Seismicity of the 2018 Kīlauea Volcano eruption

Video Transcript
Download Video
Right-click and save to download

Detailed Description

The 2018 Kīlauea eruption produced unprecedented levels of seismicity in the volcano’s instrumented history. The USGS Hawaiian Volcano Observatory documented about 80,000 earthquakes during the three-month-long eruption, starting with the dramatic collapse of the Puʻu ʻŌʻō cone on April 30 and ending with the final Kīlauea summit caldera collapse event on August 5. The sequence included a magnitude-6.9 south flank earthquake, the largest for Hawaii in 45 years. HVO seismologist Brian Shiro recounts the 2018 earthquake story in this Volcano Awareness Month talk presented in Hawaiʻi Volcanoes National Park on January 28, 2020. He also describes current levels of seismicity on Hawaiian volcanoes and HVO’s ongoing efforts to improve seismic monitoring in Hawaii. USGS photo: Damage to Crater Rim Drive in Hawaiʻi Volcanoes National Park caused by the 2018 earthquakes.


Date Taken:

Length: 00:39:20

Location Taken: HI, US

Video Credits

Video production: Katherine Mulliken, Geologist, Hawaiian Volcano Observatory,


Speaker:  Brian Shiro, Seismologist, USGS Hawaiian Volcano Observatory


In 2018, we got three years of earthquakes in three months. That set the stage for what I'm going to talk about tonight. So, three years of earthquakes in three months. What did we learn from that? What have we observed? That’s what we’re here to talk about. This is some of the damage along Crater Rim Drive in the National Park.


First of all, a show of hands, who is a local resident? And who is a visitor? This is mainly for the visitors, but for all of us, it's a refresher. Here's the volcano, Kīlauea. There are different regions on the volcano. The summit, where we are now, is the site of the 10-year eruption at Halema‘uma‘u, which concluded in 2018. We have the upper East Rift Zone; the middle East Rift Zone, which is the site of the Pu‘u ‘Ō‘ō eruption, which was very long-lived from 1983 to 2018. What we're here to talk about tonight is the 2018 eruption down in Leilani, lower East Rift Zone, which occurred during the summer of that year, and the south flank, we’ll be talking about that as well. When I talk about that, it’s this whole area to the south. So that's the volcano that we're here to talk about; those are the areas.


I thought I'd start off with where we are right now. This is a typical week of earthquakes in Hawaii. This is what happened last week—400 something earthquakes on the island. The ones that are labeled are ones people felt, or at least reported felt. You can see they occur mostly on Kīlauea and Mauna Loa, with some scattered around the north part of the island as well. These are cross-sections in depth, going north-south and east-west, if you want to know what it looks like underground. So that's the typical week, 400-500 earthquakes, multiply that by four and you get about 1500 to 2000 a month.


If we go back in time in our earthquake catalog, all the way back to 1959, you can see, we’re pretty consistent. There's some ups and downs, but it always around that number, per month. But look at this. What’s going on here? This was the 2018 eruption. This is what was so unusual about that eruption. That amount of earthquakes has never been measured before in Hawaii—20 times higher than the normal rate.


Another thing ... I'll zoom in on that a little bit… here we are. Here's that eruption. Remember I said it was three years of earthquakes in three months—here they are, May, June and July mostly, also August right there. One thing to note, is that the level of earthquake activity right now, as measured by the counts per month, is twice as much as it was before the eruption. So, the volcano is still settling, it's still adjusting, and that’s producing earthquakes still to this day.


There are some questions I want to pose, and hopefully will answer by the end of the talk. Why is that plot there? Why is that seismicity rate so high? That's one of the questions. What kind of processes made it so high? We'll try to answer that. Why is it still elevated today, which I alluded to already?


If we plot the earthquakes on a map from the eruption... these are just the ones during the first few weeks, magnitude-2.7 and higher… you can see this interesting pattern here, which we'll get to a little bit later. There are three main sequences, or groupings, of these earthquakes.


The first one, the lower East Rift Zone, or LERZ, grouping occurred down here during the first two weeks or so of the eruption. You can see earthquakes per week plotted, and you can see mostly these three weeks is when the story was going on right there, and we'll talk about why in a minute, it tailed off.


