The Making of La Palma

After Iceland, there was La Palma. It could have been the Azores, of course. There are more than just two volcanic archipelagos and islands in our youngest ocean, the Atlantic Ocean. But it was La Palma. Could we have seen it coming? Eruptions at La Palma are about ten times less frequent than at Iceland, perhaps two or three times per century. Still, it would happen eventually. And to put it into perspective, the chances of an eruption in La Palma were much higher than one in Reykjanes.

Image source : Twitter @jonni_walker

The eruption began on Sunday 19 Sept, at 3:12pm, with a small bang. As is common in La Palma eruptions, the dike which had reached close to the surface was contained by the older lava flows above. Eventually the pressure reached breaking point, an explosion occurred which blew out the old lava (the brown clouds), and an opening was made for the new lava to come out. It happened on the west facing slope of the Cumbre Vieja volcano where two fissures formed, each about 200 meters long. At peak there were 11 vents in action but one vent became dominant. (Steam is apparently still rising from other vents though.) A common aspect of La Palma eruptions is that the vents form near the cones of previous eruptions, possibly because the ground there is easier to break through. This eruption too came from an area riddled with previous eruption cones, and the current vent is next to an old cone (now presumably buried). Disastrously, the erupting vent was on a slope above and close to a populated area. The lava is relatively cool and slow flowing, but as of this morning, 390 homes have been lost. A thousand people may now be homeless. A few houses may have been second homes (still a terrible loss – in spite of some comments made, most owners of holiday homes are not rich) but for the large majority the house will be everything those people possessed. We live at the mercy of the earth.

The location of the dominant vent

The ESA emergency satellite mapping service sprang into action. Images of the flows show the location and expansion better than the ground based press ever could. This is not a tourist eruption and we do not have the wealth of images and videos that Iceland has produced. Those are for ‘nice’ volcanoes, not for human disasters.

Copernicus radar images of the lava flows

Eruptions here are not highly explosive. Still, it seems to be evolving towards a bigger bangs. The eruption rate is not too high (perhaps 50 m3/s) and that also is helpful.

How will it continue? Previous eruptions have jumped between different vents at different times and that could happen this time as well. And previous eruptions have lasted between 24 and 84 days – we still have a way to go.

La Palma

The Canary Islands are a group of larger and smaller islands, are volcanic, and are located off the coast of southern Morocco. We know one of them very well from previous activity: the submarine eruption of El Hierro, south of La Palma. Tenerife has the largest volcano and Lanzarote had the longest lasting historical eruption. The entire region remains active. This differs from a conventional hot spot where one may see extinct volcanoes further from the current location of the spot. Here the heat is distributed over a larger area. There is however a bit of a gradient, in that La Palma is still in the shield building phase while islands to the east did that a long time ago. The heat may be slowly migrating west.

Canary Islands. Source: wikipedia

The fifth largest of the Canary Islands, La Palma is mountainous. The roads make for interesting driving as they wind up the sides of the mountains. Seeing upside down cars is not uncommon (on one drive I saw two). One main road goes underneath an active volcano by means of a tunnel: this may be a unicum, but do drive carefully especially when exiting the volcano (where I saw one of those two overturned cars).

From the Digital Terrain Model of the National Geographical Institute of Spain (reduced resolution). Sourced from

The mountains form an impressive volcano. The peak is 2430 meters above the sea level but this underestimates the true size of the beast. The base of the volcano is 4 kilometers below sea, making it over 6 kilometers tall. Submarine volcanoes have an advantage as they can grow much steeper and the water helps carry the weight. They grow faster and taller than their aerially exposed counterparts. Still, this is a large one.

La Palma is a complex island with multiple volcanic features. It has history. A history of five volcanoes, in fact, all of which can still be seen on La Palma.

The map above shows that there are two obvious volcanoes. The northern, large one with a hole in its side is called Taburiente, and it is extinct enough that an astronomical observatory has been build on its peak. (Mind you, another one is build on Mauna Loa.) The big hole is called the Caldera de Taburiente. Standing on the top, the Caldera is very steep (I know from experience). The smaller ridge volcano to the south is Cumbre Vieja, and it is clearly not extinct as it is currently erupting. There is a saddle between the two volcanoes called Cumbre Nueva. Cumbre Vieja, in spite of its name, is the younger of the two: the names ‘Vieja’ and ‘Nueva’ refer not to the ages of these ridges but to the forests on them.

