When the earth moved: exploring Scotland's volcanic past

Eruptions on Iceland are teaching vulcanologists about how the Inner Hebrides reacted to shifts in tectonic plates 60 millions years ago. And the evidence of volcanoes, dyke swarms and lava fields are evident today if you know where to look, says Robert Muir-Wood

The recent videos from Iceland have been spectacular. First rifts a few metres wide had cut straight through roads, pipelines and buildings. Then, a month later, the same crack was jetting a row of white-hot magma fire-fountains rising 30 metres into the sky in a kilometre long line, the magma falling back and streaming over the rocky landscape.

These scenes from the Reykjanes peninsula in the southwest corner of Iceland reveal where the plate boundary in the Northern Atlantic is widening as Greenland moves away from Europe by two centimetres a year. Along the boundary the widening is intermittent, concentrating in one segment that opens a metre or two and then, after a few years, shifts to another segment.

Hide Ad
Hide Ad

The surface rifting forms where a deep vertical crack has opened, widened, and partially filled with magma. Where the magma pressure increases, the fire fountains burst out of the same wide crack, feeding lava flows that can spread over enormous areas. Beneath the surface the magma-filled crack will solidify into an igneous dyke.

The Cuillin Hills are evidence of the Isle of Skye’s volcanic pastThe Cuillin Hills are evidence of the Isle of Skye’s volcanic past
The Cuillin Hills are evidence of the Isle of Skye’s volcanic past

If we could time travel back 60 million years, we would encounter the exact same scenes in western Scotland. The northern Atlantic had yet to open. Shetland was nearer to Greenland than to Aberdeen.

A plate boundary had detoured its way through the Inner Hebrides and out through northern Ireland. In geological terms this plate boundary was short-lived, lasting less than two million years, and the spreading ridge only achieved three to four kilometres of opening. This was a rare example of a ‘starter ocean’, a trial attempt to split a continent and form a new spreading ridge, but that in this instance did not succeed.

Typically, when a continent splits and a new spreading ridge plate boundary forms, the original rifting and volcanics sink underwater. Yet here in western Scotland the workings are all on display. It is what makes this region so unique.

Hide Ad
Hide Ad

The 60 million year old volcanoes in western Scotland formed above a ‘plume’, where hot viscous mantle rock rose from hundreds of kilometres underground and melted to form basaltic magma, just as a similar plume feeds volcanoes across Iceland today. The chain of active volcanic mountains that follows the mid-ocean spreading ridges worldwide lies underwater almost everywhere. Beneath Iceland, melting in the plume is creating double or treble the typical amount of magma, and as a result the whole pile of Icelandic basalt lava-flows and volcanoes rises above sea-level, building mountains more than 2,000m high.

Our journey through the Hebrides plate boundary should begin with the original volcano at Rum. This volcano grew into a great mountain perhaps 3,000m high, had some spectacular eruptions, but then became too steep and collapsed. As magma production increased, new volcanoes burst out simultaneously, five each along two parallel ‘volcanic ley lines’.

The eastern line includes Rum, Ardnamurchan, Mull and Arran volcanoes, with Skye making the fifth just off the line at the northern end. The western ‘volcanic ley line’ is much longer and the five volcanoes are more widely spaced. At the northern end there is the most spectacular of all St Kilda volcano, then Blackstones (now entirely underwater), next Slieve Gullion and Carlingford around the Irish border, and at its southern end Lundy 20km off the coast of Devon.

The plate boundary diverted its path through the eastern line of volcanoes because all the heat and melting had made this the weakest link. The crust began to open just like a spreading ridge, through the repeated intrusion of igneous dykes, all aligned NNW, as the opening was in the direction WSW-ENE. Good places to see these multiple dykes, each contributing one to two metres of the spreading, include the coast of Skye south of the Cuillin and the southeastern coast of Mull. The dykes may leave a ditch where they erode faster than the surrounding rock, but when the dykes are tougher they form parallel walls, as along the southern coast of Arran.

Hide Ad
Hide Ad

The dyke can only form if the crust is already extending. When the magma chamber reservoir was full and pressurised but there was nowhere nearby extending, a horizontal sill intrusion could form instead of a dyke. There are great examples of sills on the cliffs on the eastern side of Isle of Raasay. Sills of solidified basaltic magma between layers of sandstone and shale confused early geologists, who reasoned that they had somehow crystallised out of cold water.

There was more magma than was needed to fill the widening cracks, so just as in Iceland today, the magma burst out in fire fountains streaming over the landscape. In the ‘ziggurat’ landscape on the Ardmeanach peninsula in southwest Mull you see a great stack of lava flows, each flat-topped and of very constant thickness, showing that the magma flowed as readily as water. In Skye the lava pile reached 1500m thick. At Mull 1800m of lavas have survived, all erupted in less than 1.6 million years. The pile of lava-flows on Iceland is more than four times as thick, and the oldest date to 13.4 million years ago, one quarter the age of the Hebrides volcanoes.

At Skye the crust has opened by three to four kilometres of ‘seafloor spreading’. On mid ocean spreading ridges, one section of spreading ridge connects to the next through horizontal displacement faults known as transform faults. There are several candidate transform faults through Skye and the Hebrides and through northern Ireland.

Iceland has taught us a lot about how all the pieces fit together. It used to be assumed the igneous dykes were fed vertically. Monitoring of dyke formation in Iceland over the last 40 years has shown in every instance that the dykes formed when a crack burst sideways out of the magma chamber located under one of the principal volcanoes. The magma pressure could force a crack running fifty kilometres out from the volcano.

Hide Ad
Hide Ad

The Skye, Mull and Arran volcanoes were all feeding this new spreading ridge plate boundary by intruding multiple igneous dykes, identical to those that form oceanic crust. The crust widened as each dyke was intruded. If the spreading had continued, the Outer Hebrides would today be a thousand miles away on the other side of an ocean.

Eventually, after an attempt to split the whole of Britain with giant dykes, ten times as long and ten times as wide as the standard dyke, the new plate boundary shifted a few hundred miles to the west. Even after all the volcanoes had become extinct, the western volcanoes continued to influence the tectonics for the next 30 million years as a fault system, passing through Devon and northern Ireland, connecting the Alpine to the North Atlantic plate boundaries.

On your next visit to the Inner Hebrides, look out for the volcanoes, dyke swarms, and lava fields that are traces of the short-lived plate boundary that once passed through these parts. Today all these volcanoes are in ruins, like ancient castles for which only some of the dungeons and foundations remain. Yet for a volcanologist, it is the eroded ruins that that reveal a volcano’s anatomical workings.

The landscape of Britain, read with geological insight, reveals an extraordinarily rich and varied tectonic history.

​This Volcanic Isle: The Violent Processes that forged the British Landscape by Robert Muir-Wood (OUP, £20)