Explore the claim

All four arms on one touch-friendly map, made to open on a phone in the field. Toggle the free-data layers, and tap a suggested target to see which signals put it there. The detail pages explain each method; this is where you decide where to walk. Red points are outcrop targets inside your claim. Gray points are strong signals just outside the claim boundary, shown for context. Blue points are stream-sediment sample sites, the cheap field method that pairs with this map.

The suggested targets are unsupervised hypotheses: the free-data layers fused with a documented, heuristic weighting that has not been validated against ground truth (there are no assays yet). They are where-to-walk-first points to field-check by sampling or detecting, never gold indications. They are now clipped to your claim polygons and require two coinciding signals; honestly, the strongest multi-signal coincidences sit about 600 to 800 m north of the claim (the gray context points). Add your sample points and the ranking can be recomputed against real data.

The ranked targets, also as a list:

Layers and legend

Iron oxide (S2 B04/B02), magma — relative ferric staining. Lights up laterite too; read it against contacts.
Clay/hydroxyl (S2 B11/B12), magma — broad alteration anomaly. An anomaly, not a mineral ID.
Lineament density (DEM), viridis — structural complexity from multi-azimuth hillshade. The highest-signal layer, on by default.
SAR texture (S1), viridis — radar roughness, cloud-proof but canopy-limited. Context only.
EMIT Al-OH (wet-scene anomaly, dry-season check: soil/litter artifact), magma — a wet-season Al-OH anomaly in the NW corner that the dry-season check (2024-11-04, leaf-off) showed to be a bright-soil and dry-litter artifact, not white mica. Kept on the map for transparency; it is not a mineralogy signal.
Red points outcrop targets inside your claim (two coinciding signals required). Gray points strong signals just outside the claim, for context. Blue points stream-sediment sample sites.

How the targets were made

The synthesis recipe reprojects the four layers to a common grid, normalizes each with a robust 2 to 98 percentile stretch, and sums them with a documented heuristic weighting that emphasizes structure: lineament density 0.40, iron oxide 0.35, clay/hydroxyl alteration 0.25, SAR as context only. Scoring is anchored on the DEM lineament layer, which is valid everywhere and sees through canopy, so vegetation-masked optical ground cannot manufacture false anomalies. The top few percent of the priority surface is clustered, and the strongest clusters become outcrop targets; stream-sediment points are placed on drainage lines just downslope of priority ground.

The weighting is a heuristic, not a calibrated model. It has not been validated against assays because none exist yet. Recipe: research/mineral-prospecting/synthesis/fuse_targets.py.