U.S. Geological Survey

Volcano Hazards Program

This is the text-only version of the Predicting an Eruption presentation. The full presentation uses Marcromedia Flash (MX). In the text-only version below, descriptions of photographs, graphs, and animations in the Flash presentation are indented.

Predicting Volcanic Eruptions

| Home | Introduction | Instruments | Ground Deformation | Earthquakes | Exercise |
| Kilauea Slide Show | Mount St. Helens Slide Show |


"No longer shall the cities be destroyed."
Dr. Thomas A. Jagger, 1912. Founder of Hawaiian Volcano Observatory

Text version of animation of images and text:

After the opening text moves across the top of the screen (right to left) a photograph of Dr. Thomas Jagger moves from the bottom of the screen into the lower-left part of the screen. The image slowly fades to black.

Volcanoes are inspiring but their eruptions can be lethal.

Text version of animation:

A photograph of Soufriere Hills volcano appears in the right center of the screen. The ash-covered city of Plymouth, Montserrat, is in foreground. Volcanic ash billows from the volcano and is moving toward the abandoned city.

Millions of people live within the shadow of active volcanoes.

Text version of animation:

A photograph of Mount Hood, Oregon, dissolves into the photograph of Sourfriere Hills. The snow covered voclano looms over the bustling city of Portland, Oregon. Leaves of trees in the foreground are turning yellow and red because of the change in season from summer to fall.

Scientists can provide warnings of eruptions.

Text version of animation:

A photograph of a scientist working on a volcano dissolves into the photograph of Mount Hood, Oregon. The scientist is surveying benchmarks located on South Sister volcano, Oregon, which can be seen in the distance. A helicopter used to collect temperature and pressure data along the sight lines of the survey is in the photograph.

This photograph fades to black and three small versions of the photographs of the volcanoes described above appear to the right of the text.

(end of Flash page 1, Introduction)
| back to top |


Richard Treves and Steve Brantley

Richard Treves

Flash Programming
Richard Treves

Thanks to:

Staff of U.S. Geological Survey's Hawaiian Volcano Observatory, including Paul Okubo, Peter Cervelli, Steve Brantley, and Don Swanson. Also, thanks to National Park Service volunteers Bryce and Jan for writing and technical advice.


"Predicting Volcanic Eruptions" is in the public domain. Images and text in this presentation may be used freely; please credit the U.S. Geological Survey and Richard Treves. Please let us know how you are using this resource: richard@lavalog.co.uk and vhpweb@usgs.gov

(end of credits, Introduction)
| back to top |


| Introduction | Thomas A. Jaggar | Seismometers | Tiltmeters | To Do | Kids To Do


Lava shooting high into the air or the radiant heat of molten lava burn forever in the memories of those who witness Hawai`i's active volcanoes. Few places on earth allow closer or more inspirational views of eruptions. For scientists, Hawai`i's active volcanoes are windows through which one can study the forces of nature that have forged their mighty cones and sent lava flows into ancient and modern-day communities.

Forecasting when a volcano will erupt can save lives. In this section, you will learn about two methods used by scientists to track a volcano's activity and make forecasts: earthquakes and ground tilt. You'll see the types of instruments used in the early days of the world's second oldest volcano observatory, the Hawaiian Volcano Observatory. Although the methods have since improved, the basic strategy is the same as when Dr. Thomas A. Jaggar established the observatory on Kilauea Volcano in 1912.

Thomas A Jaggar

Jaggar studied the activity of volcanoes after, but not before, they had erupted. He soon recognized the need for scientists to measure make continuous on-site observations and measurements of volcanoes before, during, and after they became active. Only careful observations with the latest instruments would help scientists to forecast dangerous volcanic activity.

(end of Flash page 1, Instruments)

So committed was Jaggar to this idea that he quit his post as the head of the geology department at Massachusetts Institute of Technology and moved to the summit of Kilauea Volcano. Kilauea was ideal for a new volcano observatory. It had a molten lava lake erupting at its summit and scientists could make measurements on the volcano in relative safety. But the key to the observatory's success was Jaggar's careful recording of his observations.


Jaggar bought the latest seismometers from Japan and Germany to study Kilauea's earthquakes, and quickly discovered that Kilauea's numerous earthquakes were related to volcanic activity.

The instruments were placed in the Whitney Laboratory of Seismology beneath the observatory's office. Unfortunately, steam from the surrounding rocks made the room uncomfortably hot for Jaggar and his colleagues!

One of these early seismometers was made with a heavy weight suspended on a wire. The weight was attached to a pen and a drum as shown in the figure. When an earthquake occurred, the instrument stand and rotating drum moved together side to side while the mass and pen remained still.

(end of Flash page 2, Instruments)

Replacement for animation illustrating seismometer:

The pen drew out a line on the paper covered drum, an earthquake wave would cause large peaks and troughs on the line, a flat line indicated a quiet period.


Soon after the first seismometer started recording earthquakes, the observatory's seismologist noticed something unusual: the earthquake trace didn't follow its usual track. Scientists recognized that this was caused by tilting of the ground.? Like earthquakes, the observatory team noted a link between ground tilt and the volcanic activity of Kilauea. They installed a tiltmeter in the Whitney Laboratory and recorded the change in the earth's surface every day. The figure to the right shows this early tiltmeter. Modern instruments are electronic, and can measure tiny angles equal to the tilt produced by putting a dime (1 mm thick) under a bar 1 km long.?

