The rocky desert stretches as far as the eye can see: a fascinating vista, forbidding and seemingly unending. Distant hills shimmer in the heat and glare of the Sun. There is beauty here but it is on an inhuman and unearthly scale. The land has been baked bone-dry by many years of sun, heat and but little rain. Beneath, there is liquid to be found, but it is not water. Much of the oil which keeps our world moving was found here. There is another liquid down there too: molten rock.
Saudi Arabia is a land of contrasts, caught between different seas. To the west, the Red Sea. This side of Saudi Arabia is build on the ancient Arabian Shield, pre-Cambrian with some rocks over 3 billion years old although most is less than 1 billion years. There is old flood basalt here too, evidence of a tortured past. Along the Red Sea is the youngest part of the country: a low-lying strip of land less than 100 kilometer wide, a young scar in an old land. To the east, the Persian Gulf. Here the land becomes younger; this region is called the Arabian Platform, to distinguish it from the old shield. The oil is found at the eastern edge of Saudi Arabia. The Ghawar field, on the Persian Gulf, is the largest oil field in the world. Once, the tropical Tethys ocean was here. At times the Thetys became oxygen-depleted: the anoxic events lead to widespread die-off, forming thick layers of black mud. Over millions of year the mud turned to oil. The Tethys disappeared 10 million years ago but the oil remained: riches of the past. It is hard to believe that this unyielding desert was once a tropical ocean, teaming with life. How times change.
Two mountain ranges are stretched out along the Red Sea. The Hajiz mountains begin near the Gulf of Aqaba, and reach over 2 kilometers high; the range peters out near Jeddah. To the south are the Asir mountains, a higher range with peaks near 3 kilometers tall: the range reaches into Yemen where it becomes even higher. The mountains form the western edge of the Arabian Shield. To the west, they drop precipitously to the plain along the Red Sea (‘Hajiz’ meaning ‘barrier’), forming an escarpment. To the east, in-land, the drop is much more gradual. The Asir mountains are the wettest part of the Arabian peninsula, and some high areas receive an annual rainfall which would not be out of place in the UK. The rains have decreased over time. 5000 years ago, there was a major river flowing from the Hajiz towards Kuwait. Now, only a trace remains: the Wadi Bisha. Climate change -for the worse- is not new. There is a third mountain range, on the Indian ocean side, but located in Oman rather than Saudi Arabia. Near the centre of Saudi Arabia, east of the Hajiz mountains, are the remains of an ancient continental microplate. The microplate contains a memory of the formation and break-up of Pangea, and of the Rodinia supercontinent which came before. Continents tend to break up along old sutures, re-opening old wounds and leaving slivers in their wake.
And now a new wound has formed; the Earth has stirred and the Arabian Shield has split. Half is stranded in Eritrea and Ethiopia, where it is called the Nubian Shield. The Red Sea has invaded the wound. Arabia became a peninsula, surrounded by the Red Sea, the Gulf of Aden, the Arabian Sea (part of the Indian Ocean), the Gulf of Oman, the shallow Persian Gulf (also known as the Arabian Gulf), and the Gulf of Aqaba. The split has formed the Arabian plate, the youngest of the main continental plates. The plate is slowly rotating anti-clock wise, pushing into and fracturing the Golan heights.
The stirrings started in (aptly named) Afar, where the Red Sea, the Gulf of Aden and the Ethiopian (or East African) rift meet. Afar is a perfect example of a triple junction, distinguished by the typical 120 degree angles between the three rifts. The Afar triple junction is not actually located in the Red Sea, but on land in Africa, at the westernmost point of Djibouti. The three rifts have fragmented old Africa into three plates: the Arabian plate, Nubia (Africa), and the Somalian plate.
To see more evidence of this process, one has to look under water, an appealing thought in this dry land. The seafloor topography (see the Landsat image above) shows the Carlsberg Ridge within the Indian Ocean: this is the spreading centre that long ago split India from Africa and pushed it into Asia. Near Arabia the Carlsberg Ridge jumps north along a fault (the Owen Fracture Zone); it continues on the other side into the Arabian Sea, but than changes direction and enters into the Gulf of Aden. Near the Afar triple junction it disappears. A little beyond, in the middle of the Red Sea another spreading ridge can be seen. Back in the Indian Ocean, there is a subduction zone along the coast of Pakistan and Iran (the Makran trench, location of an M8.1 earthquake in 1945) but it is comparatively small.
