Principles Of Relative Dating

Relative dating is less of a technique and more of a way of thinking you have to train yourself into after enough field seasons

I still remember my first real solo stratigraphic column in the Navajo Sandstone. I was 24, had a freshly sharpened Brunton, and thought I understood cross-bedding. Then I hit a thrust fault that folded the entire section upside down. My entire relative age model collapsed because I hadn't properly verified my original orientation before committing to a sequence. That mistake cost me three weeks of backtracking and re-measuring. It taught me something no textbook does: relative dating is only as good as your structural check. Superposition is the simplest one and also the one people mess up most because it only applies to originally horizontal strata. If someone hands you a tilted or folded sequence without pointing out the structural context, your superposition calls are worthless. The law states that in an undisturbed sedimentary sequence, the lowest layer is oldest and each successively higher layer is younger. Simple statement, catastrophic when applied blindly. Cross-cutting relationships come next and they apply to anything that interrupts existing rock. A dike cutting through layered sediment is younger than those layers. A fault that offsets those same layers is younger still. I've seen junior geologists get tripped up here when multiple intrusions overlap. The key is mapping every contact methodically before assigning ages. You draw the cross-cuts on your map sheet in different colors, and the youngest feature is always the one cutting the most other things. That's your visual shortcut for field work.

Inclusions are where most students lose points on exams but also where field geologists make expensive mistakes. If a fragment of rock A is trapped inside rock B, then rock A has to be older than rock B. This gets complicated fast with clastic sedimentary rocks because every clast inside a conglomerate is older than the conglomerate itself. The conglomerate age only tells you when deposition happened, not when those individual clasts formed. I've spent entire days screening gravels from a single outcrop just to figure out which clast population matched the regional basement and which came from somewhere else entirely. Fossil succession ties into this but operates on a different timescale. Faunal assemblages change through time in a predictable way, and once you identify the index fossil, you can correlate widely separated sections. This is how we built the first geological time scale before radiometric dating existed. Walther's Law governs the vertical stacking of facies though. It says that sediments deposited in adjacent environments will stack vertically in the same order those environments appear laterally. A beach facies overlying a tidal flat overlying a deep marine shale tells you a transgression happened. Reverse that sequence and you have a regression. Get this wrong and your entire paleoenvironmental interpretation flips. The practical workflow in the field goes something like this. You locate a clean exposure, preferably one with minimal weathering on the face itself. You measure a measured section at a scale that matches your project needs. A detail study might be one centimeter per division. A regional mapping project might be one meter. You note grain size changes, color shifts, fossil content, and any visible contacts. You photograph everything with a scale bar. You return to camp and plot it before your memory fades.

One specific problem I ran into recently involved a sequence in the Appalachian Valley and Ridge province where repeated thrust duplication made individual beds reappear three times across a two-kilometer stretch. My initial correlation placed beds that were actually separated by four kilometers of section right next to each other. The workaround was to use trace element geochemistry on the shales as a fingerprinting tool. Each depositional event had a slightly different signature, so even though the lithology looked identical, the chemistry let me separate the duplicates. It added about six hours of lab work but saved me from drawing an impossible structural model on my map. Concordance and unconformities deserve a standalone mention because they break the relative dating chain. A disconformity is an erosional surface between parallel layers. An angular unconformity involves tilted or folded layers beneath horizontal ones. A nonconformity cuts through igneous or metamorphic rock. Each one represents missing time, and that missing time is usually larger than the adjacent preserved sections. The Great Unconformity in the Grand Canyon erases nearly a billion years of record. When you're doing relative dating across an unconformity, you can establish that something is missing but you cannot determine exactly how much without additional data. The main limitation of relative dating is that it gives you sequence, not absolute age. Two layers might be clearly one above the other with no ambiguity about which is older, but without radiometric dating you have no idea whether that relationship spans a million years or a hundred million. I've worked on projects where relative dating established the framework perfectly, and the actual numerical ages turned out to be 40 percent different from initial paleomagnetic estimates. The relative sequence was still correct. The timeline was not.

Another limitation surfaces in metamorphosed terrains. Once rock has gone through greenschist or amphibolite facies, primary sedimentary structures are often obliterated. Cross-beds, graded bedding, mud cracks, fossil assemblages, everything gets overprinted by foliation and recrystallization. In those settings relative dating becomes nearly impossible without finding unmetamorphosed equivalents nearby. I've seen entire published maps rely on questionable correlations in these zones because the alternative was admitting the area couldn't be dated relatively at all. The biggest pitfall I see beginners fall into is treating relative dating as a standalone solution. It works best when combined with radiometric methods, paleomagnetic data, and biostratigraphic correlation. Use it to build the skeleton, then layer in the numbers afterward. When someone tells you they dated a whole basin using only relative methods, ask them where the unconformities are and whether they verified stratigraphic thickness against any absolute control points. Most can't. If you're starting out, the exercise that actually teaches you this stuff is measuring your own stratigraphic section from bottom to top in a nearby sedimentary area. Document every change. Photograph every contact. Try to correlate your section with one from a different outcrop five kilometers away. When the correlation fails, figure out why. That failure teaches you more than any successful correlation ever will.