Replacement for figure of the Clinoscope:

Now generally called 'Tiltmeters' an early instrument at the Hawaiian Volcanoes Observatory was known as a 'Clinoscope'. A heavy weight was suspended with piano wire from a tripod about 2 m tall. The weight was fixed with a rod and sharp point. As the ground tilted, the pointer remained still but a piece of paper with points of the compass marked would move as the frame moved. Scientists would record the relative movement of the point every day.

(end of Flash page 3, Instruments)

To do:

Be a Seismometer

Take a piece of string and tie a heavy weight to it such as a apple. Move your hand forward and backwards while trying to keep the weight in one place. You are acting much like Jaggar's seismometer would have; your arm and hand are the stand, the backwards and forwards motion is like the shaking an earthquake would have produced and the apple is like the heavy weight. The seismogram is produced by the relative motion of you hand and the apple.

See How a Volcano Produces Tilting

Take a big envelope and a balloon. Put the balloon in the envelope leaving the end of the balloon sticking out of the flap of the envelope so it can be blown into. Seal the flap of the envelope around the balloon but still leaving the end out. The envelope is now like a volcano with no lava below the surface. Blow into the balloon and watch how the surface of envelope changes as you inflate the balloon. You may like to put a spirit level on different parts of the envelope to see how the tilt varies. Try and work out where the best place to put your spirit level 'tiltmeters' would be to detect the volcano filling with lava.

If you stick a pin through the envelope to produce an 'eruption' of lava you will see a rapid change in tilt across the envelope similar to that found after a volcano has just erupted.

(end of ?To Do?, Instruments)

Kids To Do:

Draw a Picture of Jaggar's life

Take a piece of paper and split it into 6 parts as in the figure below. In each section draw a picture showing a part of Jaggar's life.

  1. Working at a University in America
  2. Sailing on a boat to come to Hawaii
  3. Driving up to the summit of Kilauea on a road with lots of potholes
  4. Jaggar looking at the lake of lava
  5. Measuring the tilt of the volcano
  6. Running away from the volcano when it erupted

(end of ?Kids To Do?, Instruments)
| back to top |

Ground Deformation

| Mount St. Helens | Dome-Building Eruptions | Magma Causes Ground Movement | Thrust Fault | Tiltmeter | Earthquakes | Conclusion | To Do | Kids To Do |

Mount St Helens Volcano

On May 18, 1980 Mount St. Helen's erupted explosively (discussed as a case study later). This led to the decision to create the Cascades Volcano Observatory so that scientists could carefully study the volcano's continuing activity. Scientists quickly learned to predict the volcano's later eruptions. In this section, you will discover how the measurements that the scientists took were used to predict most of the 21 eruptions that occurred through 1986.

Text version of images showing Mount St. Helens erupting:

When Mount St. Helens erupted on 18 May 1980 it sent a hot column of volcanic ash, gases, and pumice to about 22 km above sea level in less than 10 minutes. It also caused a landslide that filled 22 km of the North Fork Toutle River valley with rock debris to an average depth of 45 m. In places the deposit is 195 m thick!

Dome-Building Eruptions

Most of the eruptions produced very pasty lava flows (right image) that spread slowly across the ground (less than 10 m per hour). These lava flows were about 300 m in diameter and as thick as 50 m. Each new lava flow erupted on the previous solidified flows. In this way, the flows built a very large mound shaped feature called a volcanic dome.

Text version of images showing growth of dome:

In October 1980 glowing lava (870 degrees C) erupted and spread for several days, forming a circular dome 300 m in diameter.

Before each eruption, magma (molten rock below ground) rose into the core of the dome. The rising magma caused the crater floor to split, forming new faults, and to tilt outward away from the dome. Pressure from the rising magma also triggered earthquakes. Scientists used the faults, ground tilt, and earthquakes to predict nearly all of the dome-building eruptions.

Text version of images showing dome's lava flows or lobes:

By September 1981 the dome was 620 m wide and 165 m tall. A large rock (nicknamed 'Federal building') some 13 m tall fell from the dome.

(end of Flash page 1, Ground Deformation)

Magma causes Ground Movement

The animation to the left shows magma rising into the dome before erupting on top of the dome.

The movement of the thrust fault (left) and amount of tilting (right) increases as the molten rock gets closer to erupting on top of the dome. This pattern was the key to predicting eruptions at Mount St. Helens.

Text version of animation showing the lava dome erupting:

A cross section through a dome is shown. A batch of magma is shown rising towards the top of the dome. As it rises the right-hand side of the dome moves outward because of the pressure exerted by the magma from the inside. Similarly, upward pressure exerted by the rising the magma pushes on the rock in the crater floor to create a fault at the left base of the dome. The fault, called a thrust fault, moves as the magma rises to the top of the dome.

Thrust Fault on Crater Floor

Many thrust faults formed on the crater floor before eruptions at Mount St. Helens. Rising magma pushed (or thrusted) the upper part of thrust fault over the lower part. Scientists flew by helicopter into the crater every few days to measure the movement on the thrust faults with a tape measure.

When a new thrust fault appeared, metal stakes were pounded into the crater floor on both sides of the fault. As the fault moved the distance between the stakes shortened.

Text version of image showing scientists measuring movement on a thrust fault:

Two scientists are using a measuring tape to measure the distance between two metal posts located on both sides of a thrust fault. One post is located on top of the moving thrust fault (left side of image) and the other is located on the stable part of the crater floor. Arrows on the image indicate the direction of movement of the thrust fault.

(end of Flash page 2, Ground Deformation)

Thrust Fault Animation and Plot

A closer look at the thrust faults that ruptured the crater floor shows the pattern of movement that scientists recognized before eruptions at Mount St. Helens.