The Owen Fracture Zone is a 1000-km long strike-slip fault. This is the meeting place between the Arabian and Indian plate. It is considered a transform fault (i.e. with sideways motion), with a relative motion of about 3 mm per year, which would make Arabia a remarkably slow-moving continent. In reality, the velocity of the Arabian plate is a more respectable, albeit still slow, 1-2 cm per year, but the Indian and Arabian plate have almost the same velocity and cancel each other out. Along the Owen Fracture Zone, on the Arabian side, is the Owen ridge, reaching up to 2 km above the sea floor. It is a sizable mountain chain, but fully hidden underneath the ocean.
The Red ocean?
The Afar plume, which initiated the break-up of this part of Africa, was active 27 million year ago. The Red Sea began as a continental rift which formed over a short period 24 million years ago. The sea first entered the basin about 20 million years ago. Sea-floor spreading along the southern Red Sea dates to about 5 million year ago, but only in the past million year has it started in the northern Red Sea.
The Red Sea is currently 2 kilometers deep: it is becoming a proper deep-sea basin, and is considered a proto-oceanic basin. But it took a long time to develop a sea-floor spreading ridge. The Gulf of Aden rifted at approximately the same time, but sea floor spreading started here already 16 million year ago: the Red Sea took twice as long. The Ethiopian rift is even slower: so far a few lakes have formed, but nothing more: it is widening at no more than 2.5 mm per year and the faults along its edges are now inactive. Before sea floor spreading can start, the rift has to drop below sea level. This is still a long way from happening in the Ethiopian rift.
The Red Sea has taken 24 million years to grow to 350 kilometer width. That is not particularly fast: it has grown at about 1.5 cm per year. The Gulf of Aden has a similar problem: although it had a 10-million-year head start in ocean-forming, it is still only 500 kilometer wide. The geological advertisements promised more than has been delivered. Why this reluctance to become an ocean?
The spreading rate of the Red Sea is four times slower than the Atlantic Ocean which is one of the slower mid-oceanic spreading ridges. The North Atlantic also had a very slow start, similar to the Red Sea, but it did in the end become oceanic. There is hope. The location of the Red Sea spreading centre is a difficult one. The spreading centre is trying to push Arabia north, but this runs into immovable Asia. Iran is giving it some space by moving north (pushing up its own mountain chain in the process), and the Red Sea has been able to fill the space vacated by Iran. It also tries to push Africa south, but Africa is a very big and very deep continent and is hard to shift. The Red Sea is thus meeting stiff resistance. It is truly a titanic battle between an irresistable force and not one but two immovable continents. The going is tough.
This may in fact be an excellent example of the ‘pull’ versus ‘push’ controversy: is continental drift caused by spreading centres pushing new crust out, or by subduction zones pulling old crust in? The Red Sea has plenty of northward push, but there is little or no pull acting on Arabia, and it isn’t working well. Arabia is becoming compressed. The Ethiopian rift is working in an east-west direction, trying to push Somalia out to sea, but this is going even slower. Continents apparently drift little without help from subduction zones. For continents, push doesn’t come to shove: push without pull won’t do.
The images below depict the evolution of the rifting, starting with a reconstruction of the continental arrangement of 30 million year ago, just before the Afar plume made its appearance. It shows how the plate separated and how the spreading ridges progressed into the rifts. The final panel shows what may happen in the future, at a time when the Persian Gulf has disappeared.
During the time of the Afar plume, 26-28 million years ago, volcanic eruptions badly affected Saudi Arabia. The outpourings of flood basalts left lava fields covering large areas, which can still be seen in the landscapes inland from the Red Sea coast. Lava fields are called ‘harraat’ (singular: ‘harrah’), or ‘harrat’ if used as part of a name. This word occurs in many localities. The old ones dating from the Afar plume appearance are appropriately called the Older Harrats. Around this time the western and southern margins of the Arabian Plate were uplifted, creating what is now the Red Sea Escarpment and the mountains beyond. Volcanic activity diminished after this.
Volcanoes re-emerged 13 million years ago, perhaps in response to the collision between Iran and Asia. A series of haraats were deposited: Uwayrid, Khaybar, and Rahat, 10-12 million years ago. These are, for obvious reasons, called the Younger Harrats.The Younger Harrats are typically less than 300 meter thick. Eruptions continued along the newly formed continental edge until the present day, albeit weaker than before.