The second topic is the south flank. This is this large area. You will recall that rectangle I put there before, but it also extends offshore quite a way. You can see earthquakes showing up out there as well, related to the 6.9 earthquake.


Finally, really the big story, is the summit with all these earthquakes shown in blue, almost 50,000, depending on how you count them, at the summit during the eruption.


We'll start off with the lower East Rift first. Here's an example of the lava covering one of the roads in Leilani Estates.


What was happening before that eruption? This shows the two months before the eruption—I'll explain what's going on here. Here's the summit, here's the caldera, here we are up here. Throughout March and April, swarms of earthquakes were happening. These are groupings of earthquakes close together in space and time. Each time you see one of these, some of them are noted by these purple bars, was a swarm of earthquakes.


What was happening at the time? The volcano was engorged with lava, we were seeing inflation at the summit, inflation at Pu‘u ‘Ō‘ō, the [summit] lava lake overflowed in late April. In fact, when that happened, there was a very productive swarm of earthquakes, these shallow orange ones right here in the upper East Rift Zone connector, that little sliver on my earlier map. That really heralded a big change that was about to happen. We didn't know what was about to happen, but on April 30, it happened.


That was the collapse of Pu‘u ‘Ō‘ō. This was what the view looked like, similar to what the view looked like, from HVO on that day. You could see a pink-looking plume rise up from the edifice of Pu‘u ‘Ō‘ō.


Later it was verified that, yes, the lava lake that was there had drained away. This used to be full of lava. Where did it go?


It went down to the [lower] East Rift Zone… this is the simple view of it. It was here, it drained down and went that way.


As it did that, it produced a lot of earthquakes. This is an example of one day's worth of earthquakes observed at one station in the middle East Rift Zone, known as JOKA.  May 2, it turns out this is the day before the eruption began. This is an example of a swarm, a lot of earthquakes close together in time.


The idea here, if you look at this cartoon, is that there is a finger or blade of magma underground.  You can think of it, is called a dike, and it's pushing its way through cracks and forcing its way to make bigger cracks to open up new space. Each time it does that, it makes an earthquake. Earthquakes happen when rock breaks. That's happening right that tip, right at the front, and basically, we can map out where the front of that dike is as it moves and that’s we were doing.


However, we knew it was going down to the lower East Rift Zone, where, it turns out, we really didn't have a lot of coverage of our stations to monitor that region. We only had two stations there before the eruption—these two. So, we scrambled, and went out there in the first week and put in more. This is one of the ones that I helped install. We dig a hole and put the instrument in the ground. You can see, we started off with that, and we put in all these. We had some help with some of these. We had a pretty good network after that, in order to accurately locate and characterize the earthquakes. Then the lava came and took out four of those ultimately. But while they were there, they provided very important data to us, and we actually rescued the equipment at two of those sites right before the lava got them.


From that, we had a good catalog of earthquakes that helped characterize where that dike was at any given time, and if there was going to be changes in the eruption. This is a video put together by our colleagues over at PTWC [Pacific Tsunami Warning Center] using our data. That's one of the swarms that happened in mid-April that I talked about earlier under the summit. You can see these were going on, keeping us quite busy even before the eruption started. These are the ones in that upper East Rift connector that really were right before the eruption began, before Pu‘u ‘Ō‘ō collapsed. Here we go... there it is. Notice them moving east. That's the 6.9 earthquake, which we'll talk about in a minute. This is the beginning of the eruption and what the earthquakes are doing. Just notice how they moved east, they kind of stopped here, and went more east in two pulses.


One thing that's really interesting… one of the interesting science results that’s come out of this, based on geochemistry, is that there were different phases to the eruption. The first lava that came out was thick and viscous and sticky. We interpret that as being older. Then, later, once that was flushed out, fresher, hotter, more fluid magma came out. We see it seismically with earthquakes, this change, as well.


