Zones, Geochemical & Geophysical Data

A saved Area can carry three kinds of context beyond its geometry: Zones — smaller AOIs inside the Area; your geochemical data — the sample dataset; and geophysical layers — your ground geophysics rendered on the Globe. Together they turn an Area from a polygon into a site model, and all three feed directly into analysis.

Zones #

A Zone is a smaller AOI inside a saved Area: the open pit, the tailings facility, a village to exclude, a patch of stable ground to use as reference. Use Zones to mask that part of the Area during analysis or to export only that part of a result.

There are three ways to get Zones:

• Auto-detected at upload — if the geometry you upload contains features inside another feature, the platform detects them and offers each one as a Zone (the Add as Zone checkbox in the geometry info).
• From the Area’s Edit window — click Add New Zone and draw it on the map.
• Right after a visualization — open the AOI DATA accordion under the run card and click Create zone: you draw the Zone directly over the rendered result, which is the natural moment to fence off something you’ve just spotted. Available when the Area is yours.

On the Globe, Zones display automatically whenever you center on an Area: the Area outline in blue, the Zones in orange with their labels. The AOI DATA accordion has toggles to show or hide the area and its labels.

Geochemical data #

Your sample dataset — the CSV, TXT or GeoJSON you attach when creating or editing the Area, with sample IDs, coordinates and mineral columns.

To see it in context, run a Visual on the Area and click the database icon (Toggle area points) on the completed Area step: your sample points display on the Globe, right over the imagery — ground truth and satellite view in one look.

Preparing your dataset #

The platform reads your file’s column headers automatically (case-insensitive), so name them from these lists:

• Sample ID — one column named id, sample, sample_id, sampleid, sample_no or samplenumber.
• Latitude — one column named lat, latitude, lat_deg, latitude_deg, y or northing.
• Longitude — one column named lon, long, lng, longitude, lon_deg, longitude_deg, x or easting.
• Measurements — at least one column containing an underscore is required; every underscored column is read as mineral data. Put the element and unit in the name: Au_ppm, Cu_pct, Fe2O3_pct.

Formatting rules that save headaches:

• Coordinates in decimal degrees (WGS84) with a decimal point, not a comma — -23.4517, never -23,4517.
• CSV and TXT are read the same way; the delimiter (comma, semicolon or tab) is detected automatically.
• For GeoJSON, the coordinates come from each feature’s geometry and the properties become your columns — the underscore convention for measurements still applies.
• Every sample must fall inside the Area’s geometry — the spatial validation blocks the upload otherwise and names the offending sample IDs, so you can fix exactly those rows.

Geophysical layers #

Uploaded per Area — at creation or from its Edit window — one GeoTIFF (.tif) per layer type:

• Elevation (DEM) — choose a preset source (SRTM, Copernicus, ALOS, NASADEM, MERIT, GMTED2010, or none) or upload your own DEM, which overrides the preset while active.
• Magnetometry (nT) — total magnetic intensity from your mag survey. Magnetic anomalies reveal what optical imagery can’t see: intrusions, altered structures and faults, even under cover.
• Gravity (mGal) — the Bouguer anomaly from your gravity survey, mapping density contrasts in the subsurface: dense ore bodies, basement highs, basin edges.

Three places to visualize them:

• From the Areas panel, choose Geophysical layers in the area’s actions menu — each layer has a Visualize button and renders with its own color scale and units.
• In the workflow, click the layers icon (Toggle geophysical layers) on the completed Area step.
• Active layers are listed as Globe layers chips in the AOI DATA accordion.

What this data is for #

• Training Machine Learning insights — your geochemical samples are the ground truth that Trained AI Models learn from, and your geophysical layers can join the model’s inputs alongside the spectral bands. The more site data the Area carries, the more your models know.
• Masking mountain shadows — for scenes whose source records the sun’s azimuth and zenith at capture, the Hillshadow mask toggle in AOI DATA combines that geometry with the Area’s elevation model to mask out terrain shadows — no more false anomalies on the dark side of a ridge. This uses the DEM only (any source works: a preset like SRTM or Copernicus, or your own upload — no custom DEM required), and applies to single scenes (mosaics have no single sun position).

Best practices #

• Define a reference Zone on ground you expect to stay stable — it gives every comparison a baseline.
• Name geophysics uploads with survey vintage (TMI_2024_airborne.tif) so the provenance survives the upload.
• In steep terrain, flip the Hillshadow mask on before interpreting mineral maps — shadows are the oldest false positive in the book.
• If a layer looks misplaced on the Globe, check the GeoTIFF’s projection before re-uploading — a missing or wrong CRS is the usual culprit.