PubTalk 8/2018 — What on Earth is going on at Kilauea Volcano?

Video Transcript
Download Video
Right-click and save to download

Detailed Description

Title: What on Earth is going on at Kilauea Volcano?

  • First significant summit explosions in nearly a century
  • Largest summit collapse volume since at least 1800
  • Voluminous fissure eruptions feeding channelized lava flow
  • Unparalleled new opportunities for understanding the volcanic system


Date Taken:

Length: 01:18:45

Location Taken: Menlo Park, CA, US


[Silence] This video is a one-hour-and-15-minute presentation of the USGS evening public lecture series titled, What On Earth is Going On at Kilauea Volcano? The presentation is being hosted by the USGS Menlo Park facility. The host welcomes the audience and introduces the speaker, Kyle Anderson, who is a USGS research geophysicist. As Kyle is giving his presentation, he is continually pointing to and referring to slides presented on the screen. The slides are a mixture of charts, graphs, and photos. At the end of the presentation, there is a question-and-answer session with members of the audience. [Silence] [multiple inaudible background conversations]

- Good evening, and welcome. I’m delighted to see such a full house tonight. You made us work hard and get extra chairs out. That’s okay. Before – my name is Leslie Gordon, and welcome to the U.S. Geological Survey here in Menlo Park. Before I introduce tonight’s speaker, I do want to tell you about next month’s speaker because I want you to come back. [laughter] Next month, Tom Brocher will be – a geophysicist – will be speaking about the 150th anniversary of the damaging 1868 Hayward earthquake. Now, a lot of people never heard of the 1868 Hayward earthquake. But I can assure you it was the last time there was a big quake on the Hayward Fault. And there’s nothing to say that it won’t happen again someday soon. It happens – and October 21st, it happens to be the 150th anniversary. And we’re using this as an opportunity to remind people to be prepared. So do join us again next month on September 27th to hear about the 150th anniversary of the 1868 Hayward earthquake. And, by the way, at the time, 1868 – that was before the big 1906 earthquake, right? [laughter] So do you think they called it the Hayward earthquake then in 1868? No. There were hardly any people living in Hayward. They called it the San Francisco earthquake. [laughter] Until, of course, the Great San Francisco earthquake happened in 1906. Tonight’s speaker – it is my great pleasure to introduce Kyle Anderson. Kyle is a research geophysicist here in Menlo Park with the California Volcano Observatory. He’s been spending quite a few months recently in Hawaii doing double duty at the Hawaiian Volcano Observatory. Ironically, I have to tell you, although both Kyle and I work here in Menlo Park, I really didn’t get to know him until we both showed up in Hawaii working on the volcano response a few months ago. So it is really a pleasure. Kyle did his undergraduate work at Whitman College in Walla Walla, Washington. He then went to Stanford University to get his master’s and Ph.D. degrees, both in geophysics. After graduating Stanford in 2012, he spent more than three years in Hawaii at the USGS Hawaiian Volcano Observatory as a Mendenhall postdoctoral fellow. So he really knows his stuff. After that, he secured a permanent position here in Menlo Park with the USGS. And his work is primarily with Mount St. Helens and the Kilauea volcano. So his work focuses on building mathematical models of volcanic systems that we can use to relate different kinds of data sets to one another. Although I promise he’s not going to be talking too much about math and data sets tonight. He has pretty pictures. [laughter] But all to understand better these volcanic processes. So, as I mentioned before, Kyle flew to Hawaii in early May and spent most of the next three months on-site responding to the Kilauea volcano eruption. So please join me in welcoming Kyle Anderson. [Applause]

