What remote sensing can, and cannot, tell you about a mineral system

Cover: mineral map of Cuprite, Nevada, made by the USGS from NASA AVIRIS imagery. The coloured zoning traces advanced argillic alteration across the district — NASA flies the spectrometer, the USGS turns the radiance into mineralogy, which is more or less the argument of this piece. Image: U.S. Geological Survey, public domain. Underlying imagery: NASA/JPL AVIRIS.

From 20 km up, the Earth looks honest. A high-altitude jet drags a spectrometer across a stretch of desert and, in less time than it takes to brew coffee, brings back a portrait of the ground built from 224 narrow bands of light — most of them well outside anything an eye can register. That portrait is real. It is also, in a very specific way, a beautiful lie.

This is the part nobody tells you about remote sensing.

01What a spectrometer actually sees

A hyperspectral imager is a prism with ambition. Where your eyes split sunlight into three rough bands — red, green, blue — an instrument like NASA JPL's AVIRIS splits the same beam into 224 narrow slices spanning the visible and shortwave infrared, from roughly 380 to 2510 nanometres.[3] Different minerals absorb different slices. Alunite has a sharp tell near 2.17 microns. Kaolinite leans on a doublet at about 2.16 and 2.20. Buddingtonite, the ammonium feldspar that turns up in some epithermal gold systems, leaves its signature around 2.12.[2] The spectrometer reads these dips like a barcode scanner, pixel by pixel, across a kilometre-wide claim block, and hands you back a mineralogy map — which species are present and roughly how much, not how many grams per tonne of anything.

That instrument flies on an aeroplane — AVIRIS is the Airborne Visible/Infrared Imaging Spectrometer, and the ER-2 that carries it cruises at about 20 km, not in orbit. But the same trick now rides above the atmosphere too: EnMAP, PRISMA, and EMIT bolted to the side of the Space Station. That is how a technique proven over one Nevada test site became something you can point at a continent.

It is genuinely magic. It is also a surface phenomenon, in the most uncompromising sense of the word.

02Three things the sky cannot tell you

The signal stops at the dust on your boots

Shortwave infrared reflectance comes from the outermost skin of the rock — micrometres to tens of micrometres, depending on grain size and opacity.[4] Basically nothing. If a pediment of caliche, a metre of alluvial fan, or a forest canopy sits between the sensor and your mineralisation, the sensor is doing a very precise study of the cover, not the ore. A large share of the world's prospective ground is obscured one way or another — vegetation, transported soils, glacial till, lake sediments, ice — which is exactly why the industry now frames its central problem as exploring under cover.[8] A perfect hyperspectral signature of grass tells you nothing about the porphyry sleeping 200 metres below it.

Alteration is not ore

This one hurts. The minerals a spectrometer is best at mapping — alunite, kaolinite, sericite, jarosite, iron oxides — belong to the alteration that surrounds an ore system, not to the ore itself. Find the alteration and you have found a promising address. You have not found a deposit.

Cuprite, Nevada is the textbook case, and it is more instructive than the textbook version. USGS researchers mapped it with AVIRIS down to the composition of individual alunites, and in doing so defined a steam-heated event nobody had recognised was there.[1] That is the part worth sitting with: the advanced argillic blanket at Cuprite formed above the water table, where sour steam condensed and ate the rock from the top down. It is not a halo wrapped around an orebody. It is a lid — a beautiful, perfectly zoned hydrothermal fingerprint that tells you a system was here and breathing, and almost nothing about whether it dropped metal anywhere you can reach. In a century of work at Cuprite no one has defined an economic deposit: a little silica, a little mercury, some copper-silver shipments early on. The plumbing was all there. The orebody has never been proven.

Every pixel is a committee

From the ER-2 at 20 km, AVIRIS Classic lays down roughly 20-metre pixels — about 400 square metres of ground apiece.[3] That single pixel is almost never one mineral. It is a weighted mix of five or six things, projected onto a single spectrum. Spectral unmixing algorithms can pull the components apart, but they introduce ambiguity, and ambiguity at planet scale becomes hallucination. Add the base rate problem — economic deposits are vanishingly rare against the number of anomalies a survey will hand you — and the instrument has just given you far more plausible places to look than any budget could ever drill. Worldwide nonferrous exploration runs about $12.4 billion a year, all in.[10] Spread across every anomaly a continental survey produces, that is not a drilling programme. It is a rounding error per target.

03What radar, magnetics and gravity add — and what they don't

The honest geophysicist owns several instruments because each one is truthful about a different thing.

Synthetic aperture radar ignores clouds and night. It sees roughness and structure, the lineaments and scarps that whisper about faulting. It says nothing about chemistry.

Magnetics and gravity reach kilometres deep. They see density and magnetic susceptibility contrasts: a buried intrusion, a sulphide body, a salt diapir, a fault offset. What they cannot do is read grade. Pyrite and chalcopyrite are both effectively non-magnetic, so a barren pyrite shell and an ore-grade chalcopyrite shell can look much the same in a magnetic survey — it is usually accessory pyrrhotite or magnetite that produces the anomaly, not the copper.[9] Induced polarisation will find disseminated sulphides, and it is no better at telling you which sulphide.

Stack all of these together and you get a richer picture. You also get more contradictions: the radar says fault, the magnetics says intrusion, the spectra say argillic, and the drilling so far says nothing for sale. Someone has to read the room. Historically that someone was a senior geologist with thirty years in the basin, an internal library of analogues and a coffee habit. There are not enough of them, and they do not scale.

04How a context-aware agent closes the loop

Here is what changes when an agent reads the picture, instead of a human pulling six layers into ArcMap by hand.

A context-aware agent does not look at a pixel in isolation. It holds the whole project at once: the regional tectonic setting, every legacy drill log on the property, the historical surface geochemistry, the structural interpretation from last quarter's magnetic inversion, the analogue deposit two basins over that produced a million ounces of gold, and the satellite cube that landed this morning. When it sees an alunite anomaly on a ridge, it does not just flag it. It asks, in writing, whether the regional architecture supports a high-sulphidation system here, whether the depth to basement is right, whether the closest historic hole tickled anything interesting, and whether the alteration zoning resembles the analogue or only the textbook.

It writes you a paragraph. The paragraph is wrong sometimes. But you can read it, argue with it and audit it. That is a vastly more useful artefact than a confidence raster.

The agent also remembers — and it is worth being precise about what that means, because the loose version of this sentence implies something we do not do. Your drill results do not train a model. They go into the record the agent reasons over: which spectra tracked grade on your ground, which targets fizzled, which ones nothing would have ranked unaided. So the next recommendation is argued against your newest holes instead of last year's. The satellite cannot do that. It passes over, and it forgets.

05A new division of labour

None of this means the satellite was wrong. The satellite was always doing the right thing, just at one specific layer of the problem. The shift is that we no longer ask the satellite to be the geologist. The satellite gathers. The agent interprets. The human decides where the rig moves on Monday.

That is the right division of labour for a planet that is running out of easy copper, and running out of patience to find the hard kind.