Impacts and Geology: deep peace?

Metamorphic rocks typically come from deep in the earth and form slowly. Simple physics shows that transferring heat into large volumes of rock (a key driver of many types of metamorphism) takes millions of years. Rocks that form the deep crust of stable cratonic areas lead the most placid of lives. They are heated for so long that they become annealed; they have achieved complete chemical, textural and thermodynamic equilibrium, like some sort of silicate-based Buddhist monk.

Some deep crustal rocks in South Africa were once in granulite nirvana and might still be there, if only they hadn’t been hit by the biggest impact known on earth. The slow and calm world of the deep crust was violently attacked from Outer Space and the shocking results are visible in the a thin-section.

The Biggest Impact Structure in the world

Vredefort from Space

Vredefort Dome in Google Earth

The Vredefort structure is a very old (2Ga), very big (>250km diameter) impact crater. It has the usual range of evidence (shatter cones, shocked minerals, coesite/stishovite) and unusually for Earth is multi-ringed. As well as being very wide, it also formed a very deep hole in the ground, which leads to a number of interesting things.

Geologically, the Vredefort structure is a dome, surrounded by an outer synclinorium. The dome itself shows an increase in metamorphic grade towards the centre, from amphibolite through to granulite facies metamorphism. This roughly corresponds to hotter rocks in the middle, so why should this be? Is this because deeper rocks are found in the middle, so we have a view of a cross-section through the pre-impact crust with pre-impact temperature differences preserved? Or, is the metamorphism related to the impact itself, with the intensity of the heating highest in the middle? To answer the question, we need to look into the metamorphic rocks in more detail

Metamorphic Petrologists find an amazing thing

I first heard about the Vredefort structure from a paper on metamorphic petrology by Stevens, Gibson and Droop from 1997 (see further reading below) . It is quite old now, but I still like it as it is such a Geological paper, and has photomicrographs of lovely rocks. It also does what Science should, gather together hard-won facts and rigorous argument to tell an exciting picture of the world. In short, it uses dry phrases like “petrogenetic modelling in the simplified KFMASH system” to show that at least 9 km thickness of rock was ‘instantaneously’ removed from the crater.

When it was published, there was still a debate about whether the structure was caused by an impact or by some other cause. It is significant as it helped to settle the argument in favour of an impact, by providing independent evidence from a detailed study of the rocks themselves.

Stevens et. al use the usual techniques of metamorphic petrology to describe the granulitic rocks towards the centre of the dome. They describe mineral assemblages and textures, and mineral chemistry. Often rocks contain multiple sets of mineral assemblages and these rocks are no exception. Using cross-cutting relationships, they constructed the following time-sequence of mineral assemblages (which I have greatly simplified):

  1. Peak metamorphism: rocks showing evidence of melting, containg garnet-orthopyroxene-cordierite
  2. Pseudotachylite veins: veins of glass, cross-cutting the peak metamorphic fabrics
  3. Post-shock overprint: recrystallisation of the pseudotachylite, coronas around minerals

Detailed analysis of these metamorphic assemblages allow them to reconstruct the history of conditions these rocks experience (a Pressure-Temperature-time  or P-T-t path). The peak metamorphism shows the kind of things that typically happen to rocks in the crust: an increase in pressure and temperature until the rocks melt, reaching 900°C and 0.6GPa pressure (about 20km depth). There is evidence of cooling, but this is not quantified. Later papers suggest the peak happened at around 3Ga, giving the rock about a billion years of peaceful cooling.

The pseudotachylite veins are linked to the impact event. They represent a fracturing event in the deep crust, where the energy released by the fracturing melts the surrounding rock.

The post-shock overprint is associated with recrystallisation of the glass at high temperatures, but also with coronas, new minerals forming around older grains. This texture is often associated with decompression, and thermobarometry on these mineral assemblages gives an estimate of pressure far lower than for the peak. Why has the pressure decreased? Stevens et. al. say because about 9km of rock has been nearly instantaneously removed from above the rocks being studied.

So to answer our question earlier, why are the rocks in the middle of the Vredefort Dome formed under hotter conditions? Most of the minerals in these rocks represent the peak metamorphism, the record of the later metamorphism is only really visible under the microscope.  The hotter rocks are in the middle, because that is where the crater was of the greatest depth.

Metamorphic petrologists continue to find amazing things

The Stephens et al. paper is quite old now and such interesting geology has continued to attract attention. This rather good online thesis gives a useful overview of current thinking. I’ll quickly draw out some interesting themes.

Post-impact rebound and shape of the crater

Theophilus Crater on the Moon. An analogy for the Vredefort structure, note central uplift and surrounding terrace. Image courtesy of Nasa via Wikimedia

Isostatic rebound is the phenomena where the crust responds to load placed upon it. The best example is in areas that were covered by major ice-caps during the recent Ice Age. Places such as Scotland and Scandinavia were covered by kilometres of ice, which pressed the crust down. Now the ice has gone, the crust is rising up again (on a scale of 1000s of years). This creates raised beaches as the crust rises and the sea doesn’t.

