Rio Helmi tries to make sense of the ongoing marathon of Agung’s eruption cycle – What is it going to do next? When will it end? When will the conspiracy theories stop? Can  these questions be answered?

Since the end of  December 2018 Mt Agung has undergone a series of intermittent eruptions averaging one every 5.6 days. I get this figure from Jackie Pomeroy, creator of Mt Agung Daily Report, as she shows me her meticulous spread sheets on her laptop. You learn to expect this kind of thing from a career economist with a passion for this volcano. Though we both are keen on accurate data, I’m too lazy to to keep meticulous logs. This is probably because my father had a Phd in economics and he would always tell me I would end badly, what with all my sloppy habits.

That notwithstanding,  I’m pretty sure that the four of us who kept an eruption watch all night long in the chill wind of the night of the 29/30th of May on a plateau above Tulamben are pathologically obsessed. That would be Jackie Pomeroy, amateur (but serious) videographer Jackie Zwahlen who I meet for the first time, Geodesy graduate Elang Erlangga and myself. The last time I saw this bright young man, Elang, was when we visited the volcano observation post in Rendang together.

Hanging out at the Rendang volcano observation post with PVMBG volcanologist Ibu Estu and ITB Geodesic graduate Elang Erlangga 

By morning it is the 6th night after the last eruption – with plenty of seismic activity since then –  we end up disappointed. What happened to the 5.6 average? What happened with all that seismic activity we were seeing? And so we are reminded that it is impossible to predict the timing of an eruption ahead of time. Well we saw a lot of shooting stars and a grand view of the Milky Way. I am the least obsessed of us all (what, did you just read least ‘dedicated’?) so I abandon them to the barren landscape – but not before eating a good part of Elang’s corn and banana rations, polishing off whatever cookies Jackie P bought and raiding their collective coffee.

I head off over the northern pass between Mt Agung and Mt Abang. This road changes everytime I ride it. One day it is covered in sand from where the dry river bed comes down from Cegi, and another day it’s a clear except in front of the school in Daya. The sharp hairpin turns as you come up the steep slope to the final pass into Pura Gae are particularly tricky: one day you can ride it and it’s in good condition, the next there will be great big chunks of asphalt missing and the shoulders turned into 30 cm ruts. When you’re negotiating those narrow switchbacks and a truck is coming down at a pace it can be heart stopping – as it happened on this ride. Which is why I never take this route at night.

No eruption shot that night – but a grand view of Agung and the Milky Way with Jupiter thrown in

This is also the route that goes closest to some of the highest villages in Ban. During the last couple of eruptions they didn’t evacuate but gathered at mustering points to watch and assess the fireworks as incandescent projectiles landed close to their hamlets. Personally I think this is a healthy development. They aren’t panicking, but gathering in a place a bit out of the danger zone to wait and see before they make their next move. There have been a lot of spectacular eruptions, but in reality they were of fairly low explosivity and posed little danger – if one is outside the 4km exclusion zone. That of course doesn’t stop the foreign press from making a huge fuss.

Above and Below: Villagers in the upper Ban hamlets like Pucang, Pengalusan, Daya and so forth are now tending to  wait and check on the volcanic eruptions before panicking and fleeing

When I get closer to the main Rendang to Suter artery, I can’t make up my mind what to do. But the bike decides to take a left turn in Pempatan, and against my bleary-eyed will I end up at the Volcano Observation Post in Rendang. As I make my way down the stairs I run into the mentally ‘diffable’ girl who now hangs out regularly at the post sweeping the steps. Some of the staff have been patiently teaching her to read – I think the atmosphere of kindness has changed her over the last couple of years.

Once there, Indonesia’s Center for Volcanology and Geological Hazard Mitigation’s (CVGHM or PVMBG in Indonesian) Head of Eastern Indonesia Hazard Mitigation, Devy Kamil Syahbana, reminds me  there are many different factors that need to be assessed in order to understand what’s going on inside any volcano.

“Deformation,  seismology,  satellite images are not the key to predicting. You can only determine that in that instance that the mountain is active, and to what extent it is. If you have enough data and knowledge then you can forecast probability – but not even the best volcanologist in the world can predict exactly the time of an eruption. It is very important that people who have not studied volcanology properly do not make interpretations from raw data, much less predictions because it is complex. Just to be able to read volcanological seismographs properly takes at least four years of study!”

He tells me this right after I witnessed an interesting scene in the observation room. I was looking at the seismogram from that morning and asked Pak Devy what a certain blip was. He said “Itu hembusan (that’s degassing)”. One of the seismologists, present said “Oh actually that’s a local tectonic…”.  At which point Pak Devy enlarged that section, pulled up the concurrent seismograms from Batur and other recorders around the mountain and using the timeline and other parameters, in 5 minutes proved why it actually was a degassing. I thought to myself: these guys are trained seismologists and they are still learning about volcanological seismography – I’m way out of my league!!! Where does that put us Facebook amateurs, thinking we understand the data that CVGHM/PVMBG is generous enough to share online, or even the 6 hourly round ups of the activity of every single active volcano in Indonesia that they publish? And some of us even go so far as trying to predict the time of eruptions etc?

