Cell size, metres.
Centre [x, z] of the simulated square, frame coordinates.
Share of fuel a dry fire consumes (wetter fuel burns less completely).
Conditions without a weather: wind [x, z] in m/s and fuel moisture.
Fuel moisture, fraction of dry weight.
Wind velocity [x, z], m/s.
What stops fire.
Rock + snow ground share (0…1) above which ground doesn't burn.
Extra firebreaks.
Roads: the world's path height modifiers (flat worlds).
Rise/run above which ground is bare rock.
Sea, lakes and rivers.
Fuel derived from biomes.
Grass load at full coverage and height, kg/m².
Load multiplier per threejs-grass preset (dry grasses burn hotter).
Fuel model per scatter species.
Biomass of species not in species, kg (counted as shrub).
Scheduled ignitions (also added by FireSim.ignite).
0…1: per-cell randomness of ignition thresholds (ragged fronts).
Lightning ignitions from the weather's thunderstorms.
Chance (0…1) that a strike on dry fuel starts a fire.
Let lightning start fires.
Cap on the fire's length-to-breadth ratio. Spread is elliptical: Anderson (1983) LB = 0.936·e^(0.2566·U) + 0.461·e^(−0.1548·U) − 0.397 (U = midflame wind, mph), with the ignition at the rear focus, so the head runs, the flanks are slow and the backing fire crawls.
Share of the weather's (10 m) wind that blows at flame height (≈ 0.3 under forest, 0.5 in the open).
Cells with less fuel than this (kg/m²) don't carry fire.
Fuel moisture (fraction of dry weight) at which fire stops spreading (Rothermel's Mx).
Succession after a fire: days for each class to return to full fuel. Grass first, then shrubs, then trees.
Grass.
Shrubs.
Trees.
Burn time per kg/m² of fuel, seconds (a cell burns residence × (0.5 + load)).
Seed for the ragged fire front and ignition rolls.
Edge length of the simulated square, metres.
Slope effect: spread × exp(slopeCoefficient × rise/run); uphill runs, downhill crawls.
Base rate of spread on dry, flat, calm ground, m/s, per fuel class.
Grass and litter.
Shrubs.
Forest (surface + crown).
Time of tick 0, seconds.
Simulation step, seconds. The sim always advances in whole ticks, so results never depend on frame rate.
Seeded, spatially correlated variation, so fronts finger and edges are ragged.
Fuel load ± share.
Fuel moisture ± share.
Patch size, metres.
Seconds between weather refreshes (wind, fuel moisture).
Wind effect on the head fire: head spread × exp(windCoefficient × wind speed in m/s).
Every wildfire setting. Plain JSON.