- Thanks, Leslie. Can everyone hear me? Very well. No? [chuckles] I heard one "no." Thank you all for coming. The talk I’m going to give tonight is one of the easiest I’ve ever given and also one of the most difficult. As Leslie said, there’s a lot of fantastic photographs, and it was really an incredible event. But it’s also a lot to fit into a 40- to 45-minute talk. So I will do my best, and I hope you enjoy it. I want to start with a warning and an introduction. So the 2018 eruption at Kilauea was really unprecedented in modern times. And we’re still in the really early stages of trying to understand what happened and interpret the data sets. And as a result, my warning is that everything I’m going to tell you tonight is preliminary, okay? [laughter] So interpretations may change as time goes on, but this is our thinking at the present moment. I’m giving the talk tonight, but really the response to the eruption was a group effort from the entire U.S. Geographical Survey’s Volcano Hazards Program. Of course, the lion’s share of the work fell on the staff at the Hawaiian Volcano Observatory, but many of us went out there to assist. This is a partial list of names in the background, and there are many more on the mainland who never went out who did a tremendous amount of work here as well. So this is really a group effort, and I apologize that I don’t have time to give full credit to individuals throughout the course of the talk. So one thing – one message I’d like you to take home with you – important point of this talk is that long-term volcano monitoring really pays important dividends. This is a photograph of Thomas Jaggar in his office in 1925. Jaggar was the founder of the Hawaiian Volcano Observatory – HVO. He founded it in 1912 under the belief that careful long-term scientific observation and investigation are key to understanding and forecasting hazardous geological processes. And that’s very much as true now as it was back then. And what I hope to show you today is that it’s that long-term record of observations – more than 100 years at Kilauea – that allowed HVO to respond as effectively as it did to this eruption crisis and to preserve life. Before I delve into Hawaii, I want to remind everyone that we are in California, and California is not just earthquake country. We have a lot of volcanoes here as well. The photograph on the right is the eruption of Lassen Peak in 1915 – photograph from the town of Red Bluff. The map here shows a number of volcanoes from Medicine Lake in the top all the way down to Salton Buttes in the south. So keep in the back of your mind that we have a lot of volcanoes here in California as well. Kilauea volcano – most of you know where it is, of course. It’s on the Big Island of Hawaii in the Pacific Ocean. The island of Hawaii has a number of volcanoes, and Kilauea is far from the most topographically impressive. That honor goes to Mauna Kea or perhaps Mauna Loa. Kilauea is sort of perched on the side down here. But what Kilauea really has is activity. It is certainly recently the most active volcano. In fact, it’s one of the most active in the world. And it’s also one of the best-monitored. As I said, HVO has been monitoring the volcano for more than a century now. So we’re really getting a good idea of the way it works. Having said that, however, there are really some fundamental questions that remain. And I’ve listed some of the ones that are most relevant to the 2018 activity on the right-hand side. Some of these questions are, what causes dangerous explosive activity? Where is magma stored, and how much is stored there? How is the summit connected with the rift zones? And what is the interplay between the tectonic and magmatic forces? So I’m talking about Kilauea. Let’s zoom in on that. And I’ll give a quick summary so you know the places I’m talking about and some of the fundamental processes that drive eruptions at the volcano. So the summit is on the left-hand side of the figure. And that’s connected to what we call the East Rift Zone. A rift zone on a volcano is a zone of weakness, you can think of, in which magma can flow relatively easily or intrude. So let’s zoom in on the summit. I’ll start by showing you where the Hawaiian Volcano Observatory is. That’s HVO perched right on the rim of the caldera – really a fantastic location for viewing activity. Sometimes maybe a little too close. [laughter] This is Kilauea Caldera – the broader structure. Within that is Halema‘uma‘u Crater. And within the crater is a lava lake, and that lava lake has been active from – well, starting in about 2008 until about 2018. So the broader caldera – that has filled and collapsed repeatedly over geological history. Halema‘uma‘u Crater – that’s the intermediate-size feature there. That last changed dramatically in 1924 when it doubled in size. And then the lava lake, really over the last 10 years. Now, that’s a photograph of the lava lake. It really is an active lake of lava that’s a couple hundred meter across, and that rises and falls with pressure on the magma reservoir, which I’ll talk about. So that’s the summit. Now let’s talk about the rift zone. So what happens at Kilauea is magma rises up beneath the summit, and it passes through some storage reservoirs, and then it goes down the East Rift Zone. So ever since 1983, actually, magma has migrated down the East Rift Zone to the Pu‘u‘O‘o vent. And there are a couple of photographs of it there on the top. It’s been erupting semi-continuously for, what, 35 years. Usually what happens is, magma migrates down, and then it reaches the surface near Pu‘u‘O‘o and forms lava flows which reach the ocean, and you’ve all seen photographs of this over the last 30-plus years. That’s sort of normal activity – or what we’ve gotten used to as being normal. Now, the Lower East Rift Zone is further down – down towards the coast on the far right-hand side. There have not been lava flows in the Lower East Rift since 1960. So historically, there were eruptions down there in 1955 and 1960. These lasted on the order of one to three  months, but nothing since that time. The photograph on the top there is some fissures in 1960. So let’s actually – let’s imagine we draw a cross-section sort of along the path of magma there, so those red lines. So let’s draw a very simplified cross-section. This is not at all to scale. [laughter] This is just to give a conceptual model of the way it works. So think of a – of a – maybe a elementary school drawing here. So the idea is magma rises up to supply Kilauea from the mantle – that’s on the bottom there, and it passes through some storage reservoirs at the summit. And I’ve drawn one here, but it’s, of course, a little bit more complicated than that. That’s connected directly to that lava lake that I showed you a photo of. And magma migrates down the East Rift Zone. That’s here. And ever since – as I said, since 1983, that magma has been erupting near Pu‘u‘O‘o. So not to give the end of the story away here, but things changed in 2018. They changed about three months ago. And instead of erupting at Pu‘u‘O‘o, magma intruded further down the rift and erupted in Leilani Estates. So that gives you a general idea of the way magma is supplying the fissures during the 2018 activity. Oh, and I should say something about the scale. It’s really quite large. So this magma reservoir is about a kilometer and a half deep. That’s about 20 kilometers from the summit down to Pu‘u‘O‘o and another 20 down to Leilani Estates. So it’s a long ways. Okay, so let’s set the stage. Now, what happened in 2018, or what has happened so far in 2018? So magma started backing up in the system this spring. So imagine – and this is a photograph of Pu‘u‘O‘o, so that’s that vent down in the rift zone. And imagine we measure the distance between two points on the opposite side of that cone. You can imagine that, if magma starts getting backed in that – in the reservoir beneath that vent, or in the rift zone there, that distance is going to increase. So you’re storing magma. You’re inflating things. That distance increases. Well, that’s exactly what happened. So this is a time series. This shows from January 1st to April 29th of this year. It shows the distance between two points about a kilometer apart on each side of that vent. Each of these blue dots is a measurement. That scale in 20 centimeters. You can see that, starting in about mid- March, that distance started to increase. And then it really accelerated in April. So that’s a – that’s an increase in length of about 20 centimeters. And that’s a lot when you’re thinking of solid rock. So as a result, HVO issued a warning that a new vent could form either at the Pu‘u‘O‘o cone or along adjacent areas of the East Rift Zone. And I’ll say ERZ throughout this talk. That means East Rift Zone. So at 2:15 p.m. on April 30th, something happened. And what happened is magma rapidly exited the system right there under Pu‘u‘O‘o, and it intruded down-rift. So the photograph on the lower right is Pu‘u‘O‘o after the crater floor collapsed. And that happened as magma evacuated to move further down the rift. This is that same time series but extended a little bit further. So you can see the inflation as magma backed up and then rapidly drained away. So where did it go? This is a map of the East Rift Zone. This is Pu‘u‘O‘o right here. And what HVO was able to detect is increasing numbers of earthquakes migrating down the rift. And they also detected ground deformation, suggesting that magma was migrating down the East Rift towards Leilani Estates. And you probably heard about Leilani Estates on the news. This is a subdivision with about 1,500 people, I believe. As a result of that, on May 1st, HVO issued a warning to residents nearby that an eruption was possible. So what sort of indications did HVO have? As I said, the seismicity was migrating down-rift. That’s not all. So this is pretty obvious. You don’t need sensitive instruments to measure these things. [laughter] Now, to be fair, these photographs were taken on May 17th. So this is a couple weeks later when some of these cracks had gotten a bit bigger. But even early on, they were certainly noticeable as you were out there in the field. Not only did these cracks form and then grow, but they emitted gases for steam and then, later on, sulfur dioxide. And this is very easy to imagine. So magma is reaching – it’s getting closer to the surface. It’s pushing the rock apart. And that’s breaking in and causing these enormous cracks. So the eruption actually began on May 3rd – 5:15 p.m. The initial eruption was a little bit different than the later part. So, over the first week, the eruption consisted of a series of small fissures stretching over about 7 kilometers. That’s about 4 miles. So a fissure is a linear feature where this magma – so it’s migrating down the rift in a tabular form. It reaches the surface as a sort of line, and it creates sort of like a curtain of fire, we call it here. So that’s a fissure eruption. Mostly early on, these were fairly small. So they threw – they threw these lava particles pretty high in the sky. There was a lot of gas, a lot of noise. Not a lot of lava flows were created. It was a fairly localized set of processes. And these fissures would jump up and down – up-rift and then down-rift. And the reason for that was probably, at least in part, that the early lava – the first material that came out was really evolved. So the geochemists could go out there, and they could get samples of this stuff, and they could compare it to what we knew was at the summit, what we know is at Pu‘u‘O‘o, and also what we know came out of the ground in 1955. And what they found is that probably the first stuff that came out of the ground was not 2018 magma. It was actually 1955 magma. So this is stuff that was stored in the ground and was pushed out as new material migrated down the rift. So this stuff was cooler, and it was more viscous. So it didn’t tend to create these really voluminous lava flows. This is what these flows and fissures look like. This is an animation of a fairly slow-moving a'a flow moving down a street in Leilani Estates. The lower right-hand side here, you can see a relatively small fissure. Again, it’s just a very small localized flow around it. There’s a small fissure through the forest right here. And here you can see some houses in the background for scale. It’s very strange to see a fissure open in a subdivision, and obviously very difficult as well. So this is a map of what those fissures look like. This map was made on May 6th, so just a few days after the start of the eruption. This is a thermal map. So what the – what HVO scientists do is they get a helicopter, they fly overhead with a thermal camera. And this is taking photographs of heat instead of visual imagery. So what comes out of that is white for very hot areas and black for very cool areas. Each of these red numbers here is a different fissure, and these were numbered in the order they appeared. So Fissure 1 is right here. That’s the first one, et cetera. What you’ll see is that most of these fissures, again, just have very small little pads of lava nearby. So not a lot of stuff coming out. But at this point, Fissure 8 is putting out a small lava flow, which has moved – well, about a block to the north. That was the first sizeable – or, at the time, it seemed sizeable – lava flow from this event. So, as if things weren’t busy enough, on May 4th [chuckles], the largest earthquake in Hawaii in 43 years happened. It was the biggest since 1975. It was a magnitude 6.9 earthquake, and that’s a good-size earthquake. It was felt 500 kilometers away, so throughout the islands. This is a map which Emily Montgomery-Brown put together. These blue arrows show GPS measurements of the way the ground moved during the earthquake. So this is a map – I should say, so the summit is up here. This is the coastline. Leilani Estates is up here. So, during the earthquake, the whole south flank of Kilauea moved about 0.7 meters – so that’s a couple feet – to the south. And that happened as this slip patch offshore here began to move. This is a 1,500-kilometer slip patch, and it extended about 25 kilometers offshore. So if that’s a little bit confusing, the next slide, which is a cross-section along that red line, I hope will clarify. So here’s a cross-section along the rift. This is Hawaii island, and there’s the Pacific Ocean up there. So, at the base of the island is what’s called a decollément. And that’s the interface between the island of Hawaii and the old oceanic crust – the seafloor, actually. So, as magma intrudes down the rift, it’s actually putting pressure on the whole south flank of the volcano. And we think that’s actually what triggered that earthquake. So there’s a very interesting interaction between tectonic, or gravitational, forces and magmatic forces in the rift during this event. Again, this is – this is research that is certainly ongoing, but this is our interpretation as of – as of now. So I’ve been talking about the earthquake and about the Lower East Rift. What’s going on at the summit at this time? So this is a photograph from early May. And this is the lava lake. Remember, I showed you a photograph earlier when the lava lake was really high. In fact, it had overflowed just before this activity started because the system was very pressurized. So, starting on May 1st, the lava lake started to drop. That surface started to drop in the – in that vent, just like draining water out of a barrel. On May 4th, that rate accelerated. And it picked up to about 2.2 meters an hour. That’s about 7 feet every hour, which, for magma, is a pretty remarkable rate of drop. You could almost see this stuff going down. This right here is an animation I’ll show you. This is a thermal camera video. So, again, it’s not visual. It’s thermal. These colors represent heat. But it shows really the same thing. And this time span is from May 2nd to May 6th, so it’s just four days. Actually, it’s May 8th. That’s incorrect. So it’s about six days. And this is accelerated, of course. But what you’re going to see is you’re going to see the surface of the lava lake drop over that six-day time period. What you’ll also be able to see is, as it drops, it’s de-stabilizing the walls of the conduit. So pieces are going to fall in. And you can see that that makes the surface very turbulent, and you can see little patches forming on the side here. So pretty remarkable process. That ultimately drained about 1,000 feet.