In the last week the live public online link to some of the seismographic recorders on the mountain have had some glitches, or have possibly even been hacked. A hue and cry went up in some FB quarters that “PVMBG is hiding/manipulating data”.  One has to ask oneself if these conspiracy theorists realize that PVMBG to date is possibly the only geological/mitigation service in the world that provides online links to this kind of live data, the only one that provides constant the 6 hourly round-ups. I know for a fact that many international volcanologists have found this remarkable and even commented on it directly to PVMBG representatives who went to Italy last year to receive an international award for their work. As Pak Devy, incredulous, said to me: “What benefit would hiding or manipulating data bring us?”

I ask the burning amateur question: “With current intervals and parameters is there a possibility of something bigger happening?”

Pak Devy makes himself comfortable, ready to spend a little time patiently explaining: “It’s very unlikely (under these circumstances) we’ll see something different from the previous eruptions. We still have the same signals – in terms of seismicity, deformation, hotspots from the satellite images – so we are not expecting a larger eruption. This is normal because if you are expecting a huge eruption you would see some significant data. The seismicity would show a significant change, indicating magma on the move. While I can’t say exactly how many chambers there are in this volcano, we know that there are two systems – the deeper system and the  shallower system. Right now with the type of activity we are seeing happening in the shallower system you just cannot expect anything dramatically bigger.”

After a brief pause, he continues: “If we had a significant change in the supply chambers of the deeper system then we would have a dramatic change in seismicity and deformation like we had in September 2017. But the condition now is (minor) fluctuation. You have small scale inflation, then deflation accompanying (small scale) eruption. That’s the rhythm of Gunung Agung at the moment. So it’s unlikely to have a sudden large scale eruption with this kind of activity.”

I can’t resist the hackneyed query: “What would make it shut down, what would be the indicator of the end of this cycle?”. At this point I’m very grateful for his patience – I know it’s part of the practice of the fasting during Ramadan but it’s also part of his personality. He’s a tad red eyed, he’s been up most of the night monitoring the volcano and writing a new 8000 word article on Agung. In any case he repositions himself comfortably again in his arm chair in the new second floor sitting room adjunct to the observation room, preparing himself to give a simple but comprehensive answer to this pesky old blogger who clearly doesn’t have any real scientific capacity.

The words come out deliberately but confidently – he lives and breathes this stuff with a spirit of inquiry and precision. In brief: “If we go back to before this eruptive cycle, this volcano only had tens of volcanic earthquakes per year. At this moment we’re still recording volcanic earthquakes and degassing activity that indicates the volcano is still active and the movement of magma is detected. Right now we can’t say this is going to stop. The fluctuations aren’t showing any decrease in their patterns nor increase. So all we can say is we don’t know when it will stop. There are no signals that this pattern is changing for now. But I think if we see the frequency of eruptions decreasing over time, and also the explosivity decreasing – those would be the signs of the volcanic eruptions coming to a stop. But volcanic activity at a very low level would continue, similar to before this cycle. For example one volcanic earthquake per month. But if we look at Mt Agung’s eruption cycles it’s always long, it’s never short. If you look at Agung’s ‘63-64 cycle, actually even in ’65 there was still activity, though minor.”

As we discuss the length of these cycles, my mind flashes back to a conversation I had with young Elang several days before. We were talking about the 1843 cycle, which by some accounts supposedly lasted 4 years, yet there is so little in the manner of records – unlike the 1710-1711 eruption cycle of which there are very clear and dramatic written records in lontar form. Possibly the most famous quote from these, “Ravines became hills, hills became ravines” describes the effect of pyrochlastic flows, lava and lahar on the northeastern slopes. Traveling through these areas many months ago with mitigation expert Eko Teguh Paripurno, whose geologically trained eye sees it all in a different persepective, I came to realize that these lontars were actually not exaggerating.

Elang suggested, based on a carbon dating report he had seen on what were the supposed remains of the 1843 eruptions, that these were greater paroxysms than 1711. The mystery is, how come there are no real written records of this cycle? Looking back into history this was a very turbulent time in the Puri (palace) of Karangasem. There was a war with the Dutch colonizers to the north in Buleleng, conflict with their Buleleng cousins, and even a heinous fratricide within the palace itself. I took the liberty to ask poet/performer/activist Cok Sawitri – of the branch of the royal family in Sidemen – if she had heard of any records in Sidemen. She did some digging around but came up with nothing substantial.