In a similar way, suddenly removing kilometres of rock from the crust will cause the remaining underlying rocks to rebound up over a period of time. In order to found out exactly how, you can either study the Vredefort structure, or go to the Moon or Mars.

As well as removing material, the impact puts a lot of heat into the remaining crust. For complex craters they start to behave as Bingham Fluids (like toothpaste). The centre of the dome rises up and the outer areas slip down towards the centre, creating a terraced rim. The large volumes of pseudotachylite in the Vredefort area are thought to be as much related to this post-impact movement as to the actual impact itself.

Shock metamorphism and recovery heating

One thing that can certainly be attributed directly to the impact, is shock metamorphism. This has been mapped across the area, with pressures of at least 30GPa (50 times greater than the ‘peak metamorphism’ above) found in the centre of the dome.

Shock metamorphism puts energy into the rocks, and as the pressure decays, this is released as heat. The post-shock overprint is prominent, both because the rocks were still hot, but also because they were further heated by the thermal overprint of the shock metamorphism. Temperatures reached up to 1000°C, with shock heating adding 500°C to already hot rocks. So, both the pre- and post-impact metamorphic events are at higher temperatures in the middle.

A natural laboratory

In order to estimate the conditions under which rocks are metamorphosed, petrologists generally have to assume they achieved chemical equilibrium. This is arguably true for slowly heated rocks, but clearly not true for shock metamorphism, or indeed for the post-shock overprint, where the rocks would have quickly cooled (since they are now close to the surface). The main thrust of this recent thesis is how getting a handle on metamorphic rocks that are not in full equilibrium can help us better understand all types of metamorphism. As with people, metamorphic rocks are always encountering changing conditions and rarely achieve total inner peace. Temporary partial equilibrium is the best most of us can manage and techniques that don’t assume metamorphic rocks achieved nirvana are to be welcomed.


Mid-crustal granulite facies metamorphism in the Central Kaapvaal craton: the Bushveld Complex connection: doi:10.1016/S0301-9268(96)00043-5. This is the 1997 paper I mention. The reference to the Bushveld complex is related to the idea that the local temperatures were elevated due to its intrusion.

Metamorphic studies in the Vredefort Dome, South Africa. This is a 2010 PhD thesis by Paula Ogilvie. I’ve mined it for its clear updates of current thinking but I’ve only started to begin to get my head around its main arguments. I can’t find this work published yet, but it will be soon, I suspect. If you’ve found my referencing a bit sparse, this is the place to come.

Be Sociable, Share!

About Metageologist

Simon Wellings trained as a Geologist but professionally has metamorphosed into something else. He retains a keen interest in Geology, facilitated mainly by the wonders of the Internet. Simon now blogs at Metageologist.
Categories: planets, Rocks & minerals
Tags: , , , ,

Comments (3)

  1. Chris Rowan says:

    9 km of crust? That’s pretty mindblowing. How interesting that it was the metamorphic petrologists who came up with the clinching evidence for an impact origin. Changes over 10s of millions of years are more their thing – at least in my mind, which may need to be adjusted.

    I’m also interested when you say:

    The large volumes of pseudotachylite in the Vredefort area are thought to be as much related to this post-impact movement as to the actual impact itself.

    Presumably this is collapse along ring faults, etc.?

  2. I’ve heard the term “suevite” applied to impact melt glass — is there a clear reason to favor pseudotachylite in some circumstances and suevite in others?

  3. I believe suevite is a surface deposit (breccia) including impact glass and other material. That certainly is the case with suevite from Scotland I happen to know about.

    Pseudotachylite can be related to non-impact faulting (caldera ring-faults, that sort of thing) and is I think always formed in the subsurface. It is typically seen in veins within rock, but in the Vredefort area there is so much of it that the rock is texturally very similar to a breccia, which I guess may lead to confusion.

    I would tentatively suggest that your sample is mis-labelled. Looking at this site makes me a bit more certain. In your sample it looks like the matrix is the melt itself, whereas the bit of Scottish suevite on my desk has a sedimentary matrix with bits of melt in it. That seems to be the important distinction.

    Links (7)
  1. Pingback: New at Erratics – the largest meteorite crater in the world | Highly Allochthonous

  2. Pingback: Impact and Geology: spherules rule | Earth Science Erratics

  3. Pingback: What you ought to know about metamorphism | Metageologist

  4. Pingback: Metamorphism: evolving ideas | Metageologist

  5. Pingback: Hurled from the sky | Metageologist

  6. Pingback: How old is plate tectonics? | Metageologist

  7. Pingback: A year of metageologising | Metageologist