon position, altitude and local Bortle sky darkness rather than treating moon phase as a generic percentage.

Sourced from annual VIIRS radiance composite data with Bortle-scale classification and azimuthal light-dome profiling covering the continental US, Alaska and Hawaii, paired with astronomical event azimuth and altitude bearings so you can plan foreground alignment before opening PhotoPills or TPE.

Converts geographic coordinates into centered-dipole geomagnetic latitude (|Λ|), computes the Kp threshold required for the auroral oval to clear the local northern horizon (Kp_horizon) or reach an overhead corona (Kp_overhead), and samples a 3-bearing northern azimuth fan (345°, 0°, 015°) at 15, 50, 100 and 150 km to verify low-angle northern horizon clearance.

Automatically resolves coastal viewpoints within 20 km of the shoreline to the nearest NOAA CO-OPS harmonic station, plotting continuous water-level curves in feet MLLW across the ±90-minute photographic window alongside High/Low tide timestamps, rising/falling stage and receding wet-sand reflection guidance.

Scores 10m sustained wind, gusts and sub-window wind curves against four independent physical stability thresholds: Water & Reflections (glassy mirror vs. chop), Wildflowers & Macro (petal sway and minimum shutter speed), Foliage & Leaves (canopy rustle), and Tripod & Long Exposure (multi-second ND vibration risk), while identifying the lowest-wind 15-minute Calm Lull.

Combines NEXRAD composite reflectivity (dBZ), echo-top heights, storm cell distance and bearing, and solar altitude (< 42° for primary rainbow geometry) to determine whether your tripod stays dry while the anti-solar rain shaft or anvil underside catches direct golden-hour illumination.

Night-sky events resolve to four shared rating tiers (Excellent, Good, Marginal, Poor) while Sunrise and Sunset cards also classify each event into a photographic Sky Archetype (such as High-Canopy Burner or Post-Storm Clearing Gamble) paired with a drive-time decision tier (Worth 90m+ Drive, Worth 45m Drive, Local <20m, or Skip Drive).

Best-moment evaluation runs at finer time resolution within the scored window, so the callout reflects short-lived peaks in cloud clearing or light quality that a single nightly average would wash out.

Combines a precomputed regional Sunrise/Sunset quality heatmap with an interactive Topographic Fog & Inversion Map layer that compares boundary-layer inversion ceilings against high-resolution DEM elevation contours, complete with a click-to-inspect point drawer.

Runs comparative scoring across up to 10 user-saved locations plus a radial scan. Free users see a notice whenever a nearby viewpoint beats their local forecast; Premium shows the destination, score gain and drive time.

Persists the complete 7-day forecast payload and hourly timelines on-device across up to 8 saved location and mode combinations with freshness tracking and 1-click pre-departure sync, allowing full day-to-day detail inspection with zero cellular connectivity.

All four modes share the same location model, forecast pipeline and rating vocabulary, with mode-specific factor weighting underneath, so the map, the outlook and the status colors behave consistently no matter what you are shooting.

Across a rolling 90-day benchmark spanning 100 stratified reference sites, our verification pipeline captures T-2h pre-event boundary-layer visibility and reconstructs post-event conditions from NOAA URMA surface analyses, live METAR/ASOS <3 km surface visibility stations and GOES-16/18 ABI Level 2 satellite retrievals, publishing live confusion matrices and lead-time curves at /transparency.

Evaluates score spread across a stratified sample of NOAA GEFS ensemble forecast members to compute High, Medium or Low confidence tiers plus a Floor–Expected–Ceiling outcome range bar, flagging volatile post-storm setups as High-Reward Gambles and suppressing the badge when ensemble member density is insufficient.

Sync Insights are computed from the physical factor sub-scores, Sky Archetype classification and ensemble Floor–Ceiling bounds. For sunrise and sunset, a language model summarizes those calculated factors into a short note (with a deterministic rule-based fallback); it never changes a score.

The breakdown panel shows each weighted factor individually alongside the multiplicative composite, with explicit data-quality tags (full, partial or degraded) whenever a remote sample is unavailable.

By blending high-resolution short-range models (NOAA HRRR 3km and HRRR-Smoke) with global ensemble models (GFS, ECMWF IFS/AIFS, NBM) across a 7-day horizon, Sky Sync brings weather, terrain and astronomy together while respecting the craft of landscape and night-sky photography.