The most recent documented eruption of a volcano in Saudi Arabia was in 1256, near Medinah (known as Madinah in Arabic, which is also how it is pronounced). It came from a fissure within the extensive Harrat Rahat. The event started on June 5th with an earthquake swarm. The strongest earthquake hit 4 days later and it was followed by fountaining from a rift, only 19 kilometer southeast of Medinah. Eye witnesses, albeit paraphrased and reported 300 years later, give a vivid account of the events (from Camp et al., 1987, Bull Volcanology, 49, 489 ):
For days the volcanic eruption was preceded by many great earthquakes which occured in Madinah at the beginning of Jumad-Al-Thani (Monday 1st June 1256). At first the movements were slight and not all of the residents of Madinah town felt them. On Tuesday, the second day of Jumad Al-Thani, the earthquakes became stronger. On Wednesday, in the third part of the night, the greatest earthquake occured, which frightened the residents. The earth tremors continued throughout the rest of the night. On Friday, a major event occured, when the ground and ceiling of houses were shaken. Eighteen earthquakes were recorded during this day. […] After the main earthquake was felt in Medinah at midday, fire appeared associated with black smoke clouds which accumulated in the atmosphere. The greatest fire covered the horizon to the east of Madinah. […] The lava flow carried along its way gravel, stones and trees. It was like a red-blue boiling river, with thundering noises. The lava flow moved toward the north of Wadi Eheline. The light of the fire was seen in Makkah, Busrah, and Taima. The historians wrote that the fire continued for three months […] The lava flow descended on the rocky ground and was as high as a long spear above the ground level […] When the lava flow came to a complete stop […] it created a dam that formed a great lake in the rainy seasons (Lake Al Habas).
(The ‘third part of the night’ means the hours before sunrise, the night being divided into three parts. ‘High as a long spear’ indicates that the lava flow was a couple of meters thick.)
The eruption formed six (unnamed) scoria cones along a 2.25-kilometer rift. Over the next 52 days several lava flows extended 10 kilometers from the rift, and one much longer flow came to within 7 kilometer of the old city, within the area of the current suburbs, turning north just in time and sparing the city. About 0.5 km3 of lava was erupted. The lava is a mixture of two components, one of which had fractionated in the mantle and came up rapidly, and one which had been stored in a higher-lying magma chamber in the crust and included some crustal melt. The presence of both a magma chamber in the mantle and one in the crust is an interesting complexity.
The 1256 flow field is located near the northern edge of the Harrat Rahat: this lava field, of which the 1256 eruption was the latest twitch, stretches from Medinah to Makkah, measuring 300 kilometers tip-to-tip, with an average width of 60 kilometers. It contains 644 scoria cones, 36 shield volcanoes and 24 domes. At times the lava flowed over the escarpment onto the coastal plain. The activity migrated north with time, and the main forcus is now south of Medinah. Future eruptions from the Harrat Rahat will probably occur in this region, and eruptions close to or within the city cannot be excluded. In fact another minor eruption occured in 641 producing several aligned cinder cones southwest of the city, lighting up the night sky.
Near the western margin of Harrat Kishb, 100 km southeast of Medinah, is the 2-km wide Wahba crater which was probably formed in a phreatic explosion. It is the largest crater of its kind in Saudi Arabia. The date is not well known but it has been estimated as 10,000 year ago, with a minimum age of 4,500 year. The crater provides a spectacular sight, but being a hole in the ground, is only visible from close-up.
Geothermal activity is also present. Hot springs with temperatures ranging from 50 to over 100 degrees occur on and off the escarpment, mainly south of Jeddah. Fumaroles are present, e.g. on Harrat Ithnayn and Harrat Khaybar, but the steam may only become visible during the coldest time of the year. More worryingly, weak fumaroles are present along a 3 kilometer line close to Medinah, coincident with occasional earthquake activity. This area is closely monitored; there is currently no indication of increasing activity.
Bedouin sources indicate that there was an eruption from one of the cones of Harrat ‘Uwayrid volcano, midway between Medinah and the Sinai, around the year 640. It is not known precisely which cone erupted and the eruption appears to have been a minor one.
