EARTH 520
Plate Tectonics and People

Hawaiian Volcano Monitoring

Print Print

The Volcanoes Exploration Program at Pu'u O'o

VEPP was a project funded by NASA, implemented by USGS, and supported by SOEST at the University of Hawai‘i at Mānoa. The project lasted for about 5 years from 2010-2015. The project goal was to make available a subset of volcano monitoring data from the Hawaiian Volcano Observatory (HVO) for educational purposes. The Web site—with data streams from the Hawaiian Volcano Observatory—provided a means for students to experience and interpret current volcano-monitoring data.

The blurb above was cribbed directly from the now-sunsetted Volcanoes Exploration Program at Pu'u O'o web site and explains the point of the project. Even though the project is over we are still able to use its data for the second part of the problem set in this lesson.

The data provided by the VEPP program comes from the Pu'u O'o vent, which is on the east flank of Kilauea, so it represents just a subset of the many monitoring activities of the Hawaiian Volcano Observatory. See the map below to find the location of Pu'u O'o. It is one of the vents in the East Rift zone, the source of most of Kilauea's recent lava flows.

map of Hawaii's East Rift Zone
Map of Kilauea volcano on the island of Hawaii. The Pu'u O'o vent is the easternmost crater in an east-west trending line of craters southeast of the main summit of Kilauea.
Figure source: usgs.gov

Instrumention

There are a variety of different instruments installed around Pu'u O'o to track what the volcano is doing in real time. The general purpose of these measurements is to alert the scientists who receive the data about whether an eruption is imminent. This goal is worthy enough, but there is also the overarching scientific goal of getting a better understanding of how the plumbing system of a volcano works in the short-term and in the long-term. Here is a list of the types of instruments that record the data we'll explore, and a brief description of each one.

Tiltmeters

Tiltmeters are installed to measure any change in the angle of the ground with respect to horizontal. The goal is to make an observation of any bulge or inflation of the ground around a vent, which could be indicative of rising magma. Once an eruption has begun, tiltmeters can also measure deflation as the source of the eruption escapes from the ground. Below is a map of the three tiltmeters installed at Pu'u O'o. The Hawaiin Volcano Observatory web site also has some detailed information about tiltmeters and what their data look like.

locations of VALVE3 tiltmeters, Hawaii
Map of tiltmeter stations installed near the Pu'u O'o vent on Kilauea. Different map colors represent different recent lava flow fields.
Figure source: usgs.gov

GPS stations

Pu'u O'o has a network of seven GPS stations (see map below) that continuously monitor the location of each station to within several millimeters by recording signals from orbiting satellites. You will recall from the New Madrid lesson in Earth 501, that GPS stations are routinely used to monitor the movement across faults and to confirm the magnitude and direction of plate motions inferred from magnetic sea-floor anomalies. At active volcanoes, GPS data is used to keep track of any surface deformation of the volcano. The HVO web site also has some more detailed information about GPS stations and some examples of data from them.

Locations of VALVE3 GPS stations, Hawaii
Map of GPS stations installed near the Pu'u O'o vent on Kilauea. Different map colors represent different recent lava flow fields.
Figure source: usgs.gov

Seismometers

Pu'u O'u has two seismometers, located in the same place (see map below), but measuring different frequency bands. One of them is a "short-period" seismometer, which is sensitive to higher frequencies, such as the harmonic tremor that is often associated with the movement of magma at active volcanoes. The other seismometer is a "broadband" instrument, which is sensitive to a wider range of frequencies than the short-period instrument, and to generally lower frequencies. Broadband seismometers are installed all over the world to monitor ordinary earthquakes, but short-period seismometers are usually only for measuring what is going on in a small locality. If you want to see recent earthquakes recorded in Hawaii, you can check out HVO's seimicity map.

location of the seismometers, Hawaii
Location of the co-located short-period and broadband seimometers near Pu'u O'o.
Figure source: usgs.gov

Webcams

Pu'u O'o has two webcams that each take a snapshot every few minutes. One of them is mounted on the north rim of the crater and looks southward into the crater. The other one is installed on the southeast flank of Pu'u O'o. It looks east across the currently active vent.

location of webcams monitoring Pu'u O'o, Hawaii
Map of the two webcams at Pu'u O'o. Arrows indicate direction of view.
Figure source: usgs.gov

An example of an Inflation-Deflation event recorded at Pu'u O'o and how I made these plots

The three plots below show seismic (top), tilt (middle), and GPS (bottom) instrument records for 17-23 March 2009 at Pu'u O'o. I found this event by searching month-by-month using the software package the HVO scientists use, which used to be available to educators when the VEPP project was going on. I looked for distinctive jumps in the seismic amplitude data. Once I found one, I looked to see if it correlated with a tilt and GPS signal. Watch the three videos below to hear my explanation of what is going on in these plots because these plots tell us about the behavior of the volcano's inner workings.