At the observatory, the distance measurements across the faults were plotted on a graph (top panel). Distance in centimeters is plotted on the x axis and time in days plotted on the y axis.

Text version of thrust fault plot:

A graph is shown, time is on the x axis, distance is on the y axis. The line starts in the top left hand corner; it is horizontal with a slight downwards dip. This shape illustrates that the thrust fault is moving so that the distance between the metal stakes is becoming less. Towards the eruption at the end of the plot the line dips sharply downwards, illustrating that the rate of movement (in centimeters) increases dramatically just before the eruption takes place.

Because the thrust faults moved away from the dome, the distance between the metal stakes became shorter before eruptions. By connecting all the data points, scientists quickly recognized a pattern in the shape of the line: there was a sharp downturn days to weeks before an eruption. The sudden downturn reflects the fact that the stakes were moving towards each other more quickly. Scientists noticed this pattern recurring many times before the dome building eruptions at Mount St. Helens.

Often scientists made a long term prediction (1 to 3 weeks before eruption) followed by a shorter prediction when they had more measurements. The second prediction was more precise (sometimes predicting an eruption in 24 hours).

Text version of the two animations showing movement of a thrust fault:

The animation is similar to the earlier animation showing the plume rising in the dome. The plume rises and just before the time of the eruption its motion is seen to increase in rate (accelerates). On the left hand side of the dome a ?slice? of rock is seen to be pushed outward by the pressure exerted from the magma inside the dome. The movement of this rock is upwards and to the left. It moves with a fairly constant rate until the magma accelerates at which time its own rate of movement accelerates.

(end of Flash page 3, Ground Deformation)


In addition to measuring thrust faults, scientists installed several electronic tiltmeters on the crater floor to record the slightest change in angle of the ground before an eruption. They discovered that pressure from the molten rock rising into the dome caused the crater floor to tilt outward away from the dome.

Text version of tiltmeter plot:

A graph is shown, time is on the x axis, distance is on the y axis. The line starts in the bottom left hand corner; it is nearly horizontal with a slight upward angle. This nearly horizontal line indicates the crater floor is tilting only slightly at the tiltimeter location. As the day of the eruption nears, however, the line begins to curve upward. A sharp upward turn indicates the crater floor tilts at an increasing rate just before the eruption takes place.

A plot of tiltmeter data (angle versus time) shows a sharp increase in tilt before an eruption. Note the sudden upturn in the data plot occurs just before the molten rock reaches the surface.

Text version of the two animations showing movement of a thrust fault::

The animation is similar to the earlier animation showing the magma rising in the dome. As the magma nears the top of the dome, the speed of movement along a thrust fault increases (accelerates). On the right hand side of the dome a tiltmeter measures the movement of the crater floor due to pressure exerted from the rising magma (in the animation, the tiltmeter appears to be located on the edge of the dome--in reality the tiltmeters were situated on the crater floor). The change in tilt of the ground is up and outward away from the dome. The data curve changes rapidly as the magma nears the top of the dome.

Text version of image showing a tiltmeter station at Mount St. Helens:

A tiltmeter station is shown by the side of the dome. Each tiltmeter sent data back to the observatory every ten minutes using a radio transmitter.

(end of Flash page 4, Ground Deformation)


Scientists use an instrument called a seismometer for recording ground vibrations caused by different types of earthquakes. As magma below a volcano forces its way towards the surface, it causes rocks to break and fracture, resulting in earthquakes. Scientists have learned that earthquakes tend to increase in number before an eruption, but not necessarily in size.

Text version of image showing a seismometer:

A scientist wearing a helmet and gas mask is shown manipulating a seismometer that will be installed on the lava dome at Mount St Helens. He is sitting on a rock and is surrounded by other rocks. The gas mask was required because of the high concentration of sulfur dioxide gas emitted from cracks and fumaroles on the dome.

One way to analyze earthquake data is to plot the number of earthquakes beneath a volcano every day. At Mount St. Helens, a graph of the number of earthquakes plotted on the x axis and time in days on the y axis often showed an increasing number of earthquakes just a few days before an eruption occurred, especially within the final 24 hours.

Text version of animation showing earthquake swarm before an eruption:

The same animation of magma rising in the dome is shown. Initially, a small number of earthquakes occur in the rocks around the moving magma (earthquakes are represented by circles that appear and then disappear). As magma approaches the top of the dome, many more earthquakes are triggered.

Text version of animation showing earthquake graph:

A graph is shown of time on the x axis and the number of earthquakes per day on the y. As time progresses to the day of the eruption, the line shows the daily number of earthquakes. When the line is nearly horizontal, the number of earthquakes has not changed significantly. When the line turns upward, the number of earthquakes per day has increased dramatically. Initially, the line is nearly horizontal with a slight upward angle. This steady curve indicates that the number of earthquakes per day is fairly constant but increasing a little each day. Just before the eruption, the curve swings sharply upward. This upturn indicates that the number of earthquakes per day increased dramatically just before the eruption occurred.


You have analyzed 3 types of data used by scientists to accurately predict eruptions of Mount St. Helens between 1981 and 1986: movement of thrust faults, tilt of the ground, and earthquakes. Each of the graphs showed a pattern of increasing movement and rock breaking as molten rock rose into the dome before it reached the surface.

(end of Flash page 4, Ground Deformation)

To Do:

How a Thrust Fault Works

Take a number of books and put them in a pile. Now arrange the bottom book to be at an angle. By pushing the other books across the first book you will see a motion between the books similar to thrust faulting. The bottom book will remain stationary as the upper books move over it. Your hand represents pressure of the moving magma directed against rocks (or in your case, the books) of the crater floor. The angle of the thrust faults at Mount St. Helens was between about 15 and 20 degrees.

