SARATH THARAYILST
WRITEUPSCONCEPTSPROJECTSLABABOUT
SARATH THARAYIL
മ
IGARATIPO: AMAZON TRIBUTARIES
/ SYSTEM

Building thoughtful software, writing notes, and shipping experiments across data, AI, and the web.

No cookies, no tracking. Preferences are stored locally in your browser. Anonymous view counts are kept server-side.

Hey, there's a story for the river above/© 2026 Sarath Tharayil/IST --:--:--
← LAB

Forest Fire

Jun 6, 2026

The Bak-Tang-Wiesenfeld forest fire model. Trees grow at rate p. Lightning strikes at rate f. Fire spreads to adjacent trees and burns out. No calibration needed — the system self-organises to a state where fire sizes have no characteristic scale. Small fires and catastrophic fires obey the same power law.

p GROW0.005
f SPARK0.00005

Green = tree · Orange = fire · Dark = ash/empty. Power law: small and large fires occur at all scales.

/ NOTES
SELF-ORGANISED CRITICALITY

The system never needs to be tuned. Starting from any state, it always drifts toward criticality — the border between order and disorder where scale-free behaviour emerges. This was Bak's central claim: many natural systems self-tune to criticality.

PARAMETERS
p (grow)Probability an empty cell grows a tree each step. Higher = denser forest, larger fires.
f (spark)Probability a tree ignites spontaneously each step. Higher = smaller, more frequent fires.
WHAT TO TRY

Set p very high and f very low to see continent-spanning fires. Set f high for a perpetual small-fire regime. The fire size chart on the right plots frequency vs. size — look for the straight line on log scale that signals a power law.