- [audience reactions]

- And really vanished from sight on around May 10th. So relatively quickly – in about a week, it started dropping and then completely disappeared from sight. So let’s go back to the simple cross-section here. Why is the lava lake dropping? What’s going on here? Well, so as I said, magma is coming up from the mantle. It goes through that reservoir and then out to Leilani Estates. What was happening is that the rate of magma going out to Leilani was way higher than the rate of material coming into the reservoir. So it’s just like letting air out of a balloon. The reservoir depressurized. And, as the reservoir depressurized, the surface of the volcano dropped – it subsided. And the lava lake dropped as well. So you depressurize the reservoir, lava drains back into it. And that’s what was happening in May. So this is a time series from April 29th up until May 16th. Each of these red dots represents the height of the surface of the lava lake. And that’s measured by a laser range finder right on the rim there. So you shoot a laser beam down, you measure that distance, and you convert that to a height above sea level. So, before this activity started, that was a little over 1,000 meters above sea level. Here you can see it starts to drop. And the magnitude 6.9 earthquake happened around here. And it really just dropped at an incredibly constant rate. In volcanology, constant rates like this are unusual. Things are always changing. So when we saw something going really steady like this, it gave us a little bit of pause. It also meant that we could [chuckles] draw a line through it and ask [laughter] what happens? It doesn’t take a geophysicist to do that. Although it does take a little bit of courage, because you don’t, of course, know what’s going to happen in the future, but we did. And we asked ourselves, when is the lava lake going to reach the height of the water table? The water table at Kilauea is about 600 meters above sea level – about. So who cares about the water? Why does that matter? Well, this is why it matters or why we thought it matters. [laughter] So this is the prevailing wisdom for what causes these sorts of explosive behavior at the summit of Kilauea volcano. So the idea is, here’s your lava lake. It drains out of sight. Let me back up. So here’s a lava lake, and it’s been there for a while. That heats the rock around the lava lake – the conduit, okay? That rock is really hot, and it presents a barrier for water. So water can’t flow through really hot rock. It turns to steam and blows away. So you can see a little – a little barrier there between the water and the magma. Water and magma don’t touch, you’re okay. Lava drains away. Well, now the conduit starts to collapse. It cools a little bit. Water can start infiltrating. Then you get water on top of magma, and you get those things mixing, and maybe a little bit of sealing, and you can get a phreatic explosion. That’s one that’s driven by steam, okay? So this was the concern. And there’s a precedence for this. This is Kilauea in May of 1924. There was a lava lake at Kilauea in 1924. There was a dike intrusion – an intrusion of magma into the Lower East Rift Zone. The lava lake drained, and explosions happened just like this. There were fairly substantial. They lasted a couple of weeks. They threw blocks – I mean, you can see the size of that thing. These things weighed many tons – up to 2 meters across – for more than a kilometer. Ash fell across a broad area of the Big Island. And even gravel-sized rocks were deposited as far as 3 kilometers away. So this was a major hazard that HVO and the Survey were concerned about. And the national park as well. The summit of Kilauea is in a national park, and of course, they have a responsibility to keep people safe. So, as a result of this analysis, on May 8th, HVO released a preliminary summit hazards assessment in which we said, the potential for the summit lava lake surface to drop to or below the elevation of the groundwater table raises the strong possibility of explosions caused by interactions of groundwater and hot rock. At the current rate of decline, the lava lake may drop below the water table near the end of this week. That was the intersection of that line with the water table. HVO is located right here. The lava lake is right here. That’s a distance of about a mile. As a result of this analysis, Hawaii Volcanoes National Park closed on that Friday, on May 11th. And it’s actually still closed months later. And also, HVO moved as well. We’ve moved a little more slowly, but we gradually migrated down to the University of Hawaii in Hilo to be a little farther from the summit. That’s partly due to the danger of the explosive eruptions and partly due to earthquake shaking, which damaged the building. But the idea was, people had to get a little further from the potential of explosive activity. So indeed, it did start. The video on the lower right hand is accelerated. So that’s 10X so you can get a better sense of the way these plumes go up into the sky. It’s actually quite pretty to watch. The first of these started on May 4th, and these were triggered – so, as I said, as the lava lake drained, the walls became destabilized, and so rocks would drop in, and those would cause little explosions. On May 4th, the first of those plumes triggered by that mechanism started. These got a little bit stronger. From May 16th to 26th, there were 12 larger explosions. They varied anywhere from 8 to 45 hours apart. And they released actually a fair amount of seismic energy. Equivalent of an earthquake of magnitude, maybe, 4.7 to 5.1. But it was released over a fairly long time period, so we didn’t actually feel much of it. Mostly what we saw were just these explosions. So the summit continued to deflate. This is not – this is not the lava lake surface height anymore, although it looks like it. This is a measure of ground deformation at the summit. And you could think of this as a proxy for inflation and deflation of the reservoir. So, as that red line goes up, it means the reservoir is inflating, and as the line goes down, it means the reservoir is deflating. And you can see that it was certainly deflating. That’s a 6.9 earthquake. It was deflating at a very constant and very high rate. The lava lake disappeared around May 10th. Then we got down here, and you can see some really interesting stair-step patterns in the data. So what was causing the – oh, I should say, actually, we were able to model this. As this was happening, we ran some very simple models. And we were able to say that, all right, so probably a couple million cubic meters of magma per day were leaving the reservoir. We’re able to say the reservoir is about a kilometer and a half deep, and it probably held a few cubic kilometers of magma. So that gives us some ideas about magma storage at the summit there. So, yeah, each of these stair steps, what were those? Each of those is an explosion, actually. So interestingly, and intriguingly, and a source of much research in the future, I think, is understanding why each of these explosions caused a little inflationary bump in our data. This is not accelerated. This is just to give you a better sense of what these explosion plumes look like at the summit. This is not the beginning. They tended to come out a little more rapidly and then slow. But they were certainly quite impressive to see. The biggest of these reached as high as 30,000 feet. What did they actually – what did they actually kick out? Well, ash, of course. You can see ash in those photographs and video. Also ballistics. Ballistics are just chunks of rock that are launched like bombs from the explosion. These are ballistics from the 2018 explosion on the old road down in the caldera. The biggest of these was 10, 20 centimeters across, so decent size. The ash fell to centimeters thick, maybe. You certainly wouldn’t want to be down there when one of these things happened. But it was actually smaller than 1924, at least based on our preliminary analysis. However, it looked a lot like 1924. [laughter] So I actually made the photograph – excuse me – black and white to make it look more like 1924, but May 18th, 1924, May 15th, 2018. Really remarkably similar. And that actually raises some interesting questions because, in 1924, they argued very strongly, or they believed, that the explosions were caused by the interaction of water and magma, like that figure I showed you earlier. Well, in analyzing the data from 2018, at least to the extent we’ve been able to do so far, there’s actually no evidence that water was driving these explosions. We think that probably they were driven entirely by magmatic gases. So that raises the question, maybe that was true in 1924 as well. And if you go back and look at their data, they did a lot of speculation. They weren’t 100% sure, either. So this actually causes us to question, maybe, some interpretations of a century ago and think a little differently about what causes explosive behavior at the summit of Kilauea. So that’s one of the really interesting changes in our thinking that I think is going to come out of this eruption. So mid-May – let’s go back to the Rift Zone now. Things changed a little bit. So, as I said, in early May, the material that was erupted was very viscous. It was very cold. It didn’t flow very far. In mid-May, that really changed. So the eruption rate really picked up, and we started getting fast-moving flows which reached the ocean. You can see right here – this is the first ocean entry, I believe, on May 23rd in the lower left-hand corner. Really impressive channels of lava starting to form. On May 27th, activity focused at Fissure 8. And if you’ve watched the news at all over the last few months, you probably heard of Fissure 8. It became semi-notorious as the source of most of the lava flows in the Lower East Rift Zone at Kilauea. So, after the 27th, and really all the way through the beginning of August, this is where most of the material was coming out of the ground in the Lower East Rift. It produced fountains. These fountains went as high as 80 meters. As the tephra – the material – came back and landed, it gradually built a cone. So you can kind of see it in the dark here. None of that was there before. This was – this was literally a subdivision with forests. And now there’s a fountain and these tremendous – a cone and these tremendous fountains. That fed a channel – really, a channel of lava, which looked a lot like a river that extended more than 10 kilometers, and it reached all the way to the ocean. Our preliminary estimates of the eruption rate as high as 150 cubic meters a second, and I’ll talk a little more about that. Oh, sorry – gas emissions. Gas emissions increased to more than 40,000 tons a day, which is absolutely huge for Kilauea. It’s much higher than before this event started, even though Kilauea had been erupting since 1983 semi-continuously. So massive gas emissions, and this strongly affected air quality on the island. That vog plume – that plume of bad air – extended as far as Guam and elsewhere, where people were actually warned to not go outside and exercise. So tremendous air quality impacts, not just on the Big Island of Hawaii. This is Fissure 8 showing on the top what those fissures look like. Really tremendously beautiful to watch in a lot of ways, although [chuckles] a difficult place to stand. [laughter] Well, a lot of it depended on the direction the wind was going. As I said, there was a lot of SO2 coming out. And if you were upwind of that, you could – you could stand here and be very comfortable. But if the wind changed, it would be very toxic very quickly. The photograph on the – or, the video on the bottom is just downslope of Fissure 8. And that’s – what’s happening is those fissures are rising up, and they’re coalescing into a river of lava which extends to the ocean. And this is what that river looked like just downslope of Fissure 8.