The main difference between this demonstration and the thrust faults on the crater floor is that the leading edge of the "book" thrust fault extends out into the air. On the leading edge of "rock" thrust faults, however, the rocks crumble or break to form a small cliff or scarp. This scarp grows in size as the thrust fault continues to move. In the crater of Mount St. Helens, the leading edge of a thrust fault often grew from a few centimeters tall to more than one or two meters tall.

(end of To Do, Ground Deformation)

Kids To Do:

Build Your Own Mount St. Helens

Bury a balloon under a pile of sand with the end of the balloon left out so that the balloon can be inflated. Blow into the balloon slowly to see the 'volcano' moving. You may have to remove some of the sand to do this successfully.

Before the volcano finally collapses you may notice sand slipping away from the balloon. In real life this was seen on Mount St. Helens, many of the eruptions occurred with 'rockfalls' where looser rocks slid off the dome before the eruption finally took place.

(end of Kids To Do, Ground Deformation)
| back to top |


| Eq Vibrations | Eq Simulation | Trigger an Earthquake | Locating Eq | Types of Earthquakes | Conclusion | Kids To Do |

Earthquakes cause Vibrations

Earthquakes almost always occur before volcanic eruptions. In this section, you will simulate an? earthquake and learn to identify two types of earthquakes that occur beneath volcanoes. You will also learn how scientists locate earthquakes.

Earthquake Simulation

The figure to the left is a cross section of the ground.

The red triangle marks the location of a seismometer that will detect ground vibrations caused by the earthquake you will trigger. The green cracks at the bottom represent deep earthquake faults. By clicking these on the next screen you will simulate earthquakes on the left and right of the cross section.

The circles in the cross section will move as the waves of an earthquake travel away from the faults.

Text version of the image showing a cross-section of the ground:

A cross section is shown through the ground surface down a few thousand meters (not drawn to scale). Trees and a house sketched on the ground surface help illustrate the cross section. The entire cross section is marked with a set of circles or balls that extend from the ground surface down to the bottom of the sketch. On top of the ground surface on the far right-hand side of the image is a red triangle illustrating the location of a seismometer. Above the seismometer, a graph draws the waves of an earthquake.

At the bottom of the grid on the left and right hand sides are located green fault icons. Clicking on the icons triggers fault movement, which causes earthquake waves to pass to the Earth's surface. The waves are detected by the seismometer and written on a graph that shows at the top of the screen.

(end of Flash page 1, Earthquakes)

Trigger an Earthquake

Click on one of the faults. When the seismometer stops recording the ground vibration at the surface, click on the other fault. Compare the two earthquake signals.

Text version of earthquake animation:

When the faults are clicked a series of earthquake waves move upward and across the cross section, represented by the balls that show movement progressively farther from the fault. A graph of the earthquake record appears above the grid of balls; the graph is produced with time on the x axis vs the ground movement on the y axis. The earthquake records for the two faults are different. When the lower right fault clicked the resulting record is large, and there is essentially no delay between the time of the earthquake and the arrival of the seismic wave (the movement of the vibration along the grid). When the lower left fault is clicked the seismic plot is much smaller, and there is a delay between the start of the earthquake and the arrival of the earthquake wave (a flat section on the plot showing no movement of the seismometer).

Which earthquake fault caused the greatest vibration at the seismometer?? ANSWER

Text version of answer box:

Because the earthquake fault on the right is closer to the seismometer, its earthquake signal is larger.

The earthquake wave from the right fault also reached the seismometer sooner than the one from the left. How can you tell?? ANSWER

Text version of answer box:

Compare the two recordings. Which one has a flat section before the first wave arrives? The flat section represents the time that it took for the initial earthquake wave to travel from the fault to the seismometer.

Now create your own earthquake. Click on a circle, drag it, and then release. The farther you drag a circle, the larger the earthquake.

(end of Flash page 2, Earthquakes)

Locating an Earthquake

Scientists locate earthquakes based on the time that it takes for an earthquake wave? to reach at least 3 seismometers. The animation to the left shows the wave of an earthquake moving outward in all directions away from its source. The wave front reached seismometer 1 first and seismometer 3 last.

Text version of the animation of an earthquake wave:

Three seismometers are shown around Kilauea caldera, an earthquake starts in the caldera and the edge of its wave is represented by a circle spreading out from a point. The circle reaches seismometer 1 first and seismometer 3 last. The travel times from the source of the earthquake to the seismometers (1, 2 and 3) are shown as 0.1, 0.2 and 0.5 seconds respectively.

By knowing the speed of an earthquake wave and how long it takes for the wave to reach a seismometer, the distance between the earthquake and seismometer can be calculated.

For seismometer 3:

Distance to earthquake source = Speed x Time

Distance = 6 km per second x 0.5 seconds

Distance = 3 km

Having calculated the distance between each seismometer and the source of the earthquake, it is possible to locate where the earthquake occurred. Roll mouse over the bottom image for more details of how the earthquake was located in the 1930's. The equipment is from the Hawaiian Volcano Observatory.

Text version of image of the Kilauea model used to locate earthquakes:

A historic upside down model of Kilauea Volcano is shown. Three strings are tied to three points on the upside down model of the volcano. The strings are held taut and meet at a point above the model (representing a distance below the ground surface). A rod runs between the meeting point of the strings (shown as white lines) and the surface of the model.