This plot is… I’ll explain it here… this is showing the progression of earthquakes from Pu‘u ‘Ō‘ō in the west, down here, all the way to Leilani in the east, up here.  So, it’s southwest, northeast. If you look at the slope of this… this is distance versus time here. You can't quite see, it's cut off,  but this is May 1, May 2, May 3. So, over a period of about two days, it travels some 20 km [12 mi] or so, this is about the speed you could walk, just a comfortable walking pace. Now we know how fast that dike was moving underground just by plotting this up. But notice how this line is pretty steep and then it rolled over and got flat. That's because the dike decided to stop where it was. It stayed right there in this spot, which happens to be right there, under the neighborhood Leilani and surrounding area. Then on May 3, that's when eruption began. This is one of the tools, or pieces of information, a tool seismology helped us learn. We even saw a little bit of the shallowing of earthquakes right before the eruption, where they started off a little bit deeper, they got a little bit shallower as it [magma] came to the surface.


The next part of that for about… sorry this is cut off… this is from May 4 to May 9 here in this part, no movement in terms of east or west movement. The earthquakes just stayed where they were. They were happy to be right there. They were parked there. All the early fissures were opening up. But notice this. Right there, on May 10, it pushed its way eastward more. This was that next phase of the eruption beginning, and it's really interesting. Again, just notice this really nice curve. The earthquakes were down here at this depth, 3 km [1.8 mi]. They pushed up to the surface at zero there and then they moved eastward.


Putting it all together, remember I had pointed out earlier that the first part of this eruption, the first sequence, really was about 2-3 weeks long. Here is the main story there: there was that first pulse I told you about, it settled down for about a week and made that second advance that we just talked about in the previous slide, it settled down and stayed there for the rest of time, so it didn’t go more east. At the time, we didn't know… it might, it could have … but it didn’t. These are those two general areas of those two eruptions. You can make some interpretations about that as well.


The end of the story is that seismically we saw a first pulse of earthquakes, this first phase correlated with the geochemistry, followed by a brief pause in seismicity, or a relative pause. The second phase came up next. This is when it pushed eastward more, and the magma eventually got hotter. Fissure 8 took over toward the end of this, and the rest is history. One thing that’s really interesting is that right about May 18, the seismicity rate drops off. That's about the time that fissure 8 is going to take over. The path has been cut, is the interpretation. The conduit is open. The rocks don't need to be broken anymore. So, the earthquakes died down, but the tremor doesn't die down. If you recall this cartoon from earlier… at the front of that dike is where the earthquakes are happening. But behind it, in that conduit, the magma is resonating and sloshing around. This creates this tremor phenomenon, which is a lower frequency—kind of hum—that that occurs, which we can also track. Here’s a map showing tremor.


One of the science results from that, which Matt Patrick talked about here a couple of weeks ago, was, we could correlate that tremor to changes in the lava coming out at fissure 8, the lava level, as well as the lava effusion rate, how fast it moves. Also, the temperature has been correlated with it. Each time this goes up and down, the lava level goes up and down, so they match pretty well. That's one of the interesting results from that.


That was the first sequence of the eruption, and it kept us very busy for the first half to the first two-thirds of May. Right in the middle of that was the magnitude-6.9 earthquake on May 4. Who felt that? OK… I don’t have to tell you about it then. This was the shaking pattern, the ShakeMap, from that earthquake.


This is another view of that same information, the ShakeMap. You can see it was most intense down in the Kalapana area, lower Puna, and less so the more up the chain you go, up the island. The earthquake occurred here. You saw this earlier, where those aftershocks made this circular pattern offshore. The interpretation of that is… this fault area, this really low-angle, almost horizontal fault between the oceanic plate below and the volcanic material above was ruptured. Basically, this whole thing was mobile, producing earthquakes, and still is to a certain degree today. We know that the earthquake occurred here on this part fault versus, say, up here where the fault is steeper, because with our seismic data, we can look at the way the seismic waves radiate out. We can tell that it was a very low-angle, almost horizontal, fault from that information. That’s what this shows—we call those “beach balls,” because they look like beach balls. But that’s one of the reasons we know this occurred along that fault—the biggest earthquakes in Hawaii all occur along this fault.


