Gravity Surveys in Mineral Exploration: The Density Story Most Reports Skip

By Sufyan · 2026-05-01 · 4 min read

I ran my first gravity survey in 2021 on a chromite block near Muslim Bagh. Spent four days hauling a CG-5 gravimeter across slopes that would make a goat reconsider its life choices. Got back a map that showed a +1.8 mGal anomaly exactly where the drillers later hit a 12-meter chromite lens.

That's when I stopped treating gravity as old-school.

Most people writing about mineral exploration in Pakistan obsess over spectral satellites. Sentinel-2 this, ASTER that. And yes, we use all of it at GeoMine AI — that's literally the platform. But there's a quieter tool that keeps proving itself, especially for dense ore bodies the spectral data can't see because they're sitting under 30 meters of cover. Gravity.

What a gravity survey actually measures

A gravimeter measures tiny variations in Earth's gravitational pull. We're talking microGals here — millionths of a Gal. The instrument is sensitive enough to detect the gravitational difference caused by you walking 10 meters closer to it. That kind of sensitive.

Why do we care? Because rocks have different densities. And when there's a body of dense ore sitting in less-dense host rock, the gravity field above it bumps up. Slightly. But measurably.

Here's the density cheat sheet I keep in my notebook:

A chromite body sitting in serpentinite gives you roughly 1.7 g/cm³ of density contrast. That's huge. That's why gravity exploration mineral work in Balochistan's ophiolite belts is so productive — chromite practically screams on a gravity map.

Gold? Gold itself is dense (19 g/cm³) but the deposits are usually too small and too disseminated to register. Don't run a gravity survey hoping to find a gold vein. Run it to find the structural traps and intrusive contacts where gold lives. That's a different conversation.

Where gravity beats satellite data — and where it doesn't

Look, I'm the satellite intelligence guy. I built GeoMine on the idea that you can pre-screen massive areas of Pakistan from orbit before anyone burns diesel getting to a remote valley. That part of the workflow isn't going anywhere.

But satellites have a hard ceiling. Spectral data reads the surface. SRTM DEM reads the shape of the surface. SAR reads roughness and moisture. None of them see through 40 meters of alluvium in the Chagai basin or 80 meters of glacial till in Skardu. Gravity does. Magnetics does. That's the gap.

So our actual workflow at geomines looks like this:

  1. Satellite stack first — Sentinel-2, ASTER, SAR, DEM. AI flags anomalies across thousands of square kilometers.
  2. Field teams ground-truth the top 5-10% of targets.
  3. On the survivors, we recommend gravity (and sometimes magnetics) before any drilling.
  4. Drill the intersection of multiple anomalies.

Gravity survey mining work isn't cheap per square km. But it's stupid-cheap compared to drilling a dry hole. One bad exploration borehole in Gilgit-Baltistan costs me anywhere from 18 to 35 lakh rupees depending on access. A gravity survey covering the same prospect? A fraction of that. The math isn't subtle.

Honestly, I used to think gravity was something for big mining companies with helicopters and quarter-million-dollar budgets. Then I worked with a small geophysics crew out of Quetta who did 200 stations on one of my chromite blocks for less than the cost of a single drill hole. Changed my mind completely.

The corrections nobody warns you about

Here's the thing about raw gravity readings — they're useless. You have to apply corrections, and if you skip any of them you'll chase ghosts for months.

Get the terrain correction wrong in a place like Hunza and your anomaly map becomes fiction. I've seen consultants hand over Bouguer maps that were basically just topography in disguise. If your gravity high perfectly matches the ridge line, somebody didn't do the math.

This is also where AI is starting to actually help, not just hype. We're integrating gravity datasets (where they exist — the Geological Survey of Pakistan has more legacy data than people realize) with our satellite layers in geomining workflows. The fusion is what creates real targeting power. A spectral anomaly alone is a maybe. A spectral anomaly sitting on top of a Bouguer high with the right structural setting? That's where I'd put my own money. And usually do.

The 15 mines I own in GB weren't picked by guesswork. Most started with satellite reconnaissance, but the ones I felt confident enough to invest serious capital in had geophysical confirmation backing the spectral story. Density doesn't lie. Spectra can be fooled by weathering, by vegetation, by a thin coating of iron oxide that mimics something more interesting underneath.

Gravity just tells you what's actually heavy down there.

And if you're sitting on a license block in Balochistan or Khyber Pakhtunkhwa wondering whether to spend money on a gravity survey before drilling — what's the density contrast you'd expect between your target ore and the host rock? If it's above 0.4 g/cm³ and your target is bigger than maybe 50 meters across, you've already got your answer.

© 2026 GeoMine AI. All rights reserved.