- Is that speeded up?

- That is not speeded up.

- [audience reactions]

- That is actually real velocity. Yeah, I’d never [chuckles] – most of us had never seen anything like this. And, to be honest, never thought that we would. This did surge. It would get higher, and it would get lower. Sometimes you could almost not see it from where this video was taken. Other times, it was really, really quite incredible.

- Is that characteristic of the pahoehoe flow? That really fast-moving …

- The Pahoa from several years ago?

- Well …

- Oh, pahoehoe.

- … there was a'a and …

- Yeah. So the question is about the type of lava that this is. Yeah. This is a very, very fluid pahoehoe. Well, it’s [chuckles] – it’s really a channel. It’s channelized flow, yeah. It’s not an a’a flow. A’a flow is what I showed you earlier, where it’s moving very slowly. Further downslope, it did form a’a. So this is Fissure 8 in the top right-hand corner feeding these channels. And these became very stable, and they existed like that for, really, most of June and July with relatively little change. They really did look like rivers. If you imagine taking out the lava and putting in water, it wouldn’t look necessarily so out of place. One fundamental question for understanding eruption is, what is the eruption rate? How quickly is stuff coming out of the ground? That helps us understand, is it changing? Is the eruption slowing, et cetera? So how do we measure that? It seems easy, but [chuckles] it’s kind of not, right? How do you actually know the rate at which – the volumetric rate at which magma is being erupted? We had a really neat tool, which was applied during this eruption, I think, maybe most – almost most successfully at any eruption. And that is unmanned aerial vehicles, or drones. So these are little – well, you’ve seen drones – larger, typically, than the ones you buy on the shelf. But the same idea. And they would fly these over the channel, and they would actually take video looking straight down. And by tracking little particles, they could estimate velocities. And the figure in the lower right-hand corner shows the velocity across – a profile across a channel near Fissure 8. So this is horizontal distance – zero to 60 meters. Velocity is on the vertical axis. That’s 9 meters per second, or about 20 miles and hour. So really incredible rates. I like to say that, if the flow is near an airport, it would actually get a speeding ticket. [laughter] It’s important to emphasize that this was after the channel was established. So, when the flows were first moving over open ground, they definitely did not move this quickly. But once they form a very efficient channel, they can move very, very rapidly. Also, those flux rates slowed as you got further downslope. But near the vent, they were incredibly high.

- How do you measure how deep it is?

- Well, [chuckles] it’s easier to measure now that things have slowed down a lot. Actually, what they use is some fluid dynamical arguments. They can look at the – sorry, the question was, how do you know how deep the flow is? And you don’t, necessarily, at the time. You can use modeling to try to interpret, you know, the cross-sectional area – or, sorry, the width of the flow. Maybe you see standing waves, so you can use fluid dynamical arguments to make some inferences. But that’s one of the biggest uncertainties. But, if you can estimate the depth, then you can get a flux rate. And that’s what they’ve done, and that’s about 130 cubic meters a second dense when you remove all of the bubbles. And that’s a pretty tremendous rate. 1955 and 1960 were on the order of 30 to 40 cubic meters a second. So much higher rates. These flows reached all the way down to Kapoho Bay. That bay filled quickly, and it destroyed hundreds of houses. Many of you may have been on the Big Island. This used to be a great place to go snorkeling and swimming, and it is all, unfortunately, completely gone now. This is an animated map showing how the flows actually evolve. So it starts from early May. This is that first little flow coming out from Fissure 8. I showed you a video of it. It jumps down-rift, jumps back up-rift again. In mid-May is when the viscosity went down. And so the effusion rate went up. And then this is Fissure 8. So we put that flow all the way around to the north. That was Kapoho Bay, completely filled in. And then, from there, the channel became relatively stable. So what you mostly see after that time is filling and broadening of that ocean entry region to the top right of the picture. One question I often get is, how much – how much has actually erupted? And one of the reasons that’s very difficult to answer is, look how much of that is in the ocean. So a lot of it has vanished into the water. If it’s on land, we can map it. We can estimate the thickness and get a volume. But if it disappears into the ocean, that’s actually very hard to say how much came out. Okay, so let’s go back to the summit again. Around the time that activity focused on Fissure 8, the vent at the summit became largely blocked. So this is what the summit looked like. This is zoomed in on Halema‘uma‘u . That’s the crater inside the bigger caldera structure. The red shows where the lava lake was. This is what it looked like before the activity started. And this is what it looked like when we were able to map it using drones with the permission of the park on June 1st. So you can see that that vent area has collapsed from the north and south. Part of the rim has collapsed. It took some of our instruments with it. And at the same time, the figure – the photos at the top right show was the degassing plume looked like before that happened, and then after. So a lot of the plume disappeared because the vent was blocked. And that changed the character of the eruption. Explosive activity, like the videos I showed you earlier, that actually diminished after this time. Collapses continued to occur, and I’ll show that, but mostly the explosions slowed and eventually ended around this time. So this is a video, or an animation, of – it’s a time-lapse, actually – a photograph taken from HVO – from the Hawaiian Volcano Observatory looking out over the caldera. So this is Halema‘uma‘u Crater right here. And you’re going to see a few things – you’re going to see – this is the lava lake right here. You’ll see some little overflows. This starts in April, I should say, and goes up through the beginning of August. You’re going to see subsidence starting over here and then over here. And in late June, extending to the back – the back side of the crater as well – the caldera. And I’ll probably play this a couple of times so you can get – try to absorb everything because there’s a lot that’s going to start happening. Another thing you’ll see here, these are the explosive plumes starting in mid-May.

