How the predicted spread is computed

Every number on this page is produced by the same code the map runs, spread.js (about 100 lines, no dependencies). Nothing is hand-drawn.

This is not a fire forecast. It shows only which way the wind is pushing and a rough speed if nothing stopped the fire. Real fires throw embers a mile or more ahead and jump freeways, and this model cannot show that. Areas outside the shapes are not safe. Never use this page to decide whether to evacuate.

What this is and is not. This is a textbook surface fire spread estimate under the assumption that nothing stops the fire: no containment, no engines, no aircraft, no roads, no irrigated yards, no wind shift. Real fires on October 16 have all of those. It is published so people can see which direction the wind is pushing the fire and roughly how fast, not as a prediction of what will burn. Evacuation decisions come from OCFA, CAL FIRE and the Sheriff, never from this page.

1. Live inputs right now

Loading current wind and recomputing…

2. The pipeline

Wind comes from the National Weather Service observation at San Clemente Pier (station F1327), refreshed every 5 minutes, with Bell Canyon and Ortega as cross-checks. Direction is the direction the wind blows from; the fire is pushed the opposite way.
Dead fuel moisture is read from the NWCG fine dead fuel moisture table (PMS 437) using temperature and relative humidity, plus 1% for an exposed afternoon slope. At 93 °F and 20% humidity this is about 4%.
Fuel is Scott and Burgan model SH5, "high load dry climate shrub," the standard stand-in for mature coastal sage scrub and chaparral.
No-wind, no-slope spread rate R₀ is computed with the Rothermel (1972) equations as corrected by Albini (1976).
Wind and slope multipliers φ_w and φ_s scale R₀. The 20-ft wind is reduced to midflame height with a 0.4 wind adjustment factor (unsheltered shrub). The wind effect is capped at the effective wind limit of Andrews, Cruz and Rothermel (2013).
Shape is Anderson's (1983) ellipse: the length-to-breadth ratio grows with wind, capped at 8. Backing and flanking rates follow Alexander (1985).
Isochrones. From each assumed front point the ellipse is grown for 1, 2 and 3 hours, and the convex hull of all ellipses is drawn. The dashed band repeats the 1-hour case with the gust speed instead of the sustained speed.

3. The equations

Rothermel rate of spread

R = I_R · ξ · (1 + φ_w + φ_s) / (ρ_b · ε · Q_ig)          [ft/min]

I_R  = Γ' · w_n · h · η_M · η_s          reaction intensity, Btu/ft²/min
Γ'   = Γ'_max · (β/β_op)^A · exp(A·(1 − β/β_op))
Γ'_max = σ^1.5 / (495 + 0.0594·σ^1.5)
A    = 133 · σ^−0.7913
β_op = 3.348 · σ^−0.8189
η_M  = 1 − 2.59·r + 5.11·r² − 3.52·r³      r = M_f / M_x  (moisture damping)
η_s  = 0.174 · S_e^−0.19                   S_e = 0.010   (mineral damping)
ξ    = exp[(0.792 + 0.681·√σ)(β + 0.1)] / (192 + 0.2595·σ)   propagating flux ratio
ε    = exp(−138/σ)                          effective heating number
Q_ig = 250 + 1116·M_f                        heat of preignition, Btu/lb

Wind and slope

φ_w = C · U^B · (β/β_op)^−E        U = midflame wind, ft/min = 88 · (20-ft mph) · 0.4
C   = 7.47 · exp(−0.133 · σ^0.55)
B   = 0.02526 · σ^0.54
E   = 0.715 · exp(−3.59e−4 · σ)
U_limit = 96.8 · I_R^(1/3)          (Andrews, Cruz & Rothermel 2013); U is capped here
φ_s = 5.275 · β^−0.3 · tan²(slope)  slope = 25% assumed for these hills

Shape and isochrones

LB = 0.936·e^(0.2566·U_mph) + 0.461·e^(−0.1548·U_mph) − 0.397     (Anderson 1983, cap 8)
HB = (LB + √(LB² − 1)) / (LB − √(LB² − 1))                          head : back  (Alexander 1985)
back rate  = R / HB
flank rate = R · (1 + 1/HB) / (2·LB)
Ellipse after t hours: head distance D_h = R·t, back distance D_b = D_h/HB,
  semi-major a = (D_h + D_b)/2,  semi-minor b = a/LB,  ignition point at the rear focus.
Flame length L_f = 0.45 · I_B^0.46 ft,  I_B = I_R · (384/σ) · R / 60   (Byram, via Albini 1976)

Fuel model SH5 (Scott & Burgan 2005)

Class	Load, tons/acre	Surface-area-to-volume, 1/ft
1-hour dead	3.6	750
10-hour dead	2.1	109
100-hour dead	0	30
Live woody	2.9	1600
Fuel bed depth	6.0 ft
Dead moisture of extinction	15%
Live woody moisture assumed	65% (late season Southern California)

4. What the map draws

Front points. The red dots are the assumed active head, placed on the streets where homes were reported impacted in the 1:02 PM radio traffic (Via La Mirada, Via De Agua, Avenida Calita). They are set by hand and listed in facts.json. If the head moves, the dots move.
Solid orange bands are the 1, 2 and 3 hour envelopes at the current sustained wind.
Dashed band is the 1 hour envelope if the current gust speed were sustained.
Direction is wind direction plus 180°. With an east wind the bands point west-southwest, toward I-5 and the coast.

5. Sanity checks

With a 15 mph 20-ft wind, 25% slope, 93 °F and 19% humidity the model gives a head rate of about 112 chains per hour (1.4 mph) and a length-to-breadth ratio of about 4. A 30 mph gust more than doubles the head rate to about 3 mph. Documented Santa Ana chaparral runs fall in this range; for example the 2017 Thomas Fire and 2007 Witch Fire had sustained head runs of 1 to 3 mph with spotting ahead of that.

6. Known limits

No ember spotting. Santa Ana fires routinely spot a quarter mile to a mile ahead and jump freeways; the model cannot.
No suppression, no fuel breaks, no houses or irrigated landscaping, no roads.
One fuel model for the whole landscape, one slope value, one wind value.
Wind is a single coastal station; canyon winds in the fire area can differ by 10 mph or more.
Containment is treated as 0% regardless of what OCFA reports, because the model cannot represent a held line.

7. Sources

Rothermel, R.C. 1972. A mathematical model for predicting fire spread in wildland fuels. USDA Forest Service INT-115.
Albini, F.A. 1976. Estimating wildfire behavior and effects. INT-30.
Scott, J.H. and Burgan, R.E. 2005. Standard fire behavior fuel models. RMRS-GTR-153.
Anderson, H.E. 1983. Predicting wind-driven wild land fire size and shape. INT-305.
Alexander, M.E. 1985. Estimating the length-to-breadth ratio of elliptical forest fire patterns. 8th Conf. Fire and Forest Meteorology.
Andrews, P.L., Cruz, M.G., Rothermel, R.C. 2013. Examination of the wind speed limit function in the Rothermel surface fire spread model. Int. J. Wildland Fire 22.
Andrews, P.L. 2012. Modeling wind adjustment factor and midflame wind speed for Rothermel's surface fire spread model. RMRS-GTR-266.
NWCG PMS 437, Fire Behavior Field Reference Guide, fine dead fuel moisture tables.

Model code drafted with an AI research assistant and checked against the published coefficients above. Report errors via GitHub issues.