Fundamental Concepts:

Continents and Oceans On Earth, at the broadest scale, we can divide the surface geology up into two domains: the continents and the oceans. This doesn't exactly correspond to the surface outlines, at least not now, as the sea level currently is quite high so the edges of the continents are partially flooded. We call these areas that are flooded but still geologically associated with - part of - the oceans the continental shelves.

The average geology - more or less, the chemical composition - of the oceans is distinct from that of the continents. Oceans are floored with basalt (iron-magnesium rich silicates, broadly) and thin layers of sediment over that, and this lies on other iron and magnesium rich rocks such as gabbro and peridotite.

The continents are richer in silica, sodium, potassium, and calcium, and correspond on average to the composition of a granodiorite, a more or less equal mix of two feldspars and quartz with minor amounts of other minerals. Notably the density of granodiorite is very significantly less (2.7 to 2.8 g/cc) than that of basalt (3.3 g/cc).

The oceans are (relatively) ephemeral in that through plate tectonics they can be recycled. Continents more or less do not. As a result, the average age of continental rocks is higher (and much more variable) than the systematic and relatively young age of continents. For example, the oldest sea floor is about 200Ma or so, and the oldest rocks are about 4Ga.

Continental areas represent the accumulation of material built by the 'plate tectonics factories' that can't be recycled back into the mantle. I'll go in to that in more detail later.

The important thing for our first step process understanding of resources is that material is added to continents at mountain belts and includes sediments and a lot of igneous rocks, and, once built, continents tend to persist. Another important concept is that very old sections of continents (cratons) are very stable and so despite supercontinents coming and going do tend to remain somewhat coherent.

A first resource: diamonds.

Diamonds are a high pressure form of graphite - in other words, carbon. The pressure required is found at depth in the mantle. It looks very likely that most diamonds are surface carbon that is carried back into the mantle at a subduction zone and then transformed to diamond, but that is somewhat controversial.

Diamond source rocks are the uppermost mantle in areas that are not unusually hot, and it turns out the best place for this is the uppermost mantle under the oldest part of continents - cratons. While part of the mantle from a chemistry point of view, zones below continents can thermally 'weld' to the continent above and so get carried along when the craton moves. We have very good reason to believe that most diamonds at surface come from areas more than 2Ga old that have been carried along beneath cratons (the South African and northern Canadian regions, leaving out the technical names) ever since, and that they come up fairly straight (more on that below). There are probably megatons of diamonds under continents, but very few make it to the surface. Finally, it appears that these areas lose their diamond endowment if mountain building is 'overprinted' on the area. This pretty much rules out diamonds in or even directly adjacent to mountain belts.

So this gives us a couple of design rules:

1. Diamonds are found in areas of old continental crust. No nearby mountains, nothing that massively disturbs the continental crust (massive volcanoes, rift valleys, ...).

2. Diamonds are rare.

Diamond Emplacement: Hosting

Diamond is metastable - it doesn't 'like' being at the Earth's surface very much, but energetically it doesn't convert to low pressure forms at surface conditions. This isn't true for igneous conditions, and paradoxically, they are brought from the upper mantle to the surface in relatively hot igneous rocks. The key is speed.

Diamonds are brought to the surface via kimberlite, high gas-content exotic magmas that act like a geological shotgun, carrying debris (mantle rocks and, uh, diamonds) up the the surface very very rapidly. During this process many diamonds are simply not going to survive.

Kimberlite comes up as sheets (called dykes) that, near surface, erupt into carrot-shaped pipes (called diatremes) that erupt suddenly creating a crater from perhaps a hundred to five hundred meters across. Much of the material is blown as debris into a rouch cone in the area and this will include both kimberlite, mantle fragments and of course the rocks through which the 'carrot' was blasted. Some will fall back into the hole - it isn't unusual to find surface material like wood well down in the kimberlite pipe due to airfall.

Kimberlite is very soft compared to most continental rocks and so will weather down, generally resulting in a broad surface depression or lake. There may be a surface mound around the depression at least initially but, like the pipe, is easily weathered down.

The problem is that the diamond content of a pipe is fairly low; it just isn't practical to mine them in a pre-technological culture.

This gives us three more design rules:

1. Diamonds are at least initially found in kimberlite, a very strange and distinct soft rock, generally in semi-curcular 'pipes.'

2. Pipes are generally weathered down and, in wet climates, may very well be infilled with lakes.

3. Except in very odd cases, or with some kind of surrogate technology in your world (magic?) pipes and dykes are too low grade to be worth pursuing.

Diamond Re-hosting

Since kimberlite is weak, the surface ejecta is weaker, and since diamonds are very very resistant at the surface, secondary deposits are very common.

The first type of deposit is a river placer. Like gold placers, diamond placers occur where river processes sort and separate materials by density. These can create - from a very large volume of source material - small pockets of diamond. Mining in the river banks or in the remants of the dried up river can result in highly attractive deposits that take very little technology to exploit.

The second type is a beach placer, and occurs where a river dumps sediment in the surf zone of an ocean or perhaps a beach at a large lake and they are again sorted and concentrated by waves.

Placers require a long history of erosion to take a large volume of kimberlite (very low percent diamond) to concentrate into highly attractive placers. However, given that kimberlites occur in old rocks and throughout earth history, this in't a huge problem!

So a couple of more design rules:

1. Given a source region and a through-going river you can get placer deposits of diamond concentrated to highly attractive degree.

2. Given that rivers and kimberlites occur through history, we might find diamond placers at the current earth surface or in ancient sedimentary rocks.