La Palma is quite a rainy island (tourists beware: the beauty of the greenery gives a hint), and erosion has carved deep valleys in the sides of the volcanoes. These are locally known as ‘barrancos’. The barrancos on Taburiente are deep and steep (causing difficult and bendy driving here), whilst they are much less prominent on Cumbre Vieja. Cumbra Vieja is young whilst Taburiente is old and weathered, as lined as the faces of the people I once saw going to work in the cold of Novosibirsk.

But there have been volcanoes here before. The steep sides of the Caldera de Taburiente have cut through the layers that make up the volcano. Not all those layers are from Taburiente: some are much older than the current volcano. Even the old can be young, when seen in comparison to those who came before. We so easily forget that no one feels as old as others think they are: we perceive our own age from our memories. But we all are old at heart: our memories give life to people who may have long since passed away. They haven’t really passed on, until all living memory of them has ended. La Palma has kept those memories of the departed, and has brought them to the open. Let’s dive in.

Submarine volcano (unnamed)

The oldest lavas that are exposed in the Caldera formed under water: pillow lavas are seen, interspersed with tuff and sediment. There are also intrusive rocks that formed from magma injection underground, with many dikes running through them. Although all this formed under water, nowadays they are on dry land: the layers were uplifted later to above sea level. In the Caldera they are seen as high as 400 meters above sea level. Some of the lava flows formed in deep water, other in shallow water: the uplift continued while the subsea volcano was active. We normally think of uplift (also called inflation) as coming from growing magma chambers in the crust. That is part of it, but uplift on this scale has a deeper origin. The mantle and lithosphere below are heating up as a hot spot moves underneath. Hotter rock has a lower density and therefore floats higher, like a cork in the bath. Everything above rises too as the heat increases. This is the main process that causes large scale uplift in volcanic regions, and it is also the reason that when the hot spot moves on, the island it created begins to sink. If you live on a volcanic island, a volcano that stops erupting can be as dangerous to your future as one that is actively burying your land.

Exactly when the submarine volcano first began to grow is not well known. The oldest accurately dated lava (using radio isotope dating) formed 1.7 million years ago. But micro fossils embedded in the sediments are older: 3 to 4 million years. This sounds old for a volcano that hasn’t moved much, but it is young for the Canary Islands. Tenerife is at least 7 million years old, and Gran Canarias may be twice that. Unlike Hawai’i, the volcanic activity here does not migrate, but there has been a slow shift in peak activity from east to west. Interestingly, this westward shift (2-3 cm/yr) is similar to the spreading rate of the Atlantic ocean at this location. The area has kept its distance to the mid-ocean rift and perhaps that is where it gets its heat from. The alternative model of a mantle plume has also not been ruled out though.

The submarine volcano began to grow 4 million years and by 1.7 million years ago may have become an island. The orientation of the dikes changed over this time. Originally they ran southwest to northeast, but as the volcano developed a feeder conduit the dikes rotated to a north-northwest to south-southeast orientation, similar to the north-south elongation that we still see in La Palma.

Garafia volcano

The island volcano that now formed is known as the Garafia volcano, and it is considered the second volcano of La Palma. It is named after a village on the north coast of the island. This volcano grew over a period of half a million years, between 1.8 and 1.2million years ago. It reached an impressive size: Garafia reached 23 kilometers across and was at least 2.5 kilometers tall: this was the first shield volcano. It’s lava flows are exposed not only in the Caldera, but also at the bottom of the deepest barrancos all around the current volcano. They are thin pahoehoe flows, with some explosive layers (lapilli). It was a basaltic volcano, typical for a volcanic island.

Around 1.2 million years Garafia came to an end when a large collapse occurred. In a way this was a typical event. Slope failures are common on the western Canary Islands. The most recent one was on El Hierro, 15,000 years ago: it left a scarp 1 kilometer tall. Garafia produced several such land slides which are seen on the ocean floor. It is not clear whether they are all part of the same event or (perhaps more likely) the collapse happened in several different events. One debris flow lies off the east coast and came from somewhere above Santa Cruz. Three debris flows are seen to the southwest. The four events happened between 0.8 an 1.2 million years ago, and together deposited some 650 km3 of debris on the ocean floor.