- [audience reactions]

- Really incredible changes. [laughter] To put it mildly. I keep saying, these are things that none of us ever thought we would actually see. We know these sorts of behaviors are possible at Kilauea, like other volcanoes, but none of us thought we would actually see it. I’ll play it one more time. So it starts here and here and then broadens to the backside there. And it really started quite suddenly and stopped, actually, quite suddenly as well. We had a GPS instrument on one of those blocks that started dropping. And that was dropping as much as 30 feet during each of these events. So this makes it look semi-smooth. In fact, it was an episodic kind of behavior, which I’ll talk more about. In each one of those slumps, that instrument was moving tens of feet, which is huge. So we flew these UAVs – these drones over the summit as often as weather would allow. And we’d build a map every time – an elevation model of the summit. And that allowed us to quantify what we just saw in that video. So here’s a movie – an animation of that. Here’s subsidence forming in the southwest and then really broadening into the broader part of the caldera there. To put this in perspective, that’s the Hawaiian Volcano Observatory. [laughter] So that is a building. [laughter] This is about 1 kilometer across, so it’s a – it’s a fairly broad area. We can actually see the old road going down into the caldera here. The parking lot was here. That was completely swallowed up. That had been closed since 2008, so no one was down there, but completely gone now. One more time-lapse animation. This is taken with a time-lapse camera set up right here looking across this part of the caldera floor to capture that broader subsidence you saw east of the crater. It really kind of looks like water, doesn’t it? It’s hard to believe that’s solid rock. And you can see that it really localizes on that big fault in the back – the backside there. I’ll do it really fast.

- [audience reactions]

- Yeah, I could do this all day. [laughter] So that motion was not continuous. The caldera was not smoothly subsiding. It was actually subsiding in these steps. So I showed you that step time series before. It was red. During the explosions, where the caldera would subside, or the – really, the reservoir would deflate, and then it would inflate rapidly. And early in the eruption, each one of these pulses, these inflation pulses, was associated with an explosion. Later on, that was not true. There was no explosion. But what there was is a slump of the caldera. The whole caldera floor would move down as much as tens of feet during each one of those events. And you can see how almost steady they are. They happened a little over a day – 30 to 40 hours. And they were correlated very closely with earthquakes – seismicity. So this figure at the bottom shows earthquakes per hour. And I should say – the time period here, this is from June 16th to June 21st, so a little over a week here – a little under a week. So this is earthquakes per hour. And what you can see is that, each time one of these subsidence events happens, the earthquake rate drops way down, almost to zero. And it ramps back up again. An event happens, and we go through the cycle again. So people who live at the summit, near the summit, in Volcano, they would know exactly where they were in this cycle. If one of these was happening at night, it was very difficult to sleep. Really tremendous numbers of earthquakes – up to 700 in one of these cycles. Now, most of these would be too small to be felt, but I spent time up there, of course, observing this. And in one six-hour shift, I felt 32 earthquakes myself, and that was just a part of a day. So I had a lot of sympathy for people who had to live up there through this and try to get a good night’s sleep. So what was happening? Why is this episodic behavior occurring? Well, it’s a really interesting question. And I guess the – we don’t know yet, but our working hypothesis right now is something like this. So here’s another simple schematic. This is that magma reservoir at about a kilometer and a half deep beneath the caldera. So magma is going out to the eruption site to the Lower East Rift at a really high rate. So pressure in that reservoir is decreasing. At some point, you could imagine that the rock over that reservoir becomes unstable, and it’s going to – it’s going to slump down, or collapse. So it collapses down, and that increases the pressure in the reservoir. And it stabilizes. Magma continues to evacuate. It slumps down again. So it’s episodic plug-slip, like a piston, response to semi-steady eruption of magma towards the rift zone. This is not a new idea. This has been speculated at other caldera collapses before, like Nakajima in Japan. And in fact, they have data there which looks very much – this sort of stair-step pattern, they have data from that eruption, which looks almost eerily similar to ours. So what’s really interesting – well, there’s a lot of really interesting things. One really interesting thing is that, even though this was occurring episodically, the actual average rate at which the caldera was subsiding was really constant. So we get these volume estimates from the – from the drone flights, so they [loud static noise] – I hope I didn’t do that. They build the elevation models. [loud staticky noise] They build the elevation models, and they can compute a volume and compare that to the previous one. So each of these red dots is one of those measurements. And this extends from June 1st all the way up to the beginning of August. And you can see how incredibly constant that is. So, again, we can fit a line to that. That’s really easy. That is a rate of a little over 12 million cubic meters per day. And, I don’t know, it sounds like a lot, but maybe it doesn’t mean much. It’s about 14 dumptrucks a second. I don’t know if that helps. [laughter] And if you assume a density, it’s about 360 tons a second. So really big numbers that are maybe hard to conceptualize, but they’re pretty enormous. This is important because we think that, to first order – so let’s go back to this figure. This plug is subsiding in response to flow out to the eruption site. So maybe those rates are about the same, right? So maybe this is approximately the eruption rate. And that’s really neat because it tells us that was relatively constant from early June, and it gives us a handle on what it might be. As you saw from the channel flux measurements, it’s not necessarily easy to get that at the eruption site. And maybe this is our best constraint on the eruption rate during the eruption. So that brings me to the initial question here. What on Earth is going on at Kilauea volcano? And the answer, right now, is not very much, actually. [chuckles] In fact, almost nothing is going on right now, which is not the answer that I thought I would be giving you when I was planning for this talk. The figure at the top is that same time series with the little stair steps. So inflation is up, and deflation is down, on that red line. Seismicity is on the bottom. Right around August 2nd, everything really quieted down. So we had our last collapse event on August 2nd. And deformation just became incredibly flat. Number of earthquakes? Really dropped off. Almost nothing now. The red line here, by the way, is cumulative earthquakes over the course of the eruption. That’s 40,000 right there. Absolutely incredible number of events. These photographs are down in the rift zone, so this is the channel. You can see how cool and black it is. And this is Fissure 8. So a little bit of steam. A few days ago, there was a little bit of lava in the bottom of that, but really nothing erupted anymore. So incredibly quiet. So if the eruption is over, and I don’t think we’re ready to call it over yet. It may be a pause, but the longer time goes on, maybe the more likely the Lower East Rift part of the eruption, anyway, is over. We can start to ask questions about, how big was this? How long did it last, et cetera? One of the questions, of course, is how much did the caldera subside? So these are cross-sections at the summit. This is measured – well, it wasn’t measured in April of 2018. It was measured before that, but it didn’t change, so let’s call that April of 2018. This is a cross-section east-west across the caldera. And then again measured in August of 2018. That subsidence is more than 500 meters. That’s the Empire State Building for scale. So really enormous collapses in the summit. So, as I said, that was 12-1/2 million cubic meters a day, and this is – this is what you get when you subside 12-1/2 million cubic meters a day for as long as we did. So we can put that into historical context as well. This is a time series from 1800 all the way up to the present. The vertical axis is summit collapse volume. So the summit of Kilauea has collapsed in the past. It’s collapsed many times, in fact. It’s collapsed in geologic history – and, of course, for those, we don’t have a very good constraint on the volume. But the written record at Kilauea goes back to about 1800, and we have some constraints on those. So these are those volumes. 1840 – right up here, about 500 million cubic meters. Where’s my mouse? 1924 was about 200 million cubic meters. And there’s 2018 up there. So it’s certainly an outlier. In fact, you could say that it’s unprecedented in Kilauea’s written history. That’s about 0.8 cubic kilometers of material subsided at the summit and probably erupted as well. Calderas have formed at other volcanoes as well. Well-documented caldera collapses are very rare. There’s really just a handful of them. Here are a few. Piton de la Fournaise in the Indian Ocean, Tolbachik in Russia, Miyakejima in Japan – all fairly recently. These volumes are all smaller than Kilauea. There were a couple, however, that were larger. Fernandina in the Galapagos in 1969 and Bárdardbunga in Iceland, both close to 2 cubic kilometers. So Kilauea 2018 is not the largest ever, but it is certainly significant. And it is also, I would argue, by far the best-documented. So before I close, I’ll leave you with a parting shot. This is [chuckles] – this is an oblique view of what the summit of Kilauea looked like, let’s say, in spring of this year. And then this is what it looks like now. And so all the data goes into this, but sometimes it’s good to step back and really get a sense of how things have changed. The parking lot that I was talking about is right here. You can see how that disappeared. The gray color is due to ash, mostly from those explosions in May. And, again, the Hawaiian Volcano Observatory is right there, still [chuckles] standing there on the rim of a much deeper caldera.

