MODEL GUIDE

NOAA HRRR-Smoke model explained

HRRR-Smoke couples frequent high-resolution weather forecasts with estimates of fire emissions to predict smoke transport across the contiguous United States.

A weather model with smoke

The High-Resolution Rapid Refresh is an atmospheric model designed for frequent updates over the contiguous United States. Its smoke capability represents emissions from biomass burning and predicts how weather transports and mixes that material. Smoke Window uses completed extended runs that reach 48 forecast hours.

The app reads the MASSDEN field for primary fine smoke near the first hybrid model level, roughly 8 meters above ground. It validates the native 1799 by 1059 Lambert conformal grid and converts kilograms per cubic meter to micrograms per cubic meter in the browser.

What the grid means

The model grid spacing is approximately 3 kilometers, so each sample describes a grid cell rather than an address. Smoke Window reprojects those cells onto an interactive Leaflet basemap without smoothing them into street-level detail. The visible blocks are a truthful reminder of the model’s scale.

Terrain is represented at model scale. Narrow valleys, ridges, urban canyons, and local drainage winds may behave differently from the grid-cell average. A city or county page therefore names its representative sample point and never claims uniform conditions across the geography.

Why inputs matter

Fire-emission estimates depend on satellite fire detections and fire radiative power. Clouds, thick smoke, canopy, and satellite timing can obscure fires. Plume-rise height, wind, mixing, rain, and deposition then shape the modeled concentration. Each new run can revise both the emissions and the meteorology.

NOAA and the National Weather Service advise using the model for trends and plume behavior rather than treating absolute concentrations as official air-quality forecasts. Smoke Window follows that distinction and pairs the model with separately labeled AirNow observations.

How Smoke Window retrieves one model record

The public HRRR archive stores GRIB2 files containing many meteorological fields. Smoke Window reads the small inventory, locates the first-hybrid-level MASSDEN record, and requests only that byte range. A Web Worker decodes the record and validates its packing template, dimensions, and Lambert grid before rendering. This selective process keeps a 17-frame forecast far smaller than downloading every field while preserving the native numerical smoke data.

From kilograms to a readable band

MASSDEN arrives in kilograms per cubic meter. The decoder converts it to micrograms per cubic meter, samples the grid cell containing the location, and assigns a Smoke Window-defined display band. The numeric conversion is physical; the Low, Light, Smoky, Heavy, and Very heavy labels are a product interface, not NOAA categories. Values below 4 form the site’s low planning threshold, which has no official health meaning.

Why completed extended cycles matter

HRRR runs are produced more often than the complete 48-hour smoke sequences used here. Smoke Window checks the extended 00, 06, 12, and 18 UTC cycles and accepts the newest one only when its hour-48 inventory is available. That choice keeps the page horizon consistent. It does not poll continuously; a full reload performs a fresh discovery check. A partially published cycle never replaces a usable complete run with a shorter forecast.

Deterministic does not mean certain

A deterministic run gives one outcome from one set of initial conditions and assumptions. It does not display a probability distribution or ensemble range. Users can still infer practical uncertainty by comparing plume edges, lead times, terrain, observations, and successive runs, but the map itself is not a confidence map. Exact concentration differences should be treated cautiously, especially at long lead times or near a sharp transport boundary.

Operational and life-safety boundaries

NOAA presents HRRR-Smoke as guidance, and National Weather Service training emphasizes trends rather than absolute air-quality interpretation. Smoke Window follows that boundary. The model can support timing decisions but cannot issue evacuation instructions, diagnose exposure, or replace an official AQI forecast. WFIGS records and AirNow observations remain separate sources with their own timestamps. Local authorities and incident agencies remain authoritative for urgent safety information.

Bottom line: NOAA HRRR smoke

HRRR-Smoke is powerful because it combines frequent high-resolution weather guidance with fire emissions on a consistent grid. Its limits follow from the same machinery: estimated fires, modeled plume rise, modeled atmosphere, and one deterministic outcome. Smoke Window exposes the run, valid time, native cells, and numeric bands so the data remains traceable. Responsible use emphasizes regional movement and timing while leaving AQI and life-safety decisions to official products.

A careful reading of NOAA HRRR smoke

HRRR-Smoke couples frequent high-resolution weather forecasts with estimates of fire emissions to predict smoke transport across the contiguous United States. Keep that immediate answer attached to the valid time, selected geography, model run, and product limits described above. Those details are part of the result, not fine print, and they should travel with any shared interpretation of this guide.

Primary sources