Pak Devy was laconic when I asked him about any similarities between our current situation and that of 1843. “I can’t comment on that because I have no data to compare. There is very little information.” He pauses, then: “But in 1963, if we look at the manuscripts written by many authors, if you compare to the 2017 cycle up til now, in terms of sequence it’s similar. They both started with very rapid ramp up of quakes at the beginning of the crisis, then we’ve seen ash erupting out, then we’ve had incandescent lava flow to the crater, and also Strombolian activity, then ash eruptions again. So that’s the similarity.”

Agung erupting in 1963 (photo AP)

“But in terms of the times – intervals, durations they are (quite) different. In 1963 the process was faster. It indicates that there was a very large volume of magma focused towards an eruption which continued to build up from February when there were a series of eruptions til March 1963 when there was the paroxysmic eruption. But 2017 the volcanic earthquakes actually decreased during October, and when it erupted on the 21stof November the (seismic) signal was very weak. Afterwards we had a series of larger eruptions. In the beginning we were worried after the 21st, that on the 25ththere was going to be the paroxysm but then at the end of November it stopped. So what we saw was that some of the magma had been released in the form of 25 million cubic meters of lava inside the crater, and more in the form of ash ejected out. From our instruments (seismic, GPS, etc) we calculated a total of 50-60 million cubic meters of magma was accumulated and then released. But since it has been released the system hasn’t yet come back to equilibrium.

So there are still some small explosions, but we haven’t seen any significant changes in eruptions since January 2019. On the 19thof January 2018 there was a significant Strombolian eruption but without sound. But now we have sound, the reason being that the lava accummulated in the crater is getting colder, so it is hardening. When the gas or the accumulation of magma erupts to the surface it has to break  through that crust which creates the explosive sound. But the characteristic of volcanic activity hasn’t changed, it’s just the surface condition. Although we can see that on average at this stage the volcano erupts every six days we still can’t predict. Volcanic activity isn’t linear, it’s stochaistic. We monitor it but we can’t predict exactly what it will do or when. It’s like a doctor who can diagnose a patient’s grave illness but can’t predict when the patient will die or recover.”

As we wind up the conversation, referring to some of the scaremongering and speculation on social media, he emphasizes that a lot of people now – “new volcanologists” he says drily –  have gained some new knowledge which they feel enables them to predict and interpret the volcano in their own ways. Pak Devy was very clear: “I don’t do that, I could make up my own narrative but that would be wrong. My job is to inform the public about what has happened and what is happening with the volcano. We hope that people don’t want to become a ‘doctor’ in a few months. I really hope that people who don’t know too much about what’s going on with the volcano don’t talk too much because all they are doing is creating confusion. Volcanologists don’t predict, they don’t even use the word predict – they use the term forecast which denotes probability. Some people think we are hiding information or predictions – what for? In 2017 we presented the worst case scenario and we got pressured for that. We don’t need to hide information. For those who have some conspiracy theories going, we can’t erase anything on the seismograms – and they are all public. For the record when I present our work in the US or Japan, the volcanologists there say “Wow, your people are crazy enough to publish six hourly updates not only on one but on all your active volcanos!”. No other volcanology institute in the world does what we do! But with all this information, does it really make people smarter?”

My head full of reflection on his words, I finally make my way out the door. Sleepless and exhausted I take the shortcut through the backroads of Bangli to Ubud. In front of me on a steep section coming into Tampaksiring a small pickup has come unstuck and blocks the road. I barely make it past through the deep ruts. When I get home I stumble into my bed and pass out, my first real sleep in about 36 hours. I wake up after two hours to find that one of my dogs has been poisoned, and by the time I can get him loaded on to my pick up to take to the vet he died. My vet offers to cremate him for me, so I take him over to Pejeng with a heavy heart. What if I hadn’t gone on that wild eruption goose chase? Would I have saved him?

The next day, refreshed after a reasonable sleep, I get cracking on the chores that have been neglected. I head off to meet a friend of a friend for lunch, and sure enough at 11:45 the first messages come in – the volcano has erupted, 15 minutes ater Jackie left the look out post. I stop by the side of the road in Ubud and post it on Facebook. Not long after I post a certain facebook page that shall not be named posts images of the volcano with the usual incandescent projectiles and ash spilling out and claims that it is a pyroclastic flow  (awan panas in Indonesian, or hot cloud, one of the deadliest phenomena volcanoes can emanate) – the admin apparently can’t tell the difference.  It was enough to force  Pak Devy to release a statement correcting the mistake.

It’s been a strange day and a half…


Below: PVMBG’s eruption notice on the 31st of May 2019



lead image is of Mt Agung’s crater seen from Tulamben as the volcano degasses on the 30th May 2019  morning.


*All text and images copyright Rio Helmi except where stated otherwise


Please feel free to scroll down and leave your comments below