On the remnant of Garafia a new volcano now began to grow. It was in pretty much the same location but extended a bit further south where the land slides had removed half of the old volcano. Eventually it fully covered both the remnants of Garafia and the older submarine (uplifted) volcano.

The walls of the Caldera the Taburiente showing the older layers. Source: The Geology of La Palma

Taburiente grew for over half a million years, between 1 million and 500,000 years ago. It reached a height of 3 km, and was 25 km in diameter. Like Garafia, it was basaltic in nature. however, during its later years the eruptions became more explosive (never devastating) and the magma became more evolved. At the same time, the eruptions began to move to the south, away from the ancient location. This formed the Cumbre Nueva ridge. By 400,000 years ago the peak was extinct but this ridge continued to be active.

History now repeated itself, as it always does. The peak of Taburiente and the side of the Cumbre Nueva collapsed into the ocean around 500,000 years ago. Compared to Garafia, this was a smaller event. The debris on the ocean floor has a volume of around 100 km3. And whereas Garafia collapsed in what may have been 3 or 4 separate events, a long time apart, Taburiente only had one. The gap it created was not yet the current Caldera de Taburiente. Erosion has deepened and widened the hole since, and also formed an erosion channel at the bottom: the Barrancos de Las Angustias. And there was new growth on the far side.

Bejenado volcano.

The collapse removed a lot of weight from the southwestern part of the edifice. This allowed new volcanic activity, from decompression and from easy access to the surface. A new volcano began to grow. This one is called Bejenado, and it forms the southern wall of the Caldera. (It is interesting and perhaps confusing that the two sides of the Caldera are from different volcanoes and have different ages.) The current Caldera looks nothing like it did after the collapse: it became enclosed only because of this new growth.

Cumbre Vieja

The new volcano filled in part of the newly formed basin. The activity did not last too long. By 150,000 years ago, all eruptions were from a new volcano, Cumbre Vieja. Unlike the previous volcanoes it formed a curved ridge. It may have formed along a radial rift zone of Taburiente, activated by the southward migration of the heat. The ridge continues into the sea, with a range of sea mounts which are equally active as the part on-land. Although Cumbre Vieja is no longer young (it has reached a height of almost 2 km), it never developed an eruptive centre. The eruptions are along the entire ridge, and are monogenetic which each one forming its own rift (normally on the flanks, at a slight angle to the ridge) and cones. Eruptions are strombolian. The lava is basaltic but more evolved lavas (phonolitic, which contain a much higher fraction of silicate) are common. (The current eruption was reported to produce tephrite, a slightly evolved version of basalt.) This volcano is very different from any of the previous four. Why that is is not clear.

On the west side the ridge has a steep edge, with a coastal platform where much of the banana plantations are. The cliff is caused by sea erosion; the platform has build up from later eruptions.

Historical eruptions occurred in 1585 (84 days), 1646 (80 days), 1677 (66 days), 1712 (56 days), 1949 (38 days) and 1971 (25 days). There is no pattern to the either the frequency or location: this volcano is all over the place. There is however a pattern of decreasing length, as if it lived of a previously formed magma reservoir – we will see whether this hold his time (in which case the eruption will be over by mid October) or not (in which case it could last until December)! There was also an eruption in the late 1400’s, around or just before the time of the Spanish settlement in 1493, but we have no historical record of this.

Events high up tend to be explosive and vents lower on the flanks tend to be effusive – this seems to hold for the current eruption as well. Eruptions often occur near older phonolitic cones. For instance, the 1677 eruption was in San Antonio volcano (a tall cone) but this cone already existed before that time.

La Palma is a fascinating place. It showcases its history well. But it is not a ‘nice’ volcano. Eruptions occur over a long area, and any one location sees lava only rarely. This encourages settlement of regions that are never safe. La Palma’s eruptions are slow and they give people time to leave. This is no Taal. But they are also destructive, as we see now. In a few months time the eruption will be over and the volcano will go to sleep for decades or centuries. But it will take people affected by the eruption a long time to recover. Those memories will not go away

Albert, September 2021

This post is almost entirely based on ‘The Geology of La Palma’ by Valentin Troll and Juan Carlos Carracedo, published as a chapter in the book The Geology of the Canary Islands’ (2001).