One end of the strings were tied to points representing the location of the seismometers (black dots). The length of each string was cut to represent the distance between the seismometer and earthquake source. The earthquake source was fixed at the point where the 3 strings were taut. The length of the rod? in the model represents the depth of the earthquake, and where the rod rests on the model marks its source location on a map.

(end of Flash page 3, Earthquakes)

Types of Earthquake

Scientists have identified several different types of earthquakes that are associated with volcanic activity. Two types of earthquake waves shown below have been converted to sound waves and compressed in time. Click on the arrows to "hear" the sound of the wave forms. For expanded view and description roll mouse over 'frequency' button.

Text version of frequency button and images of earthquake waves:

Two plots of time (x axis) versus wave height (y axis) are shown in upper left of the screen. Both plots consist of peaks and troughs that become smaller in height as time progresses, but the shapes of the plots are different. The top plot shows the trace of a long period earthquake and the bottom shows the trace of a short period earthquake. The long period trace consists of troughs and peaks that are wider or more spread out compared with the short period plot. The short period plot has large peaks and troughs initially but then the height of the peaks diminishes very quickly. The long period trace diminishes gradually over a longer time period.

Frequency is a measure of peaks per second. Long period earthquakes have peaks that are far apart (about 5 peaks per second), short period earthquakes have peaks close together (about 2 peaks per second).

'Long Period' earthquakes beneath volcanoes are thought to be caused by magma moving through rock fractures or cracks. In a similar process, air moving through an organ pipe vibrates to produce a long lasting sound.

'Short Period' earthquakes beneath volcanoes are caused by rock-breaking events, such as rising magma breaking through rocks. A similar sound wave is produced if a stick is snapped.

Both types of earthquakes often occur in groups or 'swarms' before an eruption. However, a swarm of long period events can be a stronger indicator that magma is rising toward the surface.


Scientists use a network of seismometers to record earthquakes that occur beneath volcanoes. Computers automatically analyze the timing of earthquake wave forms in order to calculate earthquake locations. Different types of earthquakes are identified in order to better forecast eruptions.

(End of ?Kids To Do?, Earthquakes)

Kids To Do:

Create a Wave in a Rope

take a rope and tie it to door handle or chair. By wiggling the other end of the rope see if you can create a wave that moves along the rope. It is best to hold the rope fairly taut and to rapidly jerk your hand. Use a longer piece of rope and see if the wave takes longer to get back to your hand as happens with earthquake waves.

Locate an Earthquake Focus

The map to the right shows an upside down map of a volcanic region which is completely flat. 3 seismometers are marked as white crosses. Copy the grid out with one square = 2cm by 2cm (or 1inch by 1inch) onto a piece of card. Cut a piece of string representing the distance of? seismometer 1 to the earthquake source 4.2cm (2.1inches) long. Repeat with 2 more pieces of string for seismometer 2 and 3, both lengths = 5.4cm (2.7 inches). Attach the string to the respective seismometer marks and by holding all the strings taut (at their loose ends) locate the position of the earthquake focus on the map. Which volcano do you think the earthquake is under?

Text version of the map:

A five by five grid is shown. Seismometers 1, 2 and 3 can be found at 1 -1 (column - row), 4 ? 1, and 3 ? 4 respectively.? Volcanoes A, B, C and D are at 1 ? 4, 2 ? 2, 3 ? 1, and 4 ? 2 respectively.

Text version of answer box:

Under B

Is the volcano focus at depth or on the surface? ANSWER

Text version of answer box:

At depth, the strings are only taut if held above the grid which (because this is an upside down model) shows the focus is at depth.

(End of ?Kids To Do?, Earthquakes)
| back to top |


| Mount St. Helens | Prediction a Success | Your Turn to Predict |

Mount St. Helens Exercise

This exercise uses the 1980s Mount St. Helens eruptions as an example of eruption prediction.

Before you predict a real eruption at Mount St. Helens (starting page 4) we will show you how scientists predicted another eruption.??

On this first page you can see data plotted up to October 21st. At this time note there is some movement of the thrust fault, the number of earthquakes is beginning to increase, and the tiltmeter is detecting a slight change in ground tilt.

Perhaps magma is beginning to rise into the dome. However, the changes in the plots are too small at this time to be certain that any eruption will occur.

Roll your mouse over the yellow items on the screen for more information.

Text version of the plots:

Three plots are shown on the left side of the screen, each showing a type of measurement (y axis) plotted against time (x axis). The top plot shows the distance that thrust faults moved (in cm); the middle plot shows the number of earthquakes beneath the volcano; and the bottom plot shows the amount of tilt on the crater floor. The plots show data from September 5th to October 20th. At this point in time (October 20), the data on the thrust fault plot shows a little movement as indicated by its steady downward dip (shortening of the distance across the thrust fault between the two metal stakes). The earthquake and tilt plots have started bottom left, and are dipping only slightly upward.

Text version of the roll over buttons:

Thrust Plot: As magma rises into the volcano the crater floor ruptures to form faults, called thrust faults. One part of the fault is pushed up and over the other part. So far this plot shows only a slight movement.

Earthquake Plot: Rising magma causes rocks to fracture and break, resulting in earthquakes. The number of earthquakes has risen but only slightly.

Tiltmeter Plot: As magma rises into the volcano the crater floor is pushed up and out from the dome. This results in a slight change in the angle of ground tilt. This plot shows a tiny but steady tilt.

Chief Scientist Comments: All of the plots suggest or indicate magma is beginning to rise into the volcano. Based on our experience with earlier eruptions it is too soon to estimate when magma might erupt at the surface. However, I am asking my team of scientists to increase their monitoring activities by making more frequent visits to the crater. I've also asked them to cancel any non-essential trips in the next month.