Why did this happen? The eruption, in short, is why it happened. When the magma was withdrawn from the upper part of the East Rift Zone, and the middle part, it went to the lower East Rift Zone. When it did that, it had to wedge its way in, and when it forced its way in, it put sideways pressure on the rock. What this is showing is a net loss of material here in the middle East Rift Zone and a net gain of material down the lower East Rift Zone from one of the models that's come out of this.


Another view of that, a sideways view if you could just slice into the volcano, as that magma was going in, it was pushing sideways. That pushed the whole south flank to the south. It might have been ready to go anyway, but this was the push it needed to get started, and it moved along that almost-horizontal fault down there.


Seismologists have taken this data, as well as GPS data, even tsunami data other data, and have created models of how the fault ruptured. These are just two examples from the literature. All the hot colors, the red, orange, yellow, show where most of the motion occurred along that surface. It starts off here, but kind of twists around and moves, and these two lobes, here and here, which is the interpretation right now, based on the available data. Some seismologists are arguing about the magnitude, by the way. We still think 6.9 is the best representation, but some people taking in other data think it’s 7.2.  These things sometimes change with further interpretation. We don't know if that will happen, but with the 1975 earthquake, it started off as a 7.2 and now we think it's a 7.7, for example. So, it's not unprecedented to think that.


This earthquake was felt statewide, more than 500 km [310 mi] away, all the way up on Oʻahu and even on Kauai. Was there damage? Did anybody have damage? A little bit? From this earthquake? The answer is “yes,” but, no one was talking about it, and for good reason.


This was happening… actually this part, fissure 8, hadn’t happened yet… but there was an eruption going on. People were evacuating their homes at this time. It was a very high intensity crisis going on. This is one of the news articles about the earthquake mentioning a power outage, mentioning some of its effects. I thought I'd share a couple of photos of its effects that I was able to find.


For example, this rock slide or landslide happened off the Chain of Craters Road near the end of the road.


A collapse at Pu‘u ‘Ō‘ō was triggered, one of many, but this one was ascribed to the earthquake that day. You can see that pinkish color I mentioned earlier.


These cracks in Leilani were reported by the residents as being due to the earthquake, so these supposedly opened up at the time of the earthquake, which is pretty interesting.


There was a tsunami, as well, that I mentioned.  The tsunami was observed statewide. It was 40 cm [15-16 inches] in Kapoho, about 30 cm [11-12 inches] in Hilo, all the way to about 5 cm [~2 inches] in Kauai. This is an example of some of those stations. Here's the radiation pattern for it. I’ll try to play this for a minute. This is in Hilo. This is 4-Mile Beach. [video playing] I'll stop it there.


If you watch the whole video, you see that island get covered and then exposed again over a period of five minutes or so. It’s pretty interesting. And 30 cm, about a foot or so, going up and down.


The aftershocks were plentiful and are continuing to this day, to a certain degree. You can see just by eye, these are all the larger ones, magnitude-2.7 and higher, that occurred during the first month or two after the earthquake. Again, there's that elliptical area that you can see quite clearly. This is interpreted as the toe of that area of the fault that's pushing its way seaward along that boundary between the two materials under the ocean. It spans about 435 miles [700 km].


One of the other models of slip is shown here, which also agrees roughly with where the earthquakes occurred. We think there was about 5 m [16 ft] of movement of the flank of the volcano just during the earthquake alone.


More about those aftershocks. The 6.9, of course, was the main shock of the sequence, of the series. But you don't know that when a sequence begins. So, the sequence, we can actually trace it back about two days to May 3, a day earlier let’s say. There was a 5.1, which at the time, felt quite big. Then the next day, there was a 5.7, which felt even bigger. I was driving in my car, and it shook the car so violently, that I knew something was going on. Janet [HVO geologist] was with me. We were about to fly in a helicopter to go check out Pu‘u ‘Ō‘ō, which had just collapsed. I said, “I'm going back to analyze this; I better go check this out.” An hour later, the 6.9 happened. So, it became the main shock. That was relegated to be a foreshock. A number of other ones happened after, in the magnitude-4 range. This orange line, by the way, this curved pattern, is a telltale pattern of aftershock decay. It shows how the energy is trying to equilibrate, trying to spread out and go back to normal over time. Seismologists can use that information to help forecast when other aftershocks might happen.