To end this post, I am reproducing a very useful overview made on Dec 21 by VC commenter Oliver:

To keep the overview, here is a summary of the previous events and facts of the eruption on La Palma:

– The eruption began on 09/19/2021 at 3:12 pm (local time) on the lower western flank of the “Cumbre Vieja” in the area of ​​the “Cabeza de Vaca” and just above the first houses of the village “El Paraiso”.

– Two eruptive fissures developed, each approx. 200 m long and running in a north-north/westerly direction. On the evening of September 19th, up to 11 vents were active at the same time. This released lava fountains that were several hundred meters high. The VAAC detected volcanic ash at an altitude of 3000 m. There were also some lightning.

– The released lava was relatively viscous and cold (approx. 1075 ° C) and steep cinder cones quickly developed around the active chimneys. An Aa lava flow was formed, the front of which was initially up to 15 m high, but later mostly reached a thickness of up to 6 m.

– The lava flow crossed the LP-212 road, moved at about 700 m per hour in a westerly direction into the area just north of “Monte Rajada” and grazed the center of the village of “El Paraiso”, but already destroyed numerous houses.

– On September 20, the lava flow moved further west along the “Camino el Pastelero” street and destroyed other buildings in the process. However, the flow was getting slower and slower. Since there are numerous cisterns and small canals in the area (for the the banana plantations on the coast), there were also some phreatic explosions and the generation of steam fountains, which also led to false reports about the opening of new vents in this area.

– The seismic activity decreased significantly after the eruption started.

– On 09/20/20201, eruptive activity was concentrated in a vent that had developed at the northern end of the eruptive fissure. The largest cinder cone had developed there as a result of ongoing Strombolian activity.

– On the evening of September 20, new vents opened about 900 m below (northwest) of the main cone (below the LP-212 road) at around 9:30 p.m. At the same time there was an earthquake with a magnitude of 3.8. At least three effusive vents were active. They showed wild spattering and released less viscous lava than before. This led to the evacuation of parts of the village “Tacande”.

– The new vents produced a second lava flow. This moved in a west to south-west direction and was observed on September 21 south of the industrial area “Punto Limpio”, where it came relatively close to the main lava flow or even united with it.

– On the morning of 09/21/2021, the front of the main lava flow was just north of the center of the village “Todoque” and moved very slow. There were 183 houses destroyed and 106 acres of land covered with lava. 6000 people were evacuated.

– The sulfur dioxide emissions were determined on September 21, 2021 at 10,000 tons per day and had increased compared to the previous day (approx. 7,000 tons).

– On the afternoon of 09/21/2021, the tremor, which had decreased slightly after the onset of the eruption, increased significantly and remained high in the evening.

– The GPS stations in the west / southwest of the island recorded a further uplift of the area on September 21, 2021 despite the ongoing eruption. Overall, a maximum lift of 25 centimeters was determined.

– On the evening of September 21, the main vent produced sustained strombolian explosions or generated a lava fountain. The height of the fountain was roughly estimated by observers at 400 – 500 m.

– On the evening of 09/21/2021 at around 8:00 p.m. (local time), a new vent that had developed on the western flank of the main cone was visible. There were individual strombolian explosions there, as well as the release of a viscous lava flow. The front of this flow was moving in a westerly direction.

– On the evening of 09/21/2021, the main stream came closer and closer to the center of “Todoque” and threatened to block the important LP 213 road. This leads down to the coast to the towns of “Puerto Naos” and “La Bombilla”, which have already been evacuated, as well as the Hotel “Sol” near the beach of Puerto Naos, where 500 tourists were.

I have compiled this information from various websites and this blog. Main sources:

This summary is certainly not complete and is not free from errors. But I hope it helps to keep track of the events.


1,323 thoughts on “The Making of La Palma

  1. A new mouth has been opened in the La Palma volcano on the morning of this Monday, specifically, at 10:30 a.m., as reported a few minutes ago by the Canary Islands Volcanology Institute (Involcan).

    Thus, as can be seen in a video shared by the researchers themselves in view of the drone, it is appreciated how the new emission source is already expelling lava as well as gases, adding to the many other cracks that have opened in recent days after the eruption of this volcano in Cumbre Vieja on September 19.