(end of Flash page 1, Exercise)

By November 9th all of the data plots show a significant increase in activity. Note the sudden up turn or down turn of the plots. Based on this dramatic change the team of scientists agreed that an eruption was likely to occur soon. Roll over the chief scientist comment button to see what prediction was made.

A kink occurs on the tiltmeter plot on October 20th. Rollover the button to see what a trained scientist thought of this kink.

Rollover the plot titles for a more detailed discussion of each plot (different from the previous page).

Text version of the plots and roll over buttons:

The graphs described above show the same information but with data shown through November 9. Since October 20, the thrust fault plot shows a steep turn downward. The other two plots both show steep upward turns.

Thrust Plot: The overall shape of the curve indicates that the thrust fault is now moving more rapidly (about 3 cm per week) compared to 2 weeks ago (less than 1 cm per week). This sharp down turn has been noted before, usually within days of an eruption. However, before making a prediction be sure to check the activity represented by the other data plots.

Earthquake Plot: The gradient of this graph has become steeper. This indicates the number of earthquakes per day under the volcano has increased dramatically. This sharp increase commonly occurs within a few days of an eruption. However, before making a prediction be sure to check the activity represented by the other data plots.

Tiltmeter Plot: Tilting of the crater floor that began a few weeks ago has shown a steady increase in the past few days. This is reflected in a sharp upturn of the plot. Such rapid tilting of the ground preceded all previous eruptions by a matter of days. However, before making a prediction be sure to check the activity represented by the other data plots.

A kink appears on the graph of tiltmeter plot, the plot suddenly takes a sharp downturn but the downturn is not sustained and the plot soon shifts back to a more shallow angle.

Text version of the plots and roll over buttons:

Tiltmeter Kink: This feature on the graph shows why several types of evidence need to be considered before making a prediction. It appears that the thrust fault is moving faster, but later measurements showed that the thrust fault slowed down. A prediction should only be made when several measurements are increasing dramatically at the same time. At the time of the 'kink' neither the number of earthquakes or tilt of the ground were changing very rapidly.

Chief Scientist Comments: Because all data plots now show significant activity I am confident that an eruption will occur within a few days and certainly within a week. Therefore, I am issuing a prediction to the public that an eruption will occur within one week.

(end of Flash page 2, Exercise)

Prediction is a Success

Two days after the eruption prediction the following statement was issued:

7:00 AM, 11 November

"Landslides from the lava dome, possibly accompanied by a small explosion, occurred at about 3:20 this morning... Minor snow melt occurred in the crater, but no significant mudflow extended beyond the crater."

The landslide and explosion triggered an eruption column to at least 5 km above sea level and ash fall was reported nearly 100 km south of the volcano.

Text version of the same plots described above:

The thrust fault plot shows a continued rapid downturn as the distance between the metal posts became shorter and shorter before the eruption. The number of earthquakes and amount of ground tilt increased rapidly up to the time the eruption occurred on November 11.

Text version of roll over buttons about Notes on Data:

Notes on Data: The thrust and earthquake data in this presentation was taken from data for the eruption of Mount St. Helens in March 1984. However, the tiltmeter data is from an eruption in 1982. During the March 1984 build-up time, a new type of tiltmeter was installed in the crater. Because, bad weather prevented the installation of a radio system until a few days before the eruption, the data does not show a gradually increasing trend that could be used in this exercise.

(end of Flash page 3, Exercise)

Your Turn

You have been promoted to Chief Scientist. The date is February 21 1982, and the data collected by your team of scientists is shown in the plots. Be sure to roll over the yellow boxes for additional information about the data.

Text version of plots and notes on data:

Three plots are shown on the left side of the screen, each showing the same type of measurement (y axis) plotted against time (x axis) described above. The top plot shows the distance that thrust faults moved (in cm); the middle plot shows the number of earthquakes beneath the volcano; and the bottom plot shows the amount of tilt on the crater floor. The plots show data from January 11 to February 21. So far one thrust fault plot has started in the top left of the graph and is horizontal with a slight downward dip, another is plotted but only since the 19th February. The tiltmeter and earthquake plots have started bottom left and are dipping slightly upward.

Thrust Plot: your thrust fault specialist reports: 'The two lines are two separate thrust faults. They have both been moving at a constant but very slow speed since we started measuring them.?

Earthquake Plot: Your earthquake specialist reports 'there is very little activity below Mount St. Helens but the number of earthquakes per day has been constant since December 14th.?

Tiltmeter Plot: Your tiltmeter specialist reports 'The tiltmeter has only been operational for the? past 2 weeks but we are measuring a steady tilt of the ground near the lava dome. Maybe we are seeing the very early signs that magma is rising towards the surface.?

Using your experience from the previously described eruption click on one of the yellow boxes below:

1. The data shows nothing should be done

2. Do not issue a prediction but step up the measurements of the volcano (costs more money)

3. Issue a press release predicting that the volcano will erupt within a week

Answer Box (when 1, 2, or 3 statements above are clicked):

1.? It is true that the data plots indicate very little activity is occurring beneath the volcano. However, because of the thrust and tilt data it may be sensible to take some other action. Try again.

2.? This is probably the best action to take. The data may be showing the early signs of activity but there is too much uncertainty in the data to issue a specific prediction. However, increasing your measurements of the volcano is a good strategy.

3.? You might be correct in thinking that magma is rising to shallow levels beneath the dome. However, none of the data plots show a significant increase yet (up turn or down turn). Only when this pattern occurs is it possible to issue an accurate prediction.

