Sometimes, a volcano is nothing more than a hole in the ground. There are beautiful, though deeply eroded, cone volcanoes around Taal: Mount Banahaw, Mount Mariveles, Mount Natib – especially the latter also has a large caldera. But Taal is not like that. It is a lake, with only a little cone sticking out. The BBC thought it was worthy of the label ‘small’. But that is deceptive. The lake itself is the sign. It is the scar of a series of large eruptions which collapsed the land. The volcano never really recovered. The old crater rim is still visible on the photo above (Christina H Koch, on board the ISS). Its immensity is clear when compared to other, nearby volcanoes. Only Natib comes close, with its 40 km2 caldera. Taal caldera (actually a double caldera) is almost 500 km2, though.
This may sound worrying. With a caldera this size, we could be in VEI 8 territory. (We are not in Toba land though. That caldera is 5 times larger.) But it isn’t like that and it has never done a VEI 8, although it may have reached VEI 7. Taal’s enormous caldera formed in collapse rather than expulsion. The content went somewhere, but only a small fraction ended up in the sky. Probably. Still, this is not a volcano to take lightly, not even by the BBC. Luckily, stupendous calderas are invariably old. It happened a long time ago.
But not quite. One (and only one) of the major eruptions has been dated. It happened 5380 +- 170 BP (radiocarbon years – they differ slightly from calendar years). That is not quite yesterday, but almost yesteryear. This volcano is perhaps not as safe as we would like.
It is always worth looking at a map. A map beats even the best TomTom satnav, if you want to understand the lie of the land. The land of the Philippines is clearly rather sea-rich. The country has over 7000 islands! That doesn’t happen by accident: this is a region with some active geology. After all, all deep sea islands are volcanic in origin, and all mountain chains indicate careless driving by plates. Manila, the capital, is on the northernmost of the major islands, between a bay and a funny looking lake, called Laguna Bay. The lake south of Manila is Taal. It becomes worrying when you realize that the middle of the three fingers of Laguna Bay is a 200 km2 caldera. Taal caldera is 50 km south. Pinatubo is 100 km northwest of Manila and Mayon is 250 km southeast. Was this really the wisest place to locate the capital?
(In addition, the West Valley Fault runs through Manila towards Taal: this fault may be primed for an M7 earthquake. I just mention it.)
The sea also hides secrets. The Philippines are squeezed between two plates. On the west side is the Sunda plate, bordered by the Manila Trench, and on the east is the Philippine sea plate, bordered by the Philippine trench the 3rd deepest trench in the world. Both are subducting underneath the Philippines. The Manila trench feeds Pinatubo, and the Philippine trench feeds Mayon. The Philippines are sheared by the conflicting motions: a left-lateral fault zone runs along the entire country. In Luzon, the island with Manilla, the subduction is dominated by the Manila Trench. The Philippine plate has only recently started subducting this far north, and it is still shallow. Further south its subduction is better developed.
Taal is part of the Southwest Luzon volcanic field, fed by the Manila trench subduction. It has over 200 volcanoes, some younger than 10,000 years. There are so many because part of the field is monogenetic. The most famous of these are the seven lakes of San Pablo, all seven maars. Mount Banahaw, 50 km southeast of Manilla, has erupted twice since 1600; its 1730 eruption ruptured its crater lake, causing a flood that destroyed a town. It is not all Taal here.
This volcanic field is related to The Macolod Corridor is a 40 km wide rift zone, running southwest to northeast. It is about 1 million years old. The volcanoes it has produced are basaltic. Taal is affected by this rift zone: the southwestern part of the lake shows a graben with this alignment.
But this is not the only rift zone in the area. A second one runs from Taal northwards, in the direction north-northwest to south-southeast. This is an extensional zone, and is less dominant. Very little has been published on it – nowadays it is most obvious from the alignment of the roads in this location! It has been suggested that Taal forms the intersection of these two rift zones.
The volcanoes are caused by a combination of the rift and the subduction zone below. The magma is formed in the subduction layer: it is oceanic basalt, with the addition of melt from several kilometers of sediment that accumulated on this oceanic floor. There is no evidence of crustal melt, suggesting that the rift makes it easier for the magma to percolate up but does not generate melt itself.
A magma chamber exists 1 km below Volcano Island. A second layer of magma is found at 7-10 km depth. Eruptions are caused by magma migrating from the deeper to the shallow reservoir, although not all such intrusions cause eruptions.
34 eruptions have been reported from Taal since 1572. This included 2020. But only 25 of those are confirmed. The 1754 event may have been the largest of the historical eruptions, and may have reached VEI 5. A few events have been plinian, but the most common type are phreato-magmatic explosions. This after all is a thoroughly wetted area! All eruptions have been on or next to Volcano Island. There are some eye witness reports that sound like fissure eruptions. And the eruption have come from all four corners of the island: the central crater does not have a monopoly on eruptions. From 1572 to 1645 and again from 1749 to 1911, eruptions came from the central crater. Between 1707 and 1731 the eruptions came from various craters on the outer flanks (the 1716 and 1731 eruptions occurred in the lake), as did the 1965 and 1977 events. The 2020 explosion was again from the main crater. Eruptions tend not to move from inside to outside the crater, or the other way, within the same eruption. But this may be a rule that is there to be broken!
