Federal Data Sources & Disclaimers

Forecasts are probabilistic estimates, not guarantees. Point forecasts may not capture hyperlocal terrain effects in mountainous areas.

Open-Meteo aggregates multiple NWP models. Individual model accuracy varies by region, elevation, and season. Mountain terrain forecasts carry higher uncertainty.

Observation quality varies by station type. ASOS/AWOS stations are NWS-maintained and generally reliable. RAWS and volunteer networks have less QC. Sensors can drift; gap-fill algorithms may interpolate missing data.

HRRR is among the most accurate short-range convective models available but still misses localized terrain effects at scales below 3 km. Mountain passes, valley drainages, and north-facing slopes may behave differently than grid-point forecasts suggest. Forecast skill degrades beyond 24 hours.

PeakScout storm impact scores for individual activities are probabilistic estimates built from point-measurement data (SNOTEL) and forecast grids (Open-Meteo). They do not assess slope-scale terrain stability, individual route conditions, or equipment sufficiency. SNOTEL measures at a single point — wind redistribution and elevation gradients create significant local variability. Storm page severe weather banners display official NWS alert data — PeakScout does not issue, upgrade, or cancel NWS warnings.

METARs reflect conditions at the reporting station (typically an airport at low elevation). Mountain terrain above the station may experience dramatically different visibility, icing, turbulence, and ceiling. TAF accuracy degrades in complex terrain. PeakScout displays this data for situational awareness — it cannot substitute for a formal aviation weather briefing.

SNOTEL stations measure snowpack at a single point. Extrapolation to surrounding terrain introduces uncertainty — a station at 10,000 ft does not represent conditions at 13,000 ft. Sensor drift occurs; annual calibration may produce retroactive corrections.

NOHRSC grids assimilate satellite, radar, and ground observations but spatial resolution (~1 km) smooths elevation-dependent gradients. Do not use for slope-scale snowpack assessment.

Avalanche forecasts are regional-scale probabilistic assessments by trained forecasters. They do not predict avalanche activity at a specific slope. Individual terrain evaluation is the responsibility of the user. The danger scale is a communication tool, not a safety guarantee.

CAIC forecasts represent the best available professional judgment for each zone. Avalanche danger is zone-wide — local terrain, aspect, and elevation create significant variation within the forecast zone. No forecast eliminates terrain risk.

Provisional data has not been reviewed for accuracy by USGS. Approved data (marked "A") has passed QC. Stream conditions can change rapidly — especially during snowmelt and afternoon thunderstorms — faster than the 15-minute reporting interval.

FIRMS detects fires ≥ 1,000 m² in clear conditions. Small fires, fires under dense smoke, or fires that started after the last satellite pass will not appear. Hotspot location accuracy is ±500 meters.

InciWeb data is incident-management-generated and may contain inconsistencies between incidents. GIS data is approximate.

Mountain areas often lack nearby AQI monitors. Wildfire smoke can cause AQI to spike far faster than hourly readings reflect. AQI is based on ambient measurements at ground level — smoke aloft is not captured until it descends.

Park alert data is operational, not legally authoritative. Official closure orders are posted at park entrances and on park-specific websites.

Recreation.gov availability reflects reservable sites only. Walk-up sites, overflow camping, and group areas are not reflected. Site types and amenities are self-reported by managing agencies and may be outdated.

USFS trail data quality varies significantly by National Forest. Some trails have incomplete route geometries or outdated attributes. Fire restriction boundaries are approximate.

Land status boundaries are legally defined but GIS representations carry spatial accuracy limitations of ±30 meters in some areas. Verify access rights on-the-ground or with local BLM field office.

OSM is volunteer-maintained and carries no quality guarantee. Trail routes may be missing segments, have incorrect surface types, or describe trails that no longer exist. GPS tracks from PeakScout users supplement but do not replace field verification.

Aurora forecasts carry significant uncertainty at sub-continental spatial scales. A Kp of 5+ creates favorable conditions for mid-latitude viewing but does not guarantee it.

USNO astronomical calculations are accurate to within seconds under standard atmospheric conditions. Atmospheric refraction at the horizon introduces ±2 minute variation. PeakScout also implements Jean Meeus algorithms locally as a fallback, which match USNO output within 1 minute.

Automated magnitudes have uncertainties of ±0.3–0.5 magnitude units. Depth estimates for shallow earthquakes carry large uncertainties. Felt shaking reports (Did You Feel It?) are user-submitted and may be unreliable.