Speaking of which, some USGS seismologists have used the magnitude-6.9 earthquake and its aftershocks, as well as historical earthquakes, to calibrate a new product that'll be coming out soon, called the aftershock forecast product. It’s already in use around other regions of the country. It was very helpful during the Alaska earthquake that happened a couple years ago, for example. This shows just how well the model fits the data and it's a pretty good fit. Stay tuned for that.


This is just the summary of that sequence. You had the 6.9, the decay down, just kind of a nice decay down, and then… but remember I told you that the activity is elevated now compared to what it was before. See how before we were getting maybe 100 earthquakes per week, and now we're getting about 200 earthquakes per week. That's because that south flank of the volcano is still very mobile compared to what it was before. It's still going to slide. It's still settling, and it's producing more earthquakes than it would otherwise.


On to the third sequence of the eruption—the summit. This is an example of one of the ash plumes that came out in the early part of May. Here's the Jaggar Overlook. A very impressive series of events that went on there.


Here's a time-lapse… you may have seen this, but it never gets old. It's one image per day. Tens of thousands of earthquakes happen to allow those faults to move and allow that deformation to take place. It's just remarkable.


Here was the caldera before—nice circular crater. There's the parking lot.


And here it is after. It's enlarged quite a bit. There are different areas or different down-dropped areas. There’s the pit. Now there's water in this pit…we’ve heard about that in a different talk… and the parking lot has fallen in, as well as part of the road.


What were the earthquakes doing? Recall from the earlier slides, there were a lot of earthquakes in the summit, 40,000-50,000, at least, measured during the event. These were characterized by a cycle. This is showing earthquakes per hour. Earthquakes per hour, picked up in the magnitude-6.9. Settled down for a little while and then, starting on May 17, one of these plumes happened. With that was this blip on the tilt record, where it kind of raised up and then dropped back down. We didn't know exactly what it was at the time. We knew it was some kind of explosion, and this happened again and again and again. One interesting thing happened is… at first, they were kind of small in terms of energy release and in terms of numbers of earthquakes, relatively speaking, to what happened later. Then things really started to pick up. There's a story to this, that we’ve tried to tease out in the year and a half that’s happened since then. This cycle happened 62 times, and it peaked at about 150 earthquakes an hour. That's 5000 a day, or so. I know most of you felt them, and probably are glad they’re over. Also, quite surprising, it stopped very abruptly. So, the 63rd cycle was partway through, and it stopped before the collapse happened. For some reason, that last one didn't happen. We interpret these as stepwise collapses. As the magma is draining away, the caldera just slowly falls, slowly in between, and then abruptly falls during the collapse.


This is what that is showing. Basically, if a collapse just happened, it's happy, it's resting on the magma below. There's no gap there, and everything's good. This is showing the earthquake rate being low.


Over time, though, the magma is draining away—it's like pulling the plug in a bathtub. Now the floor of this part of the caldera is no longer supported. There’s a void underneath it, at least a partial void. These wiggles show that earthquakes are happening because the faults are getting stressed to try to hold this up. The strength of the rock is only so much.


Eventually they give, and a big collapse happens. It falls down, so this is lower than this thing. Now the caldera is at a lower spot. Then it’s stable for a few hours and then it happens again.


This is a day’s worth of earthquakes at one of our stations. If you watch the data on our website, you would see a similar kind of plot. You can see lots of earthquakes form right there. Then it collapses, it’s happy, is quiet for several hours… maybe you get some sleep. No more shaking. It picks up again. This is roughly a day… I'll show you in a minute how that varied. And then it goes again.


One of the really interesting observations from this is that, with these tens of thousands of earthquakes, there really were only a few dozen that were unique. By that, I mean we've seen repeating earthquakes where you can compare the shaking pattern of one earthquake with the shaking pattern of another earthquake on two different seismograms. These are different earthquakes. They have almost the exact same shaking. The interpretation is that it's really the same fault moving, just a repeat of what happened before. What this is showing… by the way, these orange lines are the 62 collapses through time, from May to August. All the red bars are the little families or clusters of these earthquakes that were similar. There was a family that started here, it picked up for a while, went off, started again and went off. It went for several days, maybe a week, before this other one took over, and this other one. You can see some of these last a week, some last a month, but they don't last forever. It's a really interesting observation.


