Canadian FWI System (Van Wagner 1987)

The FWI System uses four moisture codes and two behaviour indices, each updated daily from noon standard observations (temperature, RH, wind, 24-hr precipitation).

FFMC tracks surface litter moisture (0–101). It responds within hours — a single dry afternoon drives it toward ignition thresholds above 85.

DMC tracks duff moisture (loosely compacted organic matter, ~5 cm deep). It has a multi-week memory and reflects sustained drying between rain events.

DC tracks deep organic/mineral layer drought (weeks to months). It is the primary driver of BUI and thus of the final FWI value in dry seasons.

ISI = FFMC + wind → rate of spread. BUI = DMC + DC → available fuel. FWI = ISI × BUI → overall fire intensity.

Historical Chain Method

BurnWindow initializes the FWI chain by fetching 60 days of daily weather history from Open-Meteo's archive endpoint and walking each code forward day-by-day from Van Wagner startup defaults (FFMC=85, DMC=6, DC=15).

Convergence: FFMC is accurate after ~5 days. DMC after ~21 days. DC — and therefore BUI — after ~42 days. A cold spring with regular rain keeps DC low; a dry April rapidly drives it into the 200–400 range, pushing FWI meaningfully higher than startup defaults would show.

Caching: The chain state (FFMC, DMC, DC) is stored in localStorage keyed by lat/lon. On subsequent days, only the incremental days since the last run are fetched — typically 1–2 days of data, one lightweight API call.

Sources: Van Wagner & Pickett (1985); Lawson & Armitage (2008). Archive data: Open-Meteo /archive (free, no key required).

Dead Fuel Moisture — NFDRS Equilibrium (Nelson 1984)

Dead fuel moisture is estimated from the current temperature and RH using the Nelson (1984) Equilibrium Moisture Content (EMC) equations, which are the foundation of the US National Fire Danger Rating System (NFDRS) fuel moisture tables.

EMC is computed in three RH ranges:
 RH < 10%: 0.03229 + 0.281073·RH − 0.000578·T·RH
 10–50%: 2.22749 + 0.160107·RH − 0.01478·T
 > 50%: 21.0606 + 0.005565·RH² − 0.00035·T·RH − 0.483199·RH
where T is temperature in °F.

Time-lag adjustments (Fosberg & Deeming 1971) are then applied to approximate each size class:
 1-hr = EMC (fine fuels equilibrate within one hour)
 10-hr = EMC × 1.18 + 1.5 (small roundwood, ~1 in diameter)
 100-hr = EMC × 1.45 + 3.0 (medium roundwood, ~3 in diameter)

The 1-hr class equilibrates within an hour — pure EMC is the correct model and no lag is needed.

The 10-hr and 100-hr classes use the same 60-day historical weather chain as the FWI system. Each day the chain steps forward:
fm_today = EMC + (fm_yesterday − EMC) × e^(−24/τ)
where τ is 10 or 100 hours. After the 60-day run, both classes have fully converged — starting values are irrelevant. The 100-hr is most meaningful during drought recovery (slow to dry down) or after heavy rain (slow to absorb moisture), when it diverges significantly from a simple EMC estimate.

KBDI — Keetch-Byram Drought Index (Keetch & Byram 1968)

KBDI (0 = saturated, 800 = extreme drought) represents the net effect of evapotranspiration and precipitation on soil/duff moisture deficit.

Each day, drying is estimated from maximum temperature and mean annual precipitation (MAP):
dQ = [(800−Q)·(0.968·e^(0.0486·Tmax)−8.30)] / [1+10.88·e^(−0.0441·MAP)] × 0.001

Precipitation recharge: any daily total above 0.2 in subtracts 100× the excess (the first 0.2 in is assumed intercepted by the canopy).

BurnWindow steps KBDI forward through the 7-day Open-Meteo forecast and persists the running value in localStorage, so the index accumulates realistic history across page loads rather than resetting each visit.

Interpretation: 0–200 = low drought stress; 200–400 = moderate; 400–600 = high; 600–800 = extreme. Values above 400 indicate deep duff and mineral soil moisture deficits relevant to sustained burning and mop-up difficulty.

Sources: Nelson, R.M. (1984). A method for describing equilibrium moisture content of forest fuels. Can. J. For. Res. 14:597–600.  |  Fosberg, M.A. & Deeming, J.E. (1971). Derivation of the 1- and 10-hour timelag fuel moisture calculations for fire-danger rating. USDA Forest Service Research Note RM-207.  |  Keetch, J.J. & Byram, G.M. (1968). A Drought Index for Forest Fire Control. USDA Forest Service Research Paper SE-38.