Geochemical Anomaly Survey: Threshold & Contour Mapping (2D)

Survey Design

Sample density 70 pts

Threshold sensitivity (k·σ) 2.0σ

Hidden Ore Body

Source strength 90 ppm

Dispersion halo (σ) 160 m

Population Statistics

Background mean

0.0 ppm

Std deviation (σ)

0.0 ppm

Anomaly threshold

0.0 ppm

Anomalies found

0 / 0

How it works

This is a stylised soil / stream-sediment geochemical survey, the standard first pass of mineral exploration. A hidden ore body leaks a trace element into the surrounding soil, forming a dispersion halo — a plume elongated along a groundwater-flow / drainage direction — on top of natural background variability, which real geochemical backgrounds follow approximately log-normally:

C(x,y) = bg · lognormal(μ=0, σ=0.35)
         + S · exp( −u²/2σ_long² − v²/2σ_lat² )
u,v = (x,y) rotated into the plume's long/short axes

Only the discrete sample points are "measured" — the true field above is never drawn. Two real exploration-geochemistry techniques then reconstruct it from those points alone:

Threshold — the classic mean + k·σ rule: any sample above mean(C) + k·σ(C) of the sampled population is flagged anomalous (red ring). Raising k tightens the screen and rejects more of the background noise as false positives.
Contouring — a Gaussian kernel-regression estimate (Nadaraya–Watson) rebuilds a continuous field from the samples: Ĉ(x,y) = Σ wᵢCᵢ / Σ wᵢ, wᵢ = exp(−dᵢ²/2b²), with bandwidth b set by sample spacing. Marching squares then traces isopleth contour lines through that estimated grid at the threshold and threshold+σ, threshold+2σ levels — the same lines a geologist draws by hand on a base map to target a drill hole.

Sample (below threshold) Anomaly (above threshold) Kernel-regression isopleth