You're staring at a cross-section. Two features cut across each other. Which means one's an intrusion. One's a fault. Your job: figure out which happened first.
Sound simple? It is — until it isn't.
What Is Cross-Cutting Relationships
This is the bread and butter of relative dating in structural geology. Plus, the principle is straightforward: whatever cuts across something else is younger. So the thing being cut? Older. James Hutton figured this out in the late 1700s. It hasn't changed Practical, not theoretical..
But here's where people trip up. They see "fault" and "intrusion" and assume there's a universal rule. Faults always cut intrusions. Or intrusions always exploit faults. Now, neither is true. The only rule is the contact itself.
The XY and A Notation
In mapping and cross-section exercises, features get labeled. Fault XY. Intrusion A. On top of that, the letters don't imply sequence — they're just identifiers. The sequence lives in the geometry Which is the point..
If Fault XY offsets Intrusion A's contacts, the fault moved after the intrusion solidified. Also, both happen. If Intrusion A bakes Fault XY's gouge zone or follows the fault plane, the intrusion exploited a pre-existing fracture. Both leave evidence.
Why It Matters
Get this wrong and your whole structural history collapses.
Exploration geologists care because mineralization often rides the intersection of faults and intrusions. Hydrothermal fluids love fault zones. Practically speaking, intrusions bring heat and metal. If the fault came first, it's a plumbing system waiting for the intrusion. If the intrusion came first, the fault might've remobilized ore — or shattered it Simple, but easy to overlook..
Engineers care too. A fault that cuts an intrusion might still be active. Plus, an intrusion that welds a fault shut changes the rock mass behavior. Day to day, tunneling through one versus the other? On top of that, different support requirements. Different water problems Small thing, real impact..
Real-World Stakes
I've seen drill programs waste hundreds of thousands targeting "fault-controlled mineralization" where the fault post-dated the ore. And the fault just chopped the deposit into pieces. The real control was an earlier structure the intrusion followed Which is the point..
Conversely, a project in Nevada — the intrusion looked like it cut the fault. Turned out the fault reactivated after intrusion, offsetting the contact by meters. So naturally, the "offset" was actually a later slip surface. Core logging caught it. Surface mapping didn't Most people skip this — try not to..
How to Tell Which Came First
Field evidence. Always field evidence. Cross-sections are interpretations. The rock tells the truth — if you know what to look for.
1. Contact Relationships
Walk the contact. All of it No workaround needed..
Intrusion cuts fault: You'll see the intrusion body truncating fault gouge, breccia, or slickensided surfaces. The intrusion's chilled margin rests directly on fault rock. No fault fabric inside the intrusion. The intrusion is the younger feature.
Fault cuts intrusion: The intrusion's contacts are offset. You'll find matching chilled margins on opposite sides of the fault. The fault gouge contains clasts of the intrusion. Slickensides cut across the intrusion's internal fabric — flow banding, crystal alignment, xenolith trains That's the whole idea..
Sounds obvious. But weathering hides contacts. Day to day, vegetation covers the critical 2 meters. On the flip side, that's why you trench. Or drill.
2. Thermal Overprinting
Heat leaves fingerprints Worth keeping that in mind. Took long enough..
If the fault moved after the intrusion cooled, the fault rock shows no contact metamorphism. Think about it: just mechanical grinding. Consider this: gouge. Cataclasite. Maybe some low-temperature clay alteration from later fluids Took long enough..
If the fault was active before or during intrusion, the fault zone gets baked. Hornfels in the gouge. Recrystallized pseudotachylyte. Even so, the intrusion's contact aureole overprints the fault fabric. You might even see the intrusion following the fault plane — a dike exploiting the path of least resistance.
3. Internal Fabric Alignment
Intrusions have flow fabrics. Magmatic foliation. In practice, stretched xenoliths. Aligned phenocrysts.
If a fault cuts the intrusion, that fabric gets offset. You can measure the offset on the fabric itself — not just the contacts. This gives you slip magnitude and confirms the sequence That's the whole idea..
If the intrusion exploited the fault, the magmatic fabric often parallels the fault plane. The intrusion is the fault fill. The fabric wraps around fault-bounded blocks. It's not offset — it's conformable Simple, but easy to overlook..
4. Xenoliths and Clasts
Fault gouge with intrusion clasts? Fault is younger That's the part that actually makes a difference..
Intrusion with fault-rock xenoliths? Intrusion is younger No workaround needed..
But watch for recycled clasts. A fault reactivates, chews up an older intrusion, then a new intrusion comes along and bakes everything. The clasts in the second intrusion might look like fault gouge. On top of that, check the matrix. Check the contact metamorphism. Check for multiple intrusion phases.