    Cumbre vieja opens two other mouths and confines 3,500 people …
    There are four active mouths, after the appearance of another two outbreaks whose tail …

  2. Looks like Randall called it correctly last night; a new vent opened up on the north side of the cone. I did observe an accurate depression forming, looking dark beneath the streaming gas, late yesterday afternoon. I assume that’s the area where this new vent is.

    • This is another one because it opened at 10:30 this morning .

      ”A new mouth has been opened in the La Palma volcano on the morning of this Monday, specifically, at 10:30 a.m., as reported a few minutes ago by the Canary Islands Volcanology Institute (Involcan).

      Thus, as can be seen in a video shared by the researchers themselves in view of the drone, it is appreciated how the new emission source is already expelling lava as well as gases, adding to the many other cracks that have opened in recent days after the eruption of this volcano in Cumbre Vieja on September 19.”

      Cumbre vieja opens two other mouths and confines 3,500 people …
      There are four active mouths, after the appearance of another two outbreaks whose tail …

      A new mouth has been opened in the La Palma volcano on the morning of this Monday, specifically, at 10:30 a.m., as reported a few minutes ago by the Canary Islands Volcanology Institute (Involcan).

      Thus, as can be seen in a video shared by the researchers themselves in view of the drone, it is appreciated how the new emission source is already expelling lava as well as gases, adding to the many other cracks that have opened in recent days after the eruption of this volcano in Cumbre Vieja on September 19.

      Cumbre vieja opens two other mouths and confines 3,500 people …
      There are four active mouths, after the appearance of another two outbreaks whose tail …

  3. Iceland, Keilir.
    The seismic quakes continue. There is little sign of upward movement – highest is around 5km from the surface (a couple higher) and deepest between 6 and 8km from the surface. The most recent quakes still require verification so we can’t read much into them.
    But they keep happening! It looks like the Krýsuvík-Trölladyngja volcanic system has a few more surprises for us.

    • For a volcano declared dead The one in Geldigadalir is certainly currently smoking pretty heavily.

  4. Shows a white dot on the dark gray ashy slope below the steaming fumaroles. I have been watching this for about 10 mins but no smoke is coming from it and it apparently is not a lava bomb. Any ideas? 1:12 pm La Palma time

    • Spot still there 1 hr later and it occasionally flickers but definitely something is there

  5. I’m very cross at all the whingers here who were moaning about the Geldingadalur volcano being the only and constant source of comment on this blog. You all offended it and now it seems to have taken itself off for a hibernatory sulk for the winter. I’m not pleased. I loved that volcano since the first day and Cumbre Vieja doesn’t compare for 24/7 entertainment especially as it’s a cause of so much sadness and upheaval for so many people which Geldingadalsgos wasn’t. Come back to us all, Geldingadalsgos. I’ll personally throw anyone who complains about you into your very beautiful, elegant and once upon a time effusive crater.

    • Haha Gabrielle, I so agree with the sentiments you have written above. It is starting to smoke a lot today though so I wouldn’t depair just yet.

      • I have no doubt she’ll come back when we’re least expecting it. She’s never disappointed me yet but she took an awful pasting here and that’s got to have stung. I miss her. My evenings have gotten very long without her effervescent and scintillating company and I don’t spring awake in the morning to see what she’s been up to in my sleeping hours. 😀

    • Yes, but if we bait Geldingadalur enough, the wicked witch Katla might decide to have something to say – though that might have some bad side-effects for Icelanders (and elsewhere, perhaps, too).

  6. This video appears to show muliple points of entry to the sea now. Is this correct or my poor eyesight playing me false?

    • What you see in the video is not the cliff, but the lava delta. Since the water is shallow there, the lave already has some impressive thickness.

      • Ah thank you for that answer. I had thought I was missing something but didn’t think of that!

  7. Lava is back in the crater at Fagradalsfjell:

    This explains the heavy smoke today.

    There are also reports that there has been a small lava flow under the crust for several days, around 1 m3 or less compared to 10-15 m3 earlier. That would explain that lava has been visible in the slope towards Natthagi now and then.

      • I am wondering if that visual lava is just actually some of the recent quakes in the Keilir area shaking the crater and showing the residuaul heat in the crater walls. I would so love it to be new lava but for now I shall hold my horses until flowing lava is seen again. As you say though with that heat soufce visible it would explain the extra smoking today. Just as wind stirs up an almost extinct bonfire.