Video #1

Transcript of Video 1

Here is the blank start page for the VALVE interface.

I found it easier when I was searching for possible inflation/deflation events to start with the seismic data. So you click this little thing here and then over here you can choose which station to search for. I am going to use station STC and the short-period data from that station. I am going to leave everything else over here alone. To make the exact plot I made, these are the time periods. It starts 2009 March 17 and ended on 2009 March 23. Then I click submit and a little plot pops up. The x axis of this plot is days. HST stands for Hawaii Standard Time. It does not say the year but we know what year it is because we typed 2009 in right here. Over here is basically a sort of arbitrary amplitude scale.

What this plot is showing you is recordings of the background seismic amplitude. This is different from a normal seismogram where you would actually see waveform arrivals of individual events. This is just kind of telling you what the background level of chatter is. You can think of it that way. So the background level is basically nothing, but there is the occasional big excursion which could be anything, such as cultural noise or a sudden windstorm or a helicopter landing nearby, things like that. You can ignore those because they don't have anything to do with the volcano. When you see a sustained higher amplitude thing in the data, then you might start to think something is actually going on with the volcano. It is a good idea to narrow in on this little thing and then check the tilt data and the GPS data to see if there is a real event happening there recorded by those instruments, too.

Video #2

Transcript of Video 2

Now that we have the seismic data, let us check the tilt data to see if we can find this event there, too. The way I am going to do that is I am going to click Tilt. I am going to choose this station. I am going to choose the Radial and Tangential components. And I am also going to choose to see rainfall because we know that sometimes when there is a lot of rainfall that can lead to a spurious signal in the tilt data, so we want to make sure there is no rainfall correlated with this thing we are looking for. The start time and the end time are the same as what I had in there before so all I have to do is click Submit here. Up pops the tilt time series data for the same time period as the seismic data that we looked at before. Let us scroll down so we can see both of them at the same time. This plot has time in days on the x axis again and on the y axis is tilt in micro radians. On the y axis on the right side is rainfall in millimeters. There are three different lines plotted here so it is a little more complicated than the seismic data. The black data is the rainfall data. That means we have to use the y axis that is over on the right side to see what is going on. The blue and the green are the two different components of tilt for the same station. They are oriented with respect to 360 degrees. Let us check what is going on. I can see that this blue line is showing deflation, so that is negative motion, that is towards the caldera, down to about this time. Interestingly enough, right here is where inflation begins and that inflation begins exactly at the same time as the seismic signal. The inflation that ends with this little peak corresponds very well in time to the seismic signal. I can also see that there is not a big rainfall event associated with this. I think we can convince ourselves that rainfall is not causing this. There is something going on at the volcano that is causing both a tilt and a seismic signature at the same time. That is pretty cool.

Video #3

Transcript of Video 3

Let us see if the GPS data confirms our suspicions that there is something interesting going on that we already saw in the seismic and the tilt data. I am going to select GPS here. The first thing I am going to do is set the baseline with this station, MKPM. MKPM is a little farther away from the rest of the network stations and the point of setting a baseline is to zero out the effects of plate tectonic motion and stuff like that. That question is interesting but it is not what we are trying to look for here. The actual station I want to look at is this PUOC station because that one is the closest one to the tilt meter we looked at before. I am going to leave the start and the end time the same. Clicking submit gives me this plot right here. This is the baseline station subtracted from the station of interest. This is a three-panel plot with time on the x axis of all of them and then meters on the y axis. The three components of this station are east, north, and up. Those are referenced to some arbitrary benchmark somewhere else. One thing you can see right away from the GPS data is that it is only recorded once per day. You are not going to get anything like the fine-scale wiggles and bumps that we got with the tilt data and the seismic data where we were saying ooh here is a little peak and ooh here is this. All you can do is get a general sense of whether there was motion within reason or not. What we are looking for is something that happened between the 18th and the 19th. I guess I could convince myself that there is something going on between this dot and this dot here. There is certainly not zero motion. I would say the GPS data is not inconsistent with the event we found. So, we found an inflation-deflation event! Cool!

Animations of activity at Pu'u O'o

Watch these two animations below and take note of how the physical state of the volcano is correlated to what the instrument record shows. In your problem set you will have to use real data to tell the story of a real eruptive event at Pu'u O'o, so it will be good to have these cartoons in mind.

The first animation is of an inflation-deflation event at Pu'u O'o, created by Reid Townson with support from the VEPP project. As the magma chamber inflates, note what goes on with tilt, GPS, and RSAM data. What do those instruments record when the magma chamber suddenly deflates?

A second animation also created by Reid Townson under VEPP's support depicts the tilt recorded at both the summit of Kilauea and at Pu'u O'o during a lava fountaining event at Pu'u O'o. Both tiltmeters record a gradual increase followed by a sudden decrease, but they do not happen simultaneously in this case. Why not?