- Is there public access to the observatory now?

- So the question is, is there public access to the observatory now. Actually, the entire park is closed and has been since May 11th, so no. [chuckles] So in conclusion, then, I think, during the eruption, history really guided our thinking about what the system might do. So we have the geological record. We have the historical record. And that gives us an idea of what Kilauea is sort of capable of doing. Then we coupled that with the monitoring data – real-time monitoring data. That told us which direction – which of those possible outcomes the system is maybe evolving towards. And that was really nice because I think it allowed HVO to issue timely and accurate warnings. This was an enormous event, but no lives were lost. It was – it was really a human tragedy in a lot of ways. There were more than 700 houses destroyed. But nobody died, and that’s a major success. Second point I’d like to make is, as I said, I think no caldera collapse, even globally, has ever been as well-recorded as this. I think not even – not even close. So we really have an unprecedented opportunity to learn not just about Kilauea, but about the mechanisms that drive caldera collapse to volcanoes around the world. And really the work to – the work that we’re doing to understand that is really just beginning. There are decades of work in trying to understand the data that we’ve collected over the last several months. Finally, the magma system at Kilauea, as I’ve shown, has really changed in important ways. It was sort of stable over the last 35 years. The Pu‘u‘O‘o eruption, we kind of got used to that. Everything is different now. And, of course, the big question is, what happens next? And I actually do have the answer for that. And the answer is, time will tell. [laughter] So I’ll leave it with that. Happy to take any questions. Thanks. [Applause]

- I’m sure that many of you have questions, and we’ll get to those. I do want to ask you – we have a couple of microphones set up in the aisles. If you’re not able to get up to a microphone, I can bring you one. I really encourage you to please, please use the microphones because we web stream these lectures live, and although your voice might be loud enough for others in the room to hear you, we cannot hear you on the web stream nor on the video that we record for each of these lectures. I also want you to be very, very careful. The room is over-full. We’ve got some temporary chairs. People are sitting on the floors. Please do not trip over each other. Take your time getting to the microphone. We will not rush you. Please be careful, and I see one gentleman already waiting in line to ask your question. So please, go ahead.

- How much do you understand what’s going on below the surface? When we talk about magmatic chambers and collapses, are there voids that happen? I mean, a vacuum, and then the roof of the chamber collapses? What do you know about the air and tubes and how all that happens?

- So, in terms of how well we understand, I guess I can answer relatively. At Kilauea, we know pretty well. But it’s a volcano, which means we don’t know it very well at all. [laughter] Unfortunately, there are no direct measurements. So really fundamental things, like how much magma is stored there, you’d think, well, we should just go measure it. But how do you actually measure that? There’s no direct measurement. So we have a pretty good idea of where magma is at the summit. In fact, before this started, we thought that there was a magma reservoir at about a kilometer and a half depth below the center part of the caldera. And this event showed that there was a definitely a magma reservoir there because it collapsed. At least part of it did. So it confirmed some of our ideas. We don’t believe – so there may be void space down there. Probably not at great depths. I personally don’t envision that a lot of void space was created during this. I think that the magma was flowing out to the rift zone, and that depressurized the reservoir, and that allowed the roof to collapse, which really prevented void space from forming. I think the – in order for void space to form, the pressure has to drop so much that the rock is fairly weak, and it tends to collapse before that happens. Void space can certainly occur shallower in the system. You’ve all probably seen lava tubes, right? So there are voids underground, and some of those actually get fairly deep. So they can’t be ruled out. But I don’t think there was a lot of void space in the reservoir. Although there’s probably bubbles and things in the magma, which could be thought of as void space.

- Hi, thank you. I believe you mentioned you did some – you had some mathematical modeling earlier. To what extent is that refined now? And assuming you’re refining it now, but what’s your prediction as to how accurate that’s going to be in the future?

- Yeah. Good question. Refined and partially rendered obsolete, perhaps. [laughs] The models I worked on before are for a system which may not exist anymore. Most of my work was done to understand the Pu‘u‘O‘o eruption. And some of those things are fundamentally different now. Other aspects of the models are the same. The magma reservoir at the summit, I – well, one question is, how much magma is still stored there? There may still be some. And that’s a feature that was in my model before, and hopefully I’ll continue to try to resolve. So the models are going to – all of our models are going to require some revision and some rethinking, but, you know, it’s a – it’s a great opportunity. The Pu‘u‘O‘o eruption was a fantastic chance to understand the system because it had been erupting for so long. But this is such a perturbation to the system that allows us to get at things that I think we never would have gotten at with another 30 years of Pu‘u‘O‘o. So I think new models will be called for, and I’ll let you know in a few years how they turn out. [laughs]

- Thank you.

- Thank you.

- Hi. You talked a lot about the physical thing. I’m wondering if you can talk – and you mentioned dates, and I didn’t quite get them all clear. But you knew things were going to happen ahead. You warned people sometimes, and people left. And I’m wondering if you can give a little summary of that and how many houses were destroyed which people might have been killed if they didn’t have warnings.

- Yeah, so there were – HVO gave a number of different warnings. And, in fact, not just warnings, but HVO released daily updates and even more-than-daily updates, just to make sure that people were informed. The role of the USGS and HVO is not to directly tell people to evacuate or not. So we do the science, and we give that to the civil authorities – so Civil Defense in Hawaii. And they make the decisions about whether or not people should be evacuated. And actually, at the summit, it was a little bit different because that’s in a national park. So, at the summit, our job was to provide the science to the park and let them make those decisions. So we issued a number of warnings throughout the eruption. When magma was approaching Leilani, those warnings were released, at least in enough time – enough time in advance for people to leave. At the summit, people were evacuated and the park was closed sometime before the first explosions started. 700 – and so, the last number I had was more than 700 houses were destroyed. These – so, normally Hawaiian-style lava flows are not immediately threatening to life. You can typically get out of the way. That was a little different here. Some of these flows moved really quickly, and they isolated people. And that did happen. People had to be rescued by helicopter, for example. So it was, I would say, much more hazardous than a typical Hawaiian pahoehoe-style eruption. And I think it was – yeah, it was the combination of USGS warnings with really a lot of hard work from Civil Defense and the National Guard to keep people safe is what meant that no one died.

