Granite and Dimension Stone Mapping in Pakistan: What Actually Works from Satellite

Last month I spent three days in a jeep driving from Skardu toward Shigar, looking at a granite block that a friend wanted me to co-invest in. Beautiful stone. Pinkish, coarse-grained, the kind of thing you'd see polished on a hotel lobby floor in Dubai. But the deposit was fractured to hell. Joints every 40-60 cm. You can't cut 2.7 m × 1.8 m blocks out of rock that shatters like that.

And this is the entire problem with dimension stone in Pakistan. We have granite everywhere — the Karakoram batholith alone covers thousands of square kilometers — but commercial-grade granite is a completely different animal from geological granite. Most mine owners I talk to don't know the difference until they've burned $200,000 on access roads to a deposit that won't yield.

So let's talk about what actually works when you're trying to find dimension stone from space.

What commercial-grade granite actually needs

Before we get to satellites, you need to understand what buyers want. Because "granite" on a geological map means one thing, and "granite" to a stone factory in Karachi means something very specific.

Commercial dimension stone needs three things at minimum. First, low fracture density — you want joint spacing above 1.5 meters, ideally 3 meters plus, so you can extract large blocks. Second, uniform color and texture over a decent volume (a factory doesn't want to sell customers a countertop that changes color halfway across). Third, minimal weathering to depth. If the top 8 meters are rotten and stained, you're stripping overburden forever.

Here's the thing — none of these are things you can directly "see" on a Sentinel-2 image. Sentinel-2 gives you 10 m pixels. A single joint is centimeters wide. So we're not looking at the fractures directly. We're looking at what fractures do to the landscape.

How we actually map it

At geomines we combine four data layers when a client asks us to survey granite prospects. This is the workflow we've refined over the last two years across projects in Gilgit-Baltistan, Chagai, and the Nagar Parkar area of Sindh.

DEM analysis first, always. We start with SRTM 30 m and, where we can afford it, ALOS PALSAR 12.5 m DEM. From that we derive slope, aspect, curvature, and — this is the important one — a lineament density map. Highly fractured granite bodies show a dense mesh of lineaments running in multiple directions. Massive, unfractured plutons show few lineaments and rounded, dome-like topography. That characteristic "whaleback" outcrop shape you see in parts of the Mansehra granite? That's what you want. It's a visual signature of low fracture density.

We calculate lineament density in 500 m grid cells. Anything under about 0.8 km of lineament per km² of area gets flagged as high-potential. Above 2.4 km/km², we usually walk away — too broken.

Sentinel-2 for color and weathering. This is where geo mining meets real petrology. Fresh granite has a distinct spectral response in the visible and near-infrared bands. Weathered granite — kaolinized, iron-stained — behaves differently in bands 4, 8, and 11. We build false-color composites that make weathered zones jump out in orange-brown while fresh outcrops sit cool and gray-white. If a client sends me a target and 70% of the exposure looks weathered on Sentinel-2, I tell them honestly, don't spend money on this one yet.

ASTER for mineralogy. ASTER's SWIR bands are old (the sensor's been degrading since 2008) but for granite work they still deliver. We use them to distinguish quartz-rich granites from feldspar-dominant ones, and to pick up muscovite and biotite signatures. Buyers pay premium for specific mineral compositions — a high-quartz leucogranite from Shigar sells for a very different price than a biotite-heavy granodiorite.

SAR for surface roughness. Sentinel-1 radar backscatter tells us about surface texture. Smooth, massive outcrops return one signature; rubbly, jointed surfaces return another. It's a rough proxy but it correlates surprisingly well with block size potential in the field.

Where this breaks down (and what I got wrong)

Honestly, when I started doing granite mapping satellite work in 2022, I thought DEM lineaments were basically the whole answer. I'd flag the smooth areas, tell clients this is your zone, and move on.

I got humbled fast.

A deposit near Astore looked perfect on every remote sensing layer we had. Smooth topography, low lineament count, clean spectral signature, uniform color. We recommended it. When they trenched, they found a horizontal joint system that was completely invisible from above — sheet jointing parallel to the surface, spaced about 60 cm apart. Useless for dimension stone. You can only detect that with ground penetrating radar or actual test pits.

So now our reports explicitly say: satellite gets you to the top 10-15% of candidate targets across a license area. It cuts your ground survey cost by maybe 60-70%. It does not replace a geologist with a Brunton compass and a proper joint survey. Anyone selling breeze geo mineral analysis or any satellite service as a complete answer for dimension stone is lying to you.

Where the real opportunity sits in Pakistan

Pakistan exports around $58 million of dimension stone annually. India exports over $1.5 billion. The gap isn't geology — it's identification and extraction quality. We have granite deposits Pakistan-wide that would compete globally if they were properly surveyed and quarried with wire saws instead of blasted with dynamite (which, by the way, destroys 40-50% of otherwise-good stone).

The Nagar Parkar granite in Tharparkar is one of the most underrated deposits in the country. Beautiful pink-to-red coloration, decent block sizes, and it sits close to Karachi port. The Mansehra granite complex has zones of massive leucogranite that would sell in European markets. Parts of the Deosai plateau — where I have some of my own mines — contain granites that geologically match commercial quarries in Brazil and Norway.

What's missing is systematic mapping. Nobody has done a proper deposit-quality survey of these regions at the resolution that international buyers demand. That's the work we're trying to do, one license area at a time.

If you're sitting on a granite lease and wondering whether it's worth the capex to develop, the first question isn't "is there granite there" — obviously there is. The real question is what percentage of that granite is extractable as sound blocks above 2 m³. And that's the question satellite intelligence can start to answer before you spend a single rupee on a bulldozer.