5. Geochronology — When You Can Afford It
Ar-Ar on fault gouge. That said, u-Pb on zircon from the intrusion. (U-Th)/He on apatite for cooling ages.
Expensive. Slow. But definitive — if the samples are clean.
Gouge dating is tricky. Intrusion zircons give crystallization age. But syn-kinematic illite or muscovite can give you the fault's last movement. So open system. Fine-grained. If the fault age is younger than the intrusion age, case closed.
Just remember: a cooling age isn't a crystallization age. A fault at 47 Ma cuts it. But the fault might've reactivated an older structure. Day to day, a fault at 43 Ma doesn't — the intrusion was already solid. But an intrusion at 50 Ma that cooled through 300°C at 45 Ma? The age dates the mineral, not necessarily the structure.
Common Mistakes
Assuming All Faults Are Through-Going
They're not. That's why faults die out. They tip into folds. They split into fracture networks Simple, but easy to overlook..
You map a fault offsetting an intrusion contact. Regionally? Ten meters along strike, the fault vanishes into a monocline. Consider this: the intrusion contact is continuous there. Locally, yes. Did the fault cut the intrusion? The intrusion might post-date the main fault phase but pre-date a late reactivation Practical, not theoretical..
Map the whole structure. Not just the pretty outcrop.
Confusing "Fault" with "Fault Zone"
A fault is a surface. A fault zone is a volume That alone is useful..
An intrusion can cut the fault surface but exploit the fault zone. So the dike follows the damage zone, not the principal slip surface. In cross-section it looks like the intrusion cuts the fault. In 3D, the intrusion is in the fault That's the part that actually makes a difference..
This matters for fluid flow. The damage zone stays permeable. The principal slip surface seals. Mineralization follows the damage zone.
Ignoring Reactivation
Faults don't move once. They move, stop, move again. That said, different stress fields. Different slip senses.
Intrusion A at 100 Ma. So naturally, reactivates at 50 Ma (reverse slip). Fault XY moves at 90 Ma (normal slip). Reactivates at 10 Ma (strike-slip).
Which "Fault XY" are you talking about? Which means the structure? And the surface? The most recent movement?
In cross-section exercises, "Fault XY" usually means the structure — the composite feature. But the sequence question usually asks about the last significant movement that affects the intrusion Less friction, more output..
Clarify the question. Or state your assumption.
Trusting the Cross-Section Without Ground Truth
Cross-sections are models. They're only as good as the data.
I've seen published sections where the author assumed the fault cut the intrusion because "faults usually cut everything.Also, " The map showed no offset contacts. Think about it: no gouge with intrusion clasts. Just an interpretation.
Don't be that geologist That's the part that actually makes a difference..
The field relationships tell the real story. Measure the offsets. Collect structural data. Question whether that "fault" is actually a detachment bending bedding, or a fold hinge wearing a slip surface like a costume Worth keeping that in mind..
Misinterpreting Cooling vs. Crystallization Ages
Argon spectra can lie. In practice, not technically, but the interpretation often does. A muscovite grain gives you when it cooled below its closure temperature—often much later than when the fault actually moved. The fault might've been active for millions of years while the mineral slowly cooled and finally closed Which is the point..
Look for multiple mineral systems. In practice, biotite closes at higher temperatures than muscovite. If both give overlapping ages, you're probably capturing the main deformation phase. If they don't, the younger age likely represents final cooling, not fault activity.
Overlooking the Thermal Overprint
Intrusion heat resets everything nearby. A fault cutting a 50 Ma granite might contain minerals that grew during cooling at 45 Ma. The fault itself moved at 43 Ma, but your dating method sees 45 Ma.
Check for thermal aureoles. Look for recrystallized grains. If the fault zone shows no thermal alteration, you're probably safe. If it does, the thermal history matters more than the structural history.
The Practical Workflow
- Map first. Trace every fault trace. Note where it dies. Where it branches. Where it disappears into folds.
- Sample strategically. Collect across the fault zone—not just from one side. Look for mylonite, cataclasite, gouge.
- Date multiple minerals. Don't stop at one system. Cross-check cooling and crystallization ages.
- Question everything. That "through-going fault" that vanishes 50 meters away? It's probably a fault zone with complex geometry.
The geologic history isn't written in straight lines. It's written in overprints, reactivations, and subtle offsets that require careful reading. Trust the data, not the textbook examples.
In summary: Structural geology and geochronology are partners, not competitors. One tells you how the rocks moved; the other tells you when. Together, they tell you how the Earth actually works—one messy, beautiful deformation at a time.
