The glowing ground overlay is a live concentration field on a grid, not a texture — every cell holds a contamination value that is advected downwind, diffused into its neighbours, decayed by natural attenuation (sunlight, dilution, chemical breakdown) and, if a leak is active, replenished at the source each tick. This is the same shape of model — a Gaussian-plume dispersion approximation — that emergency planners run to estimate an evacuation radius after a real tank rupture, pipeline failure or plant explosion.
C(x,y,t+dt) = advect(C, wind·dt) · (1 − decay·dt) + emission(source)
next-cell = (1−k)·upstream_sample + k·neighbour_average // diffusion
- Leak rate — how much substance is released into the source cell each second; higher rates build a denser, faster-growing plume core.
- Wind speed / direction — advects the whole field bodily downwind; a stronger wind stretches the plume into a long thin tongue instead of a symmetric blob.
- Natural decay — the rate the substance breaks down or dilutes on its own; low-decay substances (persistent chemicals) leave contamination that lingers long after the leak stops.
- Deploy containment — simulates booms, absorbent barriers and shutoff valves around the source: a ring of strongly boosted decay traps and neutralises material before it can spread further, the way real hazmat response teams race to isolate a spill at its origin.
Real-world relevance: this mirrors why disaster response plans lean so heavily on speed and cross-sector coordination (fire/hazmat, water utilities, power grid backups) — every minute the source stays uncontained, the affected area and the eventual cleanup cost grow roughly with wind-driven plume area, not linearly with time.