Oldest rocks in Africa

Le 25 avril 2008, par Jeff,

Visiting a few of the english speaking geo-blogs decided me to dig into my photo collection and add a few of them on this site. And for a change, this will be in English — it will do no harm to the French-speaking readership to be exposed to a bit of English, and it may even lure the English speakers into the French pages of this site.

So we’ll start with some field photos of the oldest rocks of Africa.

They occur in the North-Western corner of Swaziland, just North of the town of Pigg’s Peak near Phonphoyane falls. Here, a small remnant of the "Ancient Gneiss Complex" (dark blue on the map) that underlies most of Swaziland is preserved, wedged between the flank of the (3.55—3.21 Ga) Barberton Greenstone Belt (multiple colours, the base is dark green), and the ca. 3.1 Ga Piggs Peak batholith (orange).

JPG - 121.5 ko
Geological map on Google image.
The outcrop is at about S25°56, E31°16

This is what the outcrop looks like :

JPG - 90.9 ko

Not very impressive, but at least it’s accessible (ah yes, I can hear you structural geologists muttering about "geochemists" and "road side geology"...).

Progressively zooming on the outcrop :

JPG - 105.9 ko
JPG - 125.1 ko
JPG - 124.4 ko
JPG - 113.6 ko
JPG - 106.9 ko

The rocks here are migmatitic tonalitic gneisses. You see that they are very composite (as Archaean gneiss complexes commonly are), with at least 4 components (in roughly chronological order) :

- Dark grey enclaves of amphibolite or tonalitic gneiss ;
- Lighter grey bands of tonalitic gneiss ;
- Leucocratic bands of granodioritic to granitic gneisses (probably products of partial melting of the above) ;
- Cutting white pegmatites and dykes (prob. associaited with the 3.1 Ga Pigg’s Peak batholith).

Compston and Kröner (1988) [1] took a sample from this locality (ACG150). Zircons were separated and analyzed, here is the resulting concordia diagram (if you know what to make of that — if not, sorry it’s a bit long to explain) :

JPG - 73.9 ko

Confused, isn’t it ?Looks like this rock recorded all the regional geological history. Compston and Kröner propose that this complex data set can be split into no less than 5 groups :

- The oldest group, made of magmatic-looking zircons, gives an age of 3644 ± 2 Ma, interpreted as the age of the emplacement of the protolith of the rock (more precisely, I think, of the "light grey" gneisses, but they do not record any older age for the dark enclaves) ;
- The next three groups comprise metamorphic zircons giving ages of ca. 3580, ca. 3500 and 3433 ± 4 Ma. All these ages correspond to geological events well-recorded in the nearby Barberton belt.
- Finally, the "younger" zircon are ca. 3100 Ma old, consistent with the emplacement of the Piggs Peak batholith.

So, we’re looking at 3644 ± 2 Ma old magmatic rocks, that have been metamorphozed and deformed several times.

Oh, yes : and here’s what happens when an over-enthousiastic geologist tries to take a piece of sample...

JPG - 99.6 ko

Notes :

[1] EPSL 87:13-28

Téléchargez cet article au format pdf : Enregistrer au format PDF

Attention !

Suite à un bug que je n’arrive pas à résoudre, vous êtes peut-être arrivés sur cette page, ou d’ailleurs sur n’importe quelle page du site, avec une adresse (url) incorrecte. Si c’est le cas, il y a des choses qui marcheront mal (documents liés, commentaires...).

Les adresses correctes sont de la forme

http://jfmoyen.free.fr/spip.php ?articleXXX

Toute autre version (avec des choses en plus entre le "... free.fr/" et le "spip.php") n’est pas bonne.

Si vous n’arrivez pas à écrire un commentaire ou voir une photo, vérifiez que vous pointez bien sur la bonne adresse ; si vous faites un lien vers ce site, merci d’utiliser la version correcte de l’URL.

Fatal error: Cannot redeclare boucle_traductionhtml_8577f9e6e58b382df53aae02e8a21c02() (previously declared in /mnt/133/sdb/8/7/jfmoyen/ecrire/public/composer.php(48) : eval()'d code:11) in /mnt/133/sdb/8/7/jfmoyen/ecrire/public/composer.php(48) : eval()'d code on line 187