        • You are right to be cautious. The ‘fissure’ is the contracting lava crust, and the incandescence may just be freshly exposed material. Lava can rise through cracks higher than in the crater, so you can get some flow from higher up the crater wall. But this does not look like that. However, there is some activity in the conduit, as seen in the tremor. Even if this is not it, there may be a bit more to come.

    • Doesn’t look like fresh lava, though.

      The big incandescent area looks like it is the hot interior of the walls that has become exposed as a result of a rock fall.
      If you look just below it, you can see the debris on the ground.

    • Thats just resudial heat left
      Rock is a Very good insulator, the entire fagradalshraun is still molten inside. It will take 100 s of years to cool down .. because its so deep too
      Rocks store heat well

    • Taal has stopped subsiding and now the whole systems is inflating, with vog too. Despite being excited and happy, my lust for volcanism has not been satiated just yet.

  8. The new vent near the top of the main cone seems to have stopped emitting anything already.

  9. Wow have I been out of the loop. Vulcano has been raised to yellow alert. Last eruption 1890.

    • M4.2

      02.10.2021 15:32:14 63.932 -22.181 5.7 km 4.2 99.0 1.1 km SSW of Keilir

  10. That would be the first one >4 since March, right? Often they get corrected afterwards. If not this is something.

      • Nice catch. I’m wondering if Fagradalir volcano may have jolted back into activity. I’m watching the tremor rising… (but it could go down again so not worth fretting over).

  11. Some news about Askja, presented yesterday on IMO’s website.

    “Land continues to rise at Askja.

    Scientists are closely monitoring crustal movements near Askja, but land has begun to rise there after a long period where pressure decreased and landslides took place. At a meeting of the Scientific Council, the latest measurements were reviewed, which indicate that land continues to rise on the western edge of Lake Öskjuvatn and at a fairly even speed.

    The seismic activity at Askja in recent weeks is higher than is generally measured in the area. The earthquakes are in known areas and of a similar magnitude as usual, but the frequency of earthquakes has increased and remained constant since the beginning of August.

    The Meteorological Office is working to increase monitoring at Askja by installing new measuring equipment, but conditions there are very difficult due to the weather and heavy snow after the storm earlier this week. An additional GPS meter has already been installed.

    The Science Council will meet again in two weeks’ time on the activity at Askja and will then set up possible scenarios for the development of matters at the volcano.”

    I wonder if Askja will be next instead of Grims. Grimsvötn doesn’t really gets on with it.
    Csm is accumulating at steady pace, but the steeper run up as seen in the previous eruptions, lacks.

    • The twitter above is a trace of the earthquakes on La Palma Island from early Sept until Oct 1st.

  12. Increase of the explosive activity in the last few hours #lapalmaeruption #erupcionlapalma #lapalma

  13. La Palma

    From Involcan
    “Informative update: From 3:00 p.m. (Canary time) today, October 2, 2021, the Canary Islands Seismic Network on the island of La Palma registers a rebound in the amplitude of the volcanic tremor related to an increase in activity Strombolian-type explosive.”


  14. Event Date Time UTC Local Time (*) Latitude Longitude Depth (km) Magnitude Type Mag. Int. Max. Location es2021ticru 02/10/2021 18:12:39 19:12:39 28.5648 -17.8320 11 3.6 mbLg SW VILLA DE MAZO.ILP

    3.6 mag quake on La Palm 69 minutes ago

  15. La Palma:
    The online magazine “El Time” reports today that the lava flow has destroyed the main water line that leads south and is used for watering the numerous banana plantations between Las Hoyas and El Remo. The line was destroyed although it runs underground. The line had withstood the lava flow for several days. Now a drama is looming for the banana farmers. There are plans to set up a seawater desalination plant, but this would take at least 4 weeks.By then, the banana plants are likely to have been damaged or have died completely.

    Link to the article in El Time:

  16. La Palma:
    The situation is getting more worse again. The tremor is rising and a new vent has opened to the north next to the chimney with the strongest slag throw, which promotes a well-feeded lava flow.

    • last 5 mins (just before 9pm CET) the amount of lava coming out is absolutely insane. The channels are overrun and there is new flow coming down.

      • saw those chunks being pushed into the lava tube, wish my house had such good plumbing…

Leave a Reply

Your email address will not be published. Required fields are marked *