(end of Flash page 4, Exercise)

March 3rd

Today the volcano was visible for the first time in 2 weeks. Many phone calls from the public have been received at the observatory reporting a steam plume rising from the crater.

Your tiltmeter specialist has requested a meeting with the team later today.?

Text version of the plots and notes on data:

There are three plots shown as for the previous screens. The plots show data from January 11 to March 3. So far the first thrust fault plot is nearly horizontal with a slight downward dip and the second plot (from measurements that began on February 20) shows a similar dip. The earthquake plot is dipping slightly upward. However, the tiltmeter plot is showing a moderate upturn in the plot over the last 7 days or so.

Thrust Fault Plot: Your fault specialist reports 'Our most recent measurements did not show a significant change in the movement of the thrust faults.'

Earthquake Plot: Your earthquake specialist reports 'The number of earthquakes per day is still fairly constant.'

Tiltmeter Plot: your tiltmeter specialist reports at today's staff meeting, 'about a week ago the tiltmeter data began showing an increased tilting of the ground next to the dome. This tilting has remained steady since then. Looking at the tiltmeter data alone suggests to me that the volcano is building toward another eruption.'

After examining all data plots and getting comments from your scientific team select an action to take by clicking the buttons below:

1. The data shows nothing more should be done

2. Issue a press release predicting that the volcano will erupt within 3 weeks

3. Issue a press release predicting that the volcano will erupt within a week

Answer Box (when 1, 2, or 3 statements above are clicked):

1.? Correct! Although the tiltmeter plot shows that what looks like an increase in tilt, neither of the other two measurements show a significant change. Asking your team to keep a close watch on the data in the next few days is advisable.

2.? Although the tiltmeter data appears to be showing some change? in ground tilt of the crater floor, the other 2 measurements do not show any evidence of significant change. Based on the data presented it is too soon to issue a reliable prediction.

3.? Although the tiltmeter data appears to be showing some change? in ground tilt of the crater floor, the other 2 measurements do not show any evidence of significant change. Based on the data presented it is too soon to issue a reliable prediction.

(end of Flash page 6, Exercise)

March 17th

You have received a call from the director of Mount St Helens National Volcanic Monument. She has asked for advice on whether it will be safe to lead a field trip for 20 hydrologists to the base of the crater early next week.

Your response needs to be consistent with the choice you make below.

Text version of the plots and notes on data:

The plots described above show data from January 11 to March 17. All data show significant activity is occurring. There has been a significant upturn in the tiltmeter and earthquake curves (increasing tilt of the crater floor and increasing numbers of daily earthquakes, and the two thrust faults plots show a sharp downward curve. This shows that the thrust faults are moving at an increased rate compared to the previous week.

Thrust Fault Plot: Your fault specialist reports 'Both thrust faults are moving faster than they were last week. Records show that before most other eruptions a similar change in rate occurs.?

Earthquake Plot: Your earthquake specialist reports 'The number of earthquakes per day has increased dramatically in the last week. We have taken a closer look at the location of the earthquakes of the past month and we have found a significant number of deep earthquakes (6-11 km below the volcano). We have not observed these deep earthquakes before previous eruptions. Because of these deeper earthquakes I think there is a possibility of explosive activity when the magma reaches the surface soon.'

Tiltmeter Plot: Your tiltmeter specialist reports: 'The tiltmeter data shows even more tilting than it did 2 weeks ago. We have never seen this much ground tilt without an eruption.?

Be sure to check with your specialists then select an action below.

1. The data shows nothing more should be done

2. Issue a press release predicting that the volcano will erupt within 3 weeks

3. Issue a press release predicting that the volcano will erupt within a week

Answer Box (when 1, 2, or 3 statements above are clicked):

1.? All of the measurements are showing an increasing level of activity. This action will mean 20 people will be at the base of the crater because you do not expect the volcano to erupt in the near future. Select another action. It may help to consult with your specialists.

2.? You are close. The data certainly indicates that something will happen within the next 3 weeks. However, a closer look at the data shows that the eruption is more likely to happen sooner than this prediction. A shorter prediction will be more informative for the public and could be the difference between life and death for the hydrologists and Director of the monument

3.? Correct! The best prediction is probably 1 week because all of the data are showing an increasing level of activity.

(end of Flash page 7, Exercise)

8:25 pm March 19th

An explosive eruption began at 7.27 pm sending a column of ash to an altitude of 14 km. Glowing projectiles were observed above the crater rim from an aircraft 35 km to the south. The explosion triggered a rock and snow avalanche out of the crater which was followed 30 minutes later by a surge of water that swept into the Toutle River below. The party of hydrologists would have been killed had they been there at the time.

Congratulations! You have now successfully predicted a volcanic eruption. Now check out the case studies of 2 other famous eruptions of the 20th century: Kilauea (1983) and Mount St. Helens (1980).??

Text version the plots:

All three plots show up curving (or down curving in the case of the thrust faults) plots up to the date of the eruption.

Text version of images:

A photograph of the crater and north flank of Mount St. Helens is shown. The dome has a rubbly surface covered with boulders and gas is seen escaping from many different parts of it.

A photograph of the crater and north flank of Mount St. Helens is shown. The lava dome inside the crater is clearly seen, and a plume of white steam rises from the top of the dome. Most of the volcano and crater is covered with snow except for an area around the dome that extends down the north side of the volcano. This dark area marks the path of a lahar (volcanic debris flow) that swept out of the crater down the steep north flank. The explosion of hot debris in the crater quickly melted snow and ice into a surge of water and rock debris that swept out of the crater.