- Thank you. I’m hoping that, in the spring, this will be a Nova special because I need to show this to my grandchildren. They will just be astonished by this. This is really something. My question was, did this outflow of all this magma do anything for the developing seamount that, I think, down to the southeast of the Big Island? Did it slow it? Could you see any impact on it? Or are those totally different plumes?

- Yeah, so I believe you’re referring to Lo‘ihi, which is a seamount, which is sort of the next Hawaiian island. We don’t know that this impacted that system at all, and I wouldn’t necessarily expect that it did. As far as we know, Lo‘ihi has not erupted for some time. As of now, this is a Kilauea phenomenon. It will be, I guess, remain to be seen if it has any effect on other systems nearby, including Mauna Loa, but right now, no.

- What have you learned so far from the earthquake catalog? Location, depth, pattern – any significant features emerge?

- Yeah. The earthquakes – I can speak mostly to earthquakes at the summit. So after the dike – the magma intruded into the Lower East Rift, it actually became seismically very quiet. And most of the seismicity was focused at the summit. So some relocations have been done. This is where you take the earthquake catalog, or the locations of those – or, those earthquakes, and you try to more accurately figure out where they are. And that’s really useful because you can pick out planar features. You can start to relate those to geological features, et cetera. In the caldera, it was interesting because, early on in the subsidence, in the collapse, there were these little groupings of earthquakes throughout the caldera, which seem to have almost no relation to what was happening. The subsidence was right in and around Halema‘uma‘u , but the earthquake clusters out in the east and central caldera were – yeah, central caldera. It was interesting because it was only later on, a few weeks later, when that broader part of the caldera started to subside, that some of those earthquake bands matched that reasonably well. So I think – we haven’t done a lot of analysis yet. We simply haven’t had time in terms of what that means. But I think that – I do think the earthquake relocations will give us insight into what was happening beneath the surface as the caldera collapsed.

- It sounds like it’s all caldera collapse and not anything occurring at greater depth.

- Most of the – or, most of the earthquakes were quite shallow, yeah.

- Okay. Thank you.

- I believe there was a tourist boat and that some people were injured. And there was some aerial explosions, and then the particles came down – or the globs came down. Would you describe that?

- Yeah. So there was a – there were lava tour boats that went out during the eruption. The ocean entry was tremendously incredible to see. And a lot of people wanted to see it, so there were tour boats that were out there. One of them was relatively near the shore – I think still in a legal area, but relatively near the shore, and it was struck by a volcanic bomb – a rock which was ejected during a littoral explosion. It actually came through the metal roof of the boat and injured someone. So what we think happened was probably – the eruption rate of the lava into the water was really high. And so it actually created probably tubes under the water – so lava tubes. In fact, in the helicopter, when we’re over the ocean entry, you could see strong regions of upwelling several hundred meters offshore, maybe at the end of one of those tubes. So what may have happened during that event is that one of those tubes collapsed – imploded. And so that allowed a lot of water to interact very suddenly with a lot of magma. And that would flash to steam and cause an explosion. That may very well be what happened to the boat. Those sorts of explosions were not so uncommon out there. We saw them from the helicopter. You definitely don’t want to get too close to a high-volume ocean entry like this.

- Thank you for a wonderful talk. Are traditional, you know, oil exploration techniques for, you know, making noise and listening going to help you find anything?

- Right. So I guess you’re referring to sort of active-source seismic, where you make a noise – thump the ground or something – and then you map the propagation of seismic waves. Those sorts of techniques are applied. This event was quite fast. There was [chuckles] – there was nothing like that that was done. But what we have instead of this active-source seismic, where you take trucks out and hit the ground, is we have tens of thousands of natural earthquakes. So people are certainly using those. You can – the idea here is, you map not only the locations of the earthquakes, as the one gentleman was asking about, but you can map, for example – or, detect changes in the velocities, maybe, between two seismic stations and look for changes in the properties of rock in the caldera, for example. So those direct oil exploration techniques, I don’t know of any being applied, but there are a whole host of seismic studies which will come out of the data that’s already been collected.

- I wonder how these eruptions have affected the larger-scale development of the Hawaiian islands. Have large populations shifted or moved? And what of the people whose homes have been destroyed? I wonder what’s going on in that context.

- Yeah. I can’t speak to that very directly. I’ve been so focused on the science aspect. That’s, I think, the question going forward now that the eruption has slowed, although it is important to say that it’s a little early to say that it’s over – that the Lower East Rift Zone is over. And so that gives us pause, or gives civil authorities pause in terms of thinking about what comes next. When the eruption in the Lower East Rift is over, then there are going to be a lot of decisions that have to be made about relocating people, et cetera. Those are not decisions the USGS is directly involved in making. I know that – boy, there were a whole host of impacts to the population – to tourists – tourism was down on the Big Island. That affected the local economy. The national park was closed, so that strongly impacted the town of Volcano – restaurants and things like that. So really tremendous impacts. You know, a lot of people have the perception that this eruption was destroying the entire island or something, when it was really directly impacting a relatively small swath of it. So I think a lot of people stayed home. Hopefully that will change. In terms of people living down in Leilani Estates in – along the Lower East Rift Zone, it’s a rift zone of an active volcano, and a lot of people understand that. But what happens in terms of land use decisions going forward, I think, will be interesting to see.

- This relates to a previous question, and it’s connected with what I call 3D acoustic imaging, which would give you the size of, say, the magma chamber if you could get – watch the earthquake reflections on enough locations at the same time. Has that been done?

- Those sorts of studies have been applied, and we see changes in velocities where reservoirs may be. The reservoir at the summit – the shallow one that subsided and partially collapsed during this event – is relatively small. So it’s actually very difficult to see with seismic techniques because the wavelength of those waves is too long.

- Right. Mm-hmm. And connected to that, you mentioned one place that the water level was 600 feet above sea level. I’m wondering what hydraulic pressures are keeping it that high and it doesn’t drain off to the sea.

- Yeah. It’s actually 600 meters.

- 600 meters. Of course.

- Yeah. I’m not a hydrologist, but I know there’s a lot of rain on the Big Island. [laughter]

- Right. That’s [inaudible].

- Over here.

- Oh, yeah.

- Can you tell us what the essential difference is between Kilauea and the other volcanoes on the Hawaiian island and the string of volcanoes we have on the West Coast of the United States?

- Yeah, so the fundamental difference between West Coast mainland volcanoes and volcanoes in Hawaii is those are hotspot volcanoes. So the idea is, there’s a plume of magma, or some material, coming up from the mantle. And the Pacific Plate is overriding that. And, as the plate moves, it creates a series of islands. West Coast volcanoes – Mount St. Helens, Shasta, Lassen, Rainier – these are subduction zone volcanoes. So one plate is actually diving beneath the other. And that process creates melt, which rises up into the overriding plate. So tectonically, they’re fundamentally different. And, as you can see, usually the form of the volcanoes is quite different. Hawaiian islands create these long, broad shields. The Cascade volcanoes are typically – they’re the pointier – they’re more viscous material, and they’re usually more explosive as well. So the tectonic setting is really fundamentally different. All the Hawaiian islands are – for example, on the Big Island, Mauna Loa, Mauna Kea, Kilauea – these are all shield volcanoes. So fundamentally, they’re the same class of volcano.

- Do you expect the observatory to still be usable at some point? The structure itself? And do you know what the national park is going to be like now, or at some point in the future?