The change from one mode to the other coincides with a dormant phase lasting typically 50 years (the quiet time between 1731 and 1749 was only 18 years). During a mode, eruptions are typically 5 to 15 years apart. But after the large 1754 eruption, there was half a century of quiescence and another 50 years of minor eruptions only. The large 1911 explosion was also followed by 50 years of solitude.
The largest of the historical eruptions was that from 1754. The eruptions started on May 15th and finally ceased December 2. Within this period were a few quieter phases, lasting a few weeks to a month. October was eruption-free. But there were powerful explosions throughout the 7 months. The sudden explosion on May 15th send a plume estimated at 40 km height, and caused a pyroclastic flow that destroyed a village 8 km northwest. After this, activity became just ‘furious’ with explosions ejecting stones and mud, volcanic lightning and tephra – an ordinary eruption. An intense explosion on Nov 15 causes boulders to fall into the lake, causing huge waves. November 28 was bigger, and now the volcanic tsunami was large enough to cause flooding. The biggest explosion was a day later, with complete darkness and houses collapsing under the weight of the ash.
The pyroclastic flows or base surges reached most places on the lake. To the west, the maximum distance was thought to be 16 km. However, mapping of the deposits suggests it came even further in-land and reached both the Tagaytay ridge and the coast. The dashed line on the plot shows the suspected extent. The ash reached 2 meters thick on Volcano Island and 1.1 meters on the shore. Even the region of Lemery received 10 cm.
A few days later a typhoon came and caused numerous landslides. It never rains but pours. The tephra reached Manila.
This was the worst eruption in recorded history. It is now classified as VEI 5. But there were other damaging eruptions. As an example, the 1911 eruption caused tsunami waves 3 meters high on the shores which did significant damage, and the surge reached 9 kilometers distance.
Below is the list of known eruptions. Ones in italics are not confirmed. Red dates correspond to eruptions from the main crater. The others are from the flanks and lake. The groupings into A-D are from Delos Reyes et al., Earth-Science Reviews 177, 565 (2018). Much of the information in this post is taken from that paper. (“A synthesis and review of historical eruptions at Taal Volcano, SouthernLuzon, Philippine” is the promising title – sadly the paper is locked away and largely unread behind a paywall). On their scheme, 2020 would count as the start of sequence ‘E’.
Note added: Mike K. has pointed out that the paper is no longer paywalled and is now freely accessible at https://www.sciencedirect.com/science/article/pii/S0012825216304068
Most eruptions affected Volcano island only and provided a spectacle to the on-looker but no danger. Some affected the lake, and some went beyond and did endanger people living around the lake. Taal has been responsible for around 1500 deaths in recorded history.
Surprisingly little is known about the older eruptions of Taal, and the formation of its caldera. That is at least in part due to the climate. Tephra layers are rapidly removed or decomposed by the tropical rains. But we know there have been eruptions much larger than those of the past 500 years. The existence of Lake Taal is telling. The prehistory is the tale of Lake Taal.
There are four main debris layers around Taal caldera, with scoria, ignimbrites and pyroclastic flows. The youngest pyroclastic flow is the only one that has been dated. It formed between 5380 and 6830 BP (radiocarbon dates). The volume of this eruption is estimated at 50 km3, or a large VEI 6. This flow is young enough that Taal is likely to have looked similar to what it is now. It was a hot, gas-rich, concentrated pyroclastic flow. The recent flow is centred on the lake, but not on Volcano Island. There may have been a different eruption center at that time, which was destroyed in the eruption.
The other flows are older but not well dated. The Taal Tuff is not older than 300,000 years, ignimbrites are 100,000 to 500,000 years old and pyroclastic flows are less than 140,000 years. This has been interpreted as that all four are less than 140,000 years old. A sea core has shown 16 large explosive events over the past half a million years coming from the Macolod Corridor, similar to and including the one 6000 years ago. Over this time, the eruptions have slowly become focussed on Taal. Taal Lake consists of at least two separate calderas, already showing more than one major event.
The first three of the major Taal eruptions produced andesitic to dacitic ignimbrites. The most recent one ejected basaltic andesite. This may represent evolution of the volcano. An optimistic interpretation is that it is not as explosive nowadays as it was 100,000 years ago!
An important question is whether these events reached over the 600-meter high Tagaytay ridge and made their way towards Manila. Did they reach? The Taal ignimbrite plain extend northwest to Manila Bay, so it did get close. Manila is build on several ignimbrite layers, up to 10 meters thick. Their origin is not known but Taal is one of the two main suspects.
This is the background to the recent eruption from Taal volcano. On the scale of historic eruptions, this event was not insignificant but not major. It may however herald a time of more frequent eruptions, now that the focus has shifted back to Main Crater. Main Crater has been quiet for more than century. This is unlikely to be repeated.
Taal has proven plenty of hazards: tsunamis, pyroclastic deposits, thick ash, sulfuric gasses. It is clear that Volcano Island is not suitable for habitation. After every cleansing, over time people move back to the island. This is a recipe for disaster. The shores are also dangerous but here the danger is perhaps more manageable. All houses and other structures should be able to handle tsunamis of at least 3 meters high (5 meters preferred) and able to carry a weight of 1 meter of ash. A tsunami warning system is needed. An evacuation plan should allow for rapid movement of people, including every location between Lake Taal and the coast: this may require new escape roads.
And from now on, Taal tourism may need to adopt a ‘look but don’t touch’ approach. No more swimming in its acid lake. It would be wise to remember White Island.
Albert, January 2020