Plate tectonics

So, I've been working on a plate tectonics simulator for the last few months. It's been a project idea in my head for a long time now, and I'm having a lot of fun with it. As such, I've been doing a lot of reading on the subject, and accidentally getting exposed to obscure forms of Discourse from plate tectonic deniers. And that led me to start to think about all the evidence for plate tectonics, and how beautifully well it all fits together. It's really an underappreciated paradigm shift story that hasn't gotten the history/philosophy of science attention it deserves. So this is my list, generally in chronological order, of everything plate tectonics explains really, really well. If you don't want to believe it, fine, but this is what any alternative you want to propose is up against.

Shape of the continents

They kind of match, that's odd.

Geological continuation across continents

There are a lot of rock formations, and sequences of formations, that seem to match across continents if you lined them up the way they look like they would fit.

Fossil continuation across continents

Lots of specific fossils are found in areas that also span across continents, as if there used to be a continuous habitat for those creatures.

Shape of the continental shelves

If you look at the shape of the continental shelves instead of the current coastlines, the match is even better.

Earth's bimodal elevation

If you plot the elevation of the Earth's surface, it's strikingly bimodal. Most of the continental land is within a certain band of heights, and almost all the ocean floor is within a much tighter band of depths.

Mid-Atlantic ridge

Not only do the continental edges match, there is a ridge running right down the middle of the ocean which separates them. The shape of this ridge also matches the contours on either side. Something has to be up here!

Seafloor arcs coming off ridges

There are very gently curving features which connect all these ridges on the seafloor. And if you map them, you can show that they all are concentric around a certain point. This is what you would expect to happen if that section of crust was moving, as all motion on the surface of a sphere is equivalent to rotation about a certain axis.

Oceanic crust sediment thickness

The thickness of sediment on the ocean floor gets thicker and thicker the farther from the ridge it is, as if it's older and has had more time to accumulate.

Oceanic crust ages as given by radiometric dating

You can work out how old rocks are by measuring the presence of certain radioactive elements (such as uranium) and the things they eventually decay into (such as lead). By knowing the half-life of those elements, (and making some reasonable assumptions about the original content of the rock) you can work out how old it is. And by this measure, again, all ocean crust gets older the farther it is from a ridge.

Paleomagnetism (wet)

When certain minerals solidify, they lock in the orientation of the Earth's magnetic field in a way that can be measured hundreds of millions of years later. Since the magnetic field isn't always the same -- in fact is flips polarity fairly regularly -- this can let you work out age and/or original position of rocks. If you do sweeps across the mid-ocean ridges, maybe using some very sensitive submarine detection gear the Navy is happy to give you after WWII, you can see that the magnetic signatures are almost perfectly symmetrical on either side of the ridge.

Paleomagnetism (dry)

You can do the same trick on land, and see that lots of rocks are very far from where they were originally formed.

Earth's bimodal crust age

Some parts of continents are billions of years old. In contrast, there is no oceanic crust older than a few hundred million years.

The ring(s) of fire

Plot all the earthquakes and volcanoes in the world. The vast majority of them are along lines which more or less outline the continents.

Ocean trenches

There are often ocean trenches near these areas of volcanism and earthquakes. Really, really deep ones, with no obvious explanation for how they are formed.

Wadati–Benioff zone

If you plot the depth and location of earthquakes in these areas, they line up to form a tongue pointing down from the trench deep into the mantle

Seismic tomography

If you set up a lot of seismic detectors and use tomographic algorithms on the waves detected after natural earthquakes, you can see the remnants of plates that have been subducted, matching the Wadati-Benioff pattern closely.

GPS measurements

There are thousands of GPS stations set up around the world, and these networks clearly show movement of bedrock coordinated across large areas, as if they're all part of a single plate that is moving in a mostly rigid way.

Actual physical measurements

Direct measurements across faults with lasers or even just cables also show movement. And of course, lots of evidence showing dramatic movement across faults during earthquakes, so we know large chunks of the Earth can move on at least a local scale.

Measured plate motion points at trenches or areas of collision

When the motion tells us that plates should be colliding, we see evidence of that collision in the form of subduction trenches or huge crumpled mountain ranges.

Crust density

Oceanic crust is (on average) much denser than continental crust. And the continental crust (as measured watching seismic waves as before) is ~10x thicker than oceanic, with as much going above the level of oceanic crust as below. It really looks like they're both floating on something.

Seamount chains

If you plot the depth of the seafloor, you can find long lines of mountains, some entirely submerged, some still poking above the surface as islands. Not only that, but there are multiple such lines, all parallel to each other. The ages of them are also linear -- the oldest at one end, the youngest at the other. And this pattern also matches between different, parallel lines. It looks like there are points which cause volcanoes, over which the seafloor itself is slowly moving.

Supervolcano chains

You can find similar chains within continents, but here it's in the form of huge calderas caused by supervolcano explosions.

Coastal subsidence after massive offshore earthquakes

One type of particularly powerful earthquake happens not far off the coast of places with trenches. When they happen, they generate dangerous tsunamis which imply a lot of seafloor got lifted very quickly. And on the shore near where this happened, the land drops by several meters, as if the crust was getting pushed up slowly as something was grinding down between it, then this tension gets released all at once. This has been shown indirectly in the geological record, and also directly measured with those GPS stations in the recent 2011 Tohoku earthquake of Japan.

Old, dead volcanoes where we think subduction used to be happening.

Currently there is only a fairly small section of subduction going on along the west coast of North America, the Juan de Fuca plate. In another 10 million years it will be entirely subducted under North America and the volcanism of the Cascades will come to an end. But if you run back the clock, it's obvious that this section used to be much wider. And inland along those areas are the remnants of volcanoes, whose estimated ages of death all fit nicely for when the subduction would have stopped at that point.