We can also look at that same information in a depth profile. Now the colors show each of the different families. This red color shows … that cluster or family only occurred at this depth, maybe half a kilometer [0.3 mi] depth. This purple one is right at 1 km [0.6 mi] depth. These are those repeating earthquakes. They're repeating, because it's the same fault structure failing the same way over and over again, until the geometry of the crater changes so much, that can't happen anymore. We see it's jumbled in here because there's a lot going on during that part of the collapse. That's one of the interesting results that’s come out of this.


This is that same data—same colors, in fact—showing that after you carefully relocate all those earthquakes, most of them… if you look on our website, it’ll look like a buckshot on the caldera. But when you carefully relocate them, they mostly occur along that boundary area, where the down-dropped area happened. This little animation showing with time how they bounced around different areas of the caldera. Earlier on, they were in the west, later they were in the east. Roughly what we see. The interpretation there again is that this piston, or this block, once it was formed, it kind of slipped some to the west and then it wiggled and slipped more in the east and then eventually slipped even more in the east. I’ll have a cartoon in a minute that will show that.


When every collapse happened, as you may know, they didn’t feel like normal earthquakes. They felt like you were on a ship, like you were rocking like this. Anyone remember that? I was at Cooper Center once and that happened, everyone cheered. They thought it was fun, I guess. Better there than in the Park, where it was stronger.


What this is showing is… by the way, each one of these lines is one of those 62 collapse events. This is May. This is August. I put a square on these because don't these look a little different than those? Those first ones were really long-period, as we say. They didn't have much of a jolt or the high frequency shaking. In fact, some people couldn't even feel them. They were very subtle. In fact, that comes out when… we can compute different kinds of magnitudes. There's not only one magnitude, there's different kinds. If we compare a long-period magnitude to a short-period, high frequencies, low frequencies, all those plot down here. So, they are like this. They’re self-similar. They're scattered, but they're all in this range.


What's interesting is, all the other ones that happened, starting from May 29 onward, are all up here. Much tightly clustered by the way. So two different processes are going on with these types of earthquakes. By the way, those later ones, from May 29 on, were the ones people felt a lot more strongly and widely. These are the 5.3s that everybody talks about. These are 4.7s. Still big, but lacking in high frequencies, and therefore didn't cause a whole lot of damage. That was one observation. If you look carefully, you can notice other stages in there as well, which we think are related to the rotation of this piston block as it goes down.


Another first-hand observation of the data is, if this is energy—“moment” is a fancy word for energy—notice how all those first events… this is from May 17-29—those first 10 events were right here, really low energy release. But something happened on May 29. It shot up. The next several events were up here. It went back down. Then it sort of settled out and all the rest were more or less the same, out here. There are basically three phases going on.


There's that first one, which is those really low-frequency ones that only some people could feel.


There were these second ones, which we think is when this … what’s called a ring fault, this circular or semi-circular fault around the caldera… was actually forming.


Then this last one. It’s formed already. It’s sliding down episodically.


The lower plot, by the way, is the time between each collapse. In the beginning, it was kind of scattered, but mostly low—10 to 40 hours is pretty big range, but lower than this. When that ring fault was forming, it took longer between events. Why was that? My idea is that it's breaking more rock. It’s having to do more work, so it takes longer to fail during this time, up to 60-something hours, in between events. Once that had happened, a nice surface has been formed. It can slide along that. It came into this nice steady pattern of about 30 hours between events, which crept up over time, but was pretty consistent. Why was it so consistent? I think the interpretation is just that the magma was draining out very consistently. This is controlled primarily by how fast magma is going away.


This is that in a nutshell. In the beginning there was a lava lake. It drained away. Once it had all drained away, we started to see collapses as the magma chamber itself started to drain away. The vent where the lava lake had been was still open. That's why the plumes were getting out. Each time a collapse happened, it would kick up a lot of material and get out. But that got blocked at some point, and once it was blocked, there were no more plumes happening. But it was still collapsing. The ring fault had formed, mostly in the west, it formed earlier. In the east, it seemed to grow over time. That's how we see it in cartoon form.


