Sooner or later, almost every rock on Earth is broken down into sediment, and the sediment is then carried away somewhere else by gravity, water, wind or ice. We see this happening every day in the land around us, and the rock cycle labels that set of events and processes erosion.
We should be able to look at a particular sediment and tell something about the rocks it came from. If you think of a rock as a document, sediment is that document shredded. Even if a document is shredded down to individual letters, for instance, we could study the letters and tell pretty easily what language it was written in. If there were some whole words preserved, we could make a good guess about the document's subject, its vocabulary, even its age. And if a sentence or two escaped shredding, we might even match it to the book or paper it came from.
Provenance: Reasoning Upstream
This kind of research on sediments is called provenance studies. In geology, provenance (rhymes with "providence") means where the sediments came from and how they got where they are today. It means working backward, or upstream, from the grains of sediment we have (the shreds) to get an idea of the rock or rocks they used to be (the documents). It's a very geological way of thinking, and provenance studies have exploded in the last few decades.
Provenance is a topic confined to sedimentary rocks: sandstone and conglomerate. There are ways of characterizing the protoliths of metamorphic rocks and the sources of igneous rocks like granite or basalt, but they're vague in comparison.
The first thing to know, as you reason your way upstream, is that transporting sediment changes it. The process of transport breaks rocks into ever smaller particles from boulder to clay size, by physical abrasion. And at the same time most of the minerals in the sediment are chemically changed, leaving just a few resistant ones. Also, long transport in streams can sort out the minerals in sediment by their density, so that light minerals like quartz and feldspar can move ahead of heavy ones like magnetite and zircon.
Second, once sediment arrives at a resting placea sedimentary basinand turns into sedimentary rock again, new minerals may form in it by diagenetic processes.
Doing provenance studies, then, requires you to ignore some things and visualize other things that used to be present. It's not straightforward, but we're getting better with experience and new tools. This article focuses on petrological techniques, based on simple observations of minerals under the microscope. This is the kind of thing geology students learn in their first lab courses. The other main avenue of provenance studies uses chemical techniques, and many studies combine both.
Conglomerate Clast Provenance
The big stones (phenoclasts) in conglomerates are like fossils, but instead of being specimens of ancient living things they are specimens of ancient landscapes. Just as the boulders in a riverbed represent the hills upstream and uphill, conglomerate clasts generally testify about the nearby countryside, no more than a few tens of kilometers away.
It's no surprise that river gravels contain bits of the hills around them. But it can be interesting to find out that the rocks in a conglomerate are the only things left from hills that vanished millions of years ago. And this kind of fact can be especially meaningful in places where the landscape has been rearranged by faulting. When two widely separated outcrops of conglomerate have the same mix of clasts, that's strong evidence that they once were very close together (see an example from the San Andreas fault zone).
Simple Petrographic Provenance
A popular approach for analyzing well-preserved sandstones, pioneered around 1980, is to sort the different kinds of grains into three classes and plot them by their percentages on a triangular graph, a ternary diagram. One point of the triangle is for 100% quartz, the second is for 100% feldspar and the third is for 100% lithics: rock fragments that haven't fully broken down into isolated minerals. (Anything that isn't one of these three, typically a small fraction, is ignored.)
It turns out that rocks from certain tectonic settings make sedimentsand sandstonesthat plot in fairly consistent places on that QFL ternary diagram. For instance, rocks from the interior of continents are rich in quartz and have almost no lithics. Rocks from volcanic arcs have little quartz. And rocks derived from the recycled rocks of mountain ranges have little feldspar.
When necessary, grains of quartz that are actually lithicsbits of quartzite or chert rather than bits of single quartz crystalscan be moved over to the lithics category. That classification uses a QmFLt diagram (monocrystalline quartzfeldspartotal lithics). These work pretty well in telling what kind of plate-tectonic country yielded the sand in a given sandstone.
Heavy Mineral Provenance
Besides their three main ingredients (quartz, feldspar, and lithics) sandstones have a few minor ingredients, or accessory minerals, derived from their source rocks. Except for the mica mineral muscovite, they are relatively dense, so they're usually called heavy minerals. Their density makes them easy to separate from the rest of a sandstone. These can be informative.
For instance, a large area of igneous rocks is apt to yield grains of hard primary minerals like augite, ilmenite or chromite. Metamorphic terranes add things like garnet, rutile and staurolite. Other heavy minerals like magnetite, titanite and tourmaline could come from either.
Zircon is exceptional among the heavy minerals. It is so tough and inert that it can endure for billions of years, being recycled over and over like the coins in your pocket. The great persistence of these detrital zircons has led to a very active field of provenance research that starts with separating hundreds of microscopic zircon grains, then determining the age of each one using isotopic methods. The individual ages aren't as important as the blend of ages. Every large body of rock has its own blend of zircon ages, and the blend can be recognized in the sediments that erode from it.
Detrital-zircon provenance studies are powerful, and so popular nowadays that they're often abbreviated as "DZ." But they rely on expensive labs and equipment and preparation, so they're mainly used for high-payoff research. The older ways of sifting, sorting and counting mineral grains are still useful.