In April to June 2009, in a previously seismically quiet area in Harrat Al-Shaqah (also known as Harrat Lunayyir) 150 kilometer northwest of Medina, an earthquake swarm occured similar to that in 2014 in Bardarbunga. The figure shows the distribution of the quakes, along the linear dyke: it was 20 kilometer long, and came up from more than 20 km depth at the southern end to less than 7 km at the northern end. An 8-km-long surface rupture occurred during this event, with a vertical offset of 90 cm between both sides. The 2 meter thick dyke was filled with an estimated 0.13 km3 of magma. The magma managed to get to within 2 kilometer of the surface but the last bit of crust was too rigid and it failed to break through. For a while an eruption was considered possible, and as a precaution 20,000 people were evacuated. The event happened underneath an existing lava field which may have been erupted around the year 1000.
A natural marvel is the area known as the Black and White Volcanoes. It lies within Harrat Khaybar, due north of Medina. Two volcanoes here have erupted a cream-white lava, a felsic silicate called comendite. Nearby is Jabal Qidr, a dome of pitch-black basalt, a hawaiite. The division between the two lavas is stark. There are neolithic stone walls on the black volcanic slopes, and some of them have been overtopped with lava. There may have been as many as seven eruptions since people started building the walls, and Jabal Qidr may have erupted as recently as the year 1800. The two white mountains, Jabal Abyad, at 2,093 meters and Jabal Bayda at 1,913 meters, are the two highest volcanoes of Saudi Arabia. (Both words mean ‘white mountain’, one with the female and one with the male adjective.) The pairing would have been a wonder of the world, but it is in an isolated and desolate region, difficult to reach. The contrasting lavas have come from a single magma chamber in the crust underneath, which is extremely stratified.
No less wondrous are the gems found in the area. Beautiful peridots, a form of olivine, are embedded in the basaltic harraats. The peridots from this region are of a pure green, a precious colour in the desert: the peridots are highly valued. The most recent find was in Harrat Kishb.
The rift itself, now deep below the sea, is also volcanically active. A number of volcanic islands have formed within the Red Sea, near the Afar triple point. On 30 Sept 2007 there was an eruption midway between both Red Sea coasts, on the small island of Jabal al-Tair. In 2011 and 2013 there were Surtseyan-type, submarine eruptions in the Zubair group of islands, further south along the coast of Yemen towards Djibouti, where new islands emerged from the sea. Volcanic activity further north along the younger Red Sea spreading centre is much less, although there are deep hydrothermal vents. There is no sign that the extensive flood basalt activity prior to the opening of the Red Sea will one day resume. Lesser, intermittent activity remains common, however.
The potentially active volcanoes within Saudi Arabia are offset from the Red Sea spreading centre by about 100-150 kilometers. One can speculate why this is. Around Medinah, the oldest lava fields are 2 million years old, but these are in almost the same location as the new activity. Over that time, the Arabian plate has moved by perhaps 50 kilometer due to the widening of the Red Sea. The volcanoes have moved with the plate. This means that the magma is located in the continental plate and does not come from the oceanic Red Sea spreading centre. The felsic lavas confirm that the magma has been stored in the crust and is not currently coming directly from the mantle, but the predominant lava is still basaltic, i.e. originally derived from the mantle.
The magmatic activity may be due to local crustal extension. Alternatively, the magma may have been emplaced at the start of the seafloor spreading, but this begs the question what kept it from cooling. A sub-crustal flow of magma from the triple point has also been suggested, but the level of activity does not seem to decline with distance from Afar which argues against it. Crustal extension therefore seems most plausible. Especially the northern-most Red Sea does not yet have an active spreading centre, and without new oceanic crust being formed, the movement of the Arabian Plate has to be accommodated by the crust, leaving a line of weakness around Medina.
An interesting suggestion is that the line of the Younger Harrats, which is under a slightly different angle to that of the Red Sea, is a separate rifting event. There seems little direct support for this, but if this is correct we may end up with two parallel oceans, separated by a sliver of Arabia.
State of the nation
At the current time, volcanic eruptions in Saudi Arabia tend to be minor to medium in size. Several eruptions were reported about a millennium ago but there has been little activity since 1256. In the empty desert, more recent eruptions may have been missed. An eruption in the near future is not unlikely. Although it would probably not be a large event, Saudi Arabia could still be supplementing its oil with liquid rock. Just remember the word ‘harrat’.
With thanks to Samira Alharbi