- Yeah. Those are both great questions. And I wish I had better answers for you. This eruption has been really challenging for HVO. You can imagine responding to the biggest event in generations while your office is being destroyed by earthquakes. [laughter] It’s not easy. [laughter] So the building – the state of the building, that’s a good question. There will be structural engineers going in to evaluate it, et cetera. And I don’t know what the outcome will be. There are advantages and disadvantages to being that close to a volcano. [laughter] HVO is a fantastic place to be. To be in the – in the tower and watch lava flows in the crater, it was unbeatable. But for really outlier, unusual events like this, it was – it was close. So I don’t know. In terms of the park – so, the park is still closed. They’re slated to hopefully at least partially re-open, I believe in September. It will not be the whole park initially. There’s a lot of infrastructure damage. There’s roads. They had no water. So I imagine it’ll be a gradual opening, and then there will be a lot of decisions to be made about where roads should be built, et cetera. There was a road – so I showed you that parking lot that fell in. That road was open until 2008 when the lava lake formed, and gas, et cetera, from the lava lake meant that that road has been closed for the last 10 years. It’s possible that there could be another road down there. Right now, there’s no lava lake. There’s no gas. The air is quite nice down there. So this is – I guess you could look at it as an opportunity for the park to make decisions about long-term land use going forward, and I’m sure they’re having those decisions – or, having those discussions right now.

- Two questions. One, could you tell us a little bit more about the effect that the eruption has had on the bay? And the second one is, what would have happened if the hurricane and the eruption had overlapped a little more? [laughter]

- Okay, so by the bay, you mean Kapoho Bay there?

- Yeah.

- Yeah. So the bay is gone. The interaction is, there is no more bay. The lava filled the bay completely and extended well offshore from that point. So, yeah, simple answer, unfortunately. The next question – yeah, a few people have asked me that. And it’s an interesting one. So, yeah, the activity stopped in early August, and Hurricane Lane came fairly soon after that and dumped 52 inches of rain in Mountain View. Not Mountain View, California. [laughter] Mountain View, Hawaii. There is speculation of volcanoes around the world that large amounts of rainfall can actually trigger eruptions, can change behavior. Groundwater percolates down. It interacts with magma. My suspicion is that nothing particular would have happened if the hurricane had happened while the eruption was still going on. It’s a lot of rain, but dumping that much rain on a channel of lava like that, you’re going to get a lot of steam. [laughter] But probably not much else. That’s a guess.

- Hi. I was a little confused on what caused the eruption of the smoke. You were mentioning, you know, is it a reaction to the groundwater? Or was it gas? Were you mentioning that? Or – I was a little confused there.

- Yeah. You’re probably confused because we’re probably confused. [laughter] It’s a good question. So before this event, we would have said that explosive eruptions like this at the summit happen as water interacts with magma. So far, we’ve looked at the data from this eruption, and there’s no evidence that water was involved. This is still a preliminary analysis. So what we think is it was probably magmatic volatiles. So magma has dissolved volatiles like carbon and sulfur in it. As the magma rises up, as it depressurizes, that creates bubbles. The analogy that’s usually used is, like, you know, uncapping a bottle of soda. So those can be explosive as well. So, as the – and the exact mechanism, I think we don’t know, but as the vent was collapsing, it could cause a little bit of pressurization. It could cause bubble nucleation. That is, bubbles grow and cause these explosions, which are triggered entirely by bubble growth and nucleation of, really, volatiles into magma and not by groundwater at all. So that’s our preliminary thought right now, but it’s early.

- Hi. Maybe I missed this also, but the rapid draining of the magma chamber, what was the underlying mechanism that led to that? Do you have a good feel for what happened? What that occurred?

- Yeah. That’s one thing I think we do understand really well. So it’s just like a balloon. So this balloon has an inlet and an outlet. And the inlet is magma coming up from the mantle. And that’s probably 2 or 3 cubic meters a second. That’s the long-term average rate.

- Fairly constant.

- We don’t know that it’s changed, so we assume that it hasn’t. The output rate to the Lowest East Rift Zone was, like, 150 cubic meters a second. So that difference was enormous. So it was just like letting air very rapidly out of a balloon, and the balloon deflated.

- Yeah. My question is what caused the change that increased the outflow?

- So – right, so the flow rate from the summit is caused by the pressure gradient. And a lot of it is because the Lower East Rift eruption – Leilani Estates – is actually at a much lower elevation than Pu‘u‘O‘o. It’s hundreds of meters lower. So it’s easier to drive magma to a lower elevation vent. And that’s been seen in the past. There’s actually a direct correlation between the height of vents along Kilauea’s East Rift Zone and the degree to which the summit deflates. So the – in fact, the analogy that was used is it’s like draining water from a bucket. The lower you poke a hole in the side of the bucket, the further the summit lava drains.

- Right. But the vents that were opened up down there, what is the mechanism that causes that to happen?

- Why did they open in the first place?

- Yeah. Down in the lower elevation.

- Yeah, so it’s a great question. There’s a complicated interplay of tectonics and magma. So there’s evidence that intrusions of magma in Kilauea’s East Rift Zone are semi-periodic, actually.

- Hm.

- So the south flank of Kilauea is actually moving steadily, probably gravitationally, and there may be semi-steady rates of magma storage in it as well. But either way, it seems to prime itself for something. So pressure in that zone in which magma tends to intrude is decreasing over time to a point where it just becomes easy to send a bunch of magma down the rift. And so there’s really some new evidence that that is semi-periodic. That evidence is more from the middle and upper parts of the rift. There’s not a lot of evidence for the Lower East Rift because these things happen so rarely.

- So infrequently.

- But it may be the same process. So it’s a great question, fundamentally.

- Thank you. Money well spent, by the way. USGS. Love you guys.

- [laughs] Thank you. [Applause]

- Thank you for your talk, and my question is about short-term climate effect. So historically, huge eruptions like this always cause certain kind of short-term local climate, like, change – or, not exactly like that, but, like, it impact the climate of a certain area.

- Yeah.

- Or even a local area. Sometimes even, like, someplace pretty far away from the eruption site. And is there any prediction on that?

- Yeah, so to the best of my understanding, the gas emissions from this eruption were not large enough to make any fundamental – certainly any global climate change or local. So climate effects, I believe, are minimal to none. But there were definitely very localized effects. So the gas and the fume that was coming off of Fissure 8 actually caused nucleation of clouds. They called these pyrocumulus clouds. So there were some very tremendous localized thunderstorms and rain and things right in that area that were caused by gases coming off of the lava flow and the vent. But that’s very localized, and that’s all that I know of.

- Thank you.

- Thank you.

- All right, so …

- One more question. I may have missed it, but the volume of the caldera collapse, how did that relate in volume to the erupted magma out of all the fissures?

- Yeah, that’s a great question, and I should have made that a little more clear. So we can build these elevation models at the summit, and we can get a really good number on that. It’s really hard to say how much actually erupted because a lot of it went into the ocean. So how much got erupted? When it’s on land, we can go out there, and we can – we can map it. When it’s in the ocean, that’s harder. So there’s work being done right now using sonar and other techniques to try to see how much was erupted under the water. And I think when those numbers come in, we’ll have a better idea. What we do know right now is that the eruption rates that were estimated from those channel velocities were pretty consistent with the rates of magma withdrawal from the summit. So my guess is that, at the end of the day, they’ll end up being pretty similar. So most of what left the summit was erupted in Leilani Estates.

- All right. I’m going to call the end of questions right now. I think a lot of you have been sitting for a while. [Applause] You guys are way ahead of me. I was going to say, thank you to Kyle Anderson for a wonderful talk. Thank you for joining us. And I did want to point out one thing, that there are wonderful resources online that the USGS Hawaiian Volcano Observatory has a website, and most of these photos and animations and things that Kyle has shown you are online for you for free – or, not free. Your tax dollars already paid for them. [laughter] But you can download them freely. And so go to – here’s the actual – that website. But if you just Google HVO or Hawaiian Kilauea or whatever, you’ll find it pretty quickly. There’s a photo chronology. There’s daily status updates. There are maps that were produced every day. And they just kind of scroll down, but there’s tons of good information online that you can help yourself to. So thank you for joining us, and join us again next month. We’ll talk about earthquakes. [Applause] [inaudible background conversations] [Silence]