What did it do? What were the effects? This is just a selection. The Park was closed; we're glad it's open again, but the Park was closed for a long time.


Each time a collapse happened, the scene was like this. If you were at Tina's talk the first week [of January], she showed a video of this. Later I have a video that you can hear what it sounded like. I like this one because … who knows what that is? The HVO building up there getting battered.


Here's that video, which shows the effects of shaking at HVO from two different vantage points. [video]


Listen… that’s not wind. See the dust... it’s rising. The dust would just fill the scene and it would take a while to settle out.  Every 30 hours or so that would happen.


Cars in the parking lot there would bounce edge several inches. Obviously, it created cracks, it created damage. This what the sign looks like today.


Inside [HVO], it knocked things over. There are some bookcases.


Some more. You can guess whose offices these are, if you want.


It caused damage to some [National Park] trails, many trails. This is just one, one with lesser damage actually. Thankfully, many of the trails have reopened now.


Buckling in the road, from minor buckling like this—shows there was compression here—to major buckling like this.


This is a down-drop of 20 feet [6 m], maybe. Taller than you, taller than a car—certainly not drivable. This is what the scene is like all around the west side of Crater Rim Drive right now.


Sinkholes. This one was in front of the Golf Course [subdivision]. But there's sink holes in the Park, too. That’s why Kīlauea Overlook is still closed. There's one of these right in the entrance to it.


What was it doing, all that work it was doing? It was creating this, all these steps going down to where the lake is today. The parking lot used to be over here. It’s just remarkable how much it changed, how many earthquakes it took to allow that change.


I like this shot because you can see that eastward fault, which was just eating away at the old [caldera] floor as it went. If this had continued longer, it probably would have gone more east and more north, which is off this part of the scene. It exposed lots of new areas that had never been seen. This area here, for example. Geologists are going to have fun mapping these in the future.


That was the summit sequence, in a nutshell. What happened after the eruption? It had a small little PostScript there. This was in September [2018], the last little gasp of fissure 8.


What's been going on since then? We've had inflation, both in the East Rift Zone and at the summit, which is ongoing. This is the past year or so. We know magma is coming back. This is showing radar images, by the way, from space. It’s one of the ways we can monitor this.


Earthquakes. We have lots of earthquakes still occurring, but in the usual places—at the summit of Kīlauea, the summit of Mauna Loa, the south flank that I talked about, and this area—the deep Pāhala zone, the deep lower southwest rift, as it’s known.  


I show this plot down here. This is showing the last week of eruption, when some 2500 earthquakes happened, compared to now. All the variation we've had in a year and a half since then, has been pretty consistent compared to then. So, steady as she goes is what's happening now. The volcano is adjusting to its new normal, but it's not changing very rapidly, necessarily. It’s reached a new normal for now.


In summary, there were three sequences with this eruption. There was the lower East Rift sequence associated with the intrusion of that magma and the opening of the fissures and the different phases of the opening of those fissures. There was the south flank sequence dominated by the 6.9 earthquake, and that whole zone and south flank slipping on this almost horizontal fault beneath the island, and even extending far offshore. And the summit sequence, where the caldera was enlarging itself, becoming deeper and becoming what you can go see today, creating unprecedented levels seismicity never before measured here—20 times higher rate than normal. Three years of earthquakes in three months.


What about our questions?


Why is the seismicity rate so high? There was an eruption. It was the biggest eruption in 200 years, and this eruption involved both the summit and the flank. A lot was going on that made a lot of earthquakes. That's the simple answer.


What type of processes contributed to that high rate? The three we talked about: the migration of magma cutting its way down to the lower East Rift Zone; the 6.9 earthquake, and, of course, the caldera collapse.


Finally, why is the rate still a little bit higher now? It’s because the volcano is still pretty mobile. The south flank is still moving at a higher rate than usual. There are still aftershocks happening along that fault, and there will be for quite some time.


Thank you very much for your attention. I want to thank everyone who's helped with this. There are many seismologists and technicians, and people who have helped keep up with this data—we’ll be digging out of it for several years. Especially thanks to Janet for helping put this together. Mahalo.