Pakistan Sits on Some of the Best Marble on Earth. So Why Do We Export It Raw?
A slab of Ziarat white marble sells for around $18 per square meter when it leaves Pakistan as a raw block. That same stone, cut and polished in a factory in Foshan, China, comes back into the international market at $60 or more. We mine it. Somebody else gets rich off it.
That gap has bothered me for years.
Pakistan holds roughly 297 billion tonnes of marble and granite reserves — that's not a rounded marketing number, it's the figure the mining sector keeps citing and honestly it might be conservative. We're the fourth or fifth largest marble producer on the planet depending on the year. And yet our share of the global dimensional stone trade sits somewhere near 1%. India, sitting right next door with comparable geology, does many times that.
So the reserves aren't the problem. The problem is everything after the reserves.
What the satellites actually see
When people hear "geo mine" and satellite imagery, they think gold, copper, lithium. The sexy stuff. Marble and granite get treated like an afterthought. Big mistake.
Dimensional stone is one of the easier targets to map from orbit, and it's also one of the most commercially reliable. Here's the thing — you're not looking for a needle. Marble and granite bodies are large, exposed, and they have distinct spectral signatures. ASTER data picks up carbonate absorption features in the shortwave infrared band that basically light up marble deposits. Granite gives itself away through its silica and feldspar response. We layer that on Sentinel-2 for surface detail and SRTM DEM data to understand the slope and structure of the body.
Why does slope matter for stone? Because a beautiful white marble deposit sitting on a 60-degree cliff face is a nightmare to quarry. You'll waste half your block yield just fighting gravity and fractures. The DEM tells us that before anyone drives a single truck up the mountain.
We ran a breeze geo mineral analysis pass over parts of Chitral last year and flagged granite bodies that nobody had documented as commercially interesting. The color was there. The block size potential was there. And critically — road access within a workable distance. That last part kills more marble projects than bad geology ever does.
I got this wrong early on, by the way. When I started mapping stone for GeoMine AI I obsessed over deposit size. Bigger is better, right? No. A smaller deposit with clean, un-fractured blocks and a road nearby beats a massive fractured mess every single time. Block recovery rate is the number that actually pays your bills.
The reserves nobody's fully mapped
Balochistan is the marble giant. Chagai, the areas around Nushki, Loralai — enormous onyx marble and pink marble reserves. Some of the onyx coming out of there is genuinely world-class, the kind of material Italian designers pay silly money for. But most of it moves as raw block through Karachi and disappears.
Khyber Pakhtunkhwa has the Buner and Mohmand marble, the famous Ziarat white. Then you've got granite reserves scattered through the north — Gilgit Baltistan (where my own 15 mines sit), Chitral, and pockets of the Nagar and Hunza valleys with granite that would compete with anything from Brazil or India on color and hardness.
And Sindh has the Nagarparkar granite. Pink and grey granite sitting in the Thar region that's barely scratched. Some of it's been quarried in a small way for decades, but the full extent of the dimensional stone reserves out there hasn't been properly delineated. Not with modern remote sensing anyway.
Here's what frustrates me. A lot of these deposits get "surveyed" the old way — a geologist visits, takes a few samples, writes a report, and it sits in a drawer. Nobody has a proper spatial map of block quality across the whole body. So when an investor asks "how many cubic meters of export-grade material is actually here?" nobody can answer honestly.
That's the question geo mining from space can start to answer. Not perfectly — you still need ground work to confirm block integrity — but you get a defensible first estimate over a huge area for a fraction of the cost of sending crews to walk every ridge.
Where the actual export money is
Look, raw block export is a trap. We keep doing it because it's easy and the cash comes fast. But marble export Pakistan strategy has to shift toward processed material — polished slabs, cut tiles, finished dimensional stone products. That's where the value multiplies three or four times over.
The SIFC push and the general appetite for mineral investment right now means there's finally money looking for good stone projects. But the money wants proof. It wants to see reserve maps, block yield estimates, access logistics, and a realistic production plan before it commits. "Trust me, there's great marble up that mountain" doesn't cut it anymore, and honestly it never should have.
What we do at geomines is give that proof a starting point. We map the deposit, estimate the extent, flag the structural risks from DEM and SAR, and hand over something an investor or a government mining department can actually read. Then the ground team goes in with a plan instead of a hunch.
One more thing I've learned working with actual quarry owners — the fracture pattern is everything for dimensional stone mining. Two deposits can have identical color and hardness, but if one is heavily jointed you'll pull out rubble and the other pulls out clean 3-meter blocks. SAR data helps us read lineaments and fracture density across the surface before anyone commits capital. It's not a guarantee. But it stops you from buying a lease on a shattered hill.
Pakistan's granite reserves and marble reserves are genuinely a national asset that's been treated carelessly for decades. We've got the stone. The Chinese and Italians have the processing and the maps. Two of those three things we can build ourselves — starting with knowing exactly what we're sitting on.
So the real question I keep asking mine owners who come to us: do you actually know how much export-grade block is in your lease, or are you guessing?