(end of Flash page 7, Exercise)
| back to top |

Slide Show: 1983 Eruption of Kilauea Volcano, Hawai`i

Page 1

Kilauea Volcano is the youngest and most active volcano on the island of Hawai`i. Magma rises from a depth of more than 80 km into a magma reservoir located beneath Kilauea's summit. From here, the molten rock can erupt in the caldera above or travel along the east rift zone (an area of faulting and eruption) to erupt many kilometers from the caldera. Move mouse over the image for more details.

Text version of image:

A shaded-relief map of the island of Hawai`i is shows Kilauea Volcano, including its summit caldera and the craters associated with the east rift zone.

A cross section of the ground beneath Kilauea's summit caldera is also shown. This illustration shows magma moving vertically upwards into a magma reservoir directly beneath the caldera. Arrows in the diagram show that some of this magma moves upward and into the east rift zone to collect in smaller magma storage areas. Magma can move up to the surface to erupt anywhere along the east rift zone.

Page 2

Magma began moving from the magma reservoir upward into the east rift zone of Kilauea on January 2. From there it moved east (to the right in the animation above) along the rift zone. The locations of the earthquakes (green, red and orange circles) show this movement.

Text version of image:

A map of the upper part of the east rift zone shows the locations of earthquakes during January 1-8. The locations of the earthquakes move from the upper east rift zone on January 2 progressively eastward along the rift zone through January 8. Earthquakes also appear in the summit caldera during the end of this period.

Page 3

During the same period tiltmeters along the east rift zone recorded a tilting of the ground as magma moved up below the caldera and then east (to the right) along the rift zone between January 2 and 8. Deflations marked on the figure occurred as the magma passed each tiltmeter.

Text version of image:

Three tiltmeters are shown on a map of the caldera and rift zone. The tiltmeters are located at increasing distances from the caldera. Plots of tilt against time are shown for each tiltmeter: an inflation of the ground (magma pushing the ground up and outward) is shown by an upward curve in the data; a deflation of the ground (magma receding and the ground sinking down and inward) is shown aby a downturn in the curve. The tiltmeter nearest to the caldera shows a deflation of the ground before the other two instruments detect any ground movement. The next tiltmeter registers inflation shortly after the first tiltmeter begins to deflate; this second tiltmeter then shows the ground deflating about two days later. On about January 7, the third tiltmeter registers deflation of the ground.

This pattern of ground deformation is interpreted as a ?slug? of magma moving past the tiltmeters as the magma moves along the east rift zone away from the summit magma reservoir.

Page 4

Just after midnight on January 3, after 24 hours of earthquakes, lava erupted along new fissures in the east rift zone about 15 km from the caldera.

Text version of image:

In the foreground burnt tree stumps are visible. Beyond them a continuous line of lava fountains erups from a fissure. The lava shoots into the air as high as 10m. This photograph was taken on January 5, 1983.

Page 5

After two more eruptions in February and March, the lava fountains moved to a single eruption point at a 'spatter cone', soon to be named Pu`u `O`o? by scientists.

Text version of image:

The Pu`u `O`o? crater is shown with a lava fountain hundreds of meters high. This photograph was taken in September 1983.

Page 6

During the next 6 years Pu`u `O`o erupted every 3 to 4 weeks, usually for less than 24 hours. Each eruption catapulted lava nearly 500 m above the vent in spectacular lava fountains.

Text version of image:

A brilliant red lava fountain above Pu`u `O`o is shown. The wind is blowing it away from the observer, and the molten lava quickly changes to a black color as it falls back to the ground.

Page 7

Many eruptions were of blocky `a`a, the less fluid of the two types of Hawaiian lava. In 1986 this `a`a flow reached Kalapana village, destroying 14 houses.

Text version of image:

An active lava flow is moving through an intersection in seen covering woodland and gardens to a depth of several meters. It is also spreading across a cross roads and around its edges it is causing the road to catch fire. The chaos and irregularity of the lava contrasts with the neat garden it is approaching.

Page 8

Steam explosions often occur when lava enters the sea. This is one of the several hazards associated with lava entering water.

Text version of image:

The sea and a rocky cliff coastline can be seen. Lava is entering the water, but the lava and cliff are enveloped in a cloud of steam. An explosion has just occurred. Fragments of lava can be seen flying upwards and outwards from the base of the cliff. Each fragment has a trail of steam behind it.

Page 9

In 1986, the eruption changed from the lava fountaining episodes to nearly continous eruption of lava flows that spread to the ocean along the south coast of Kilauea. Twenty years after the start of the eruption, lava continues to erupt from Pu`u `O`o in the longest eruption on the east rift zone in at least the past 600 years.

Text version of image:

A lava flow up close is shown. Fluid lava appears to spread slowly across the ground. New lava exposed to the air cools quickly to from a black crust on its surface.

(end of Flash page 9, Kilauea Volcano Slide show)
| back to top |

Slide Show: 1980 Eruption of Mount St. Helens, Washington

Page 1

The explosive eruption of Mount St. Helens, Washington, on May 18, 1980 was preceded by two months of restless activity. Thousands of earthquakes, hundreds of small steam-blast eruptions, and a bulging north flank were clear signs that magma was rising into the volcano. Scientists knew that stronger activity was possible, but the timing of such activity could not be predicted.

Text version of image:

The eruption column of the Mount St. Helens eruption on May 18, 1980 is shown. The huge column of volcanic ash, pumice, and gas billows from the new crater of the volcano.

Page 2

The first earthquakes began gradually in mid-March 1980. A rapid increase in the number of earthquakes beginning March 25 (note sharp upward trend in curve) led to the first steam-blast eruption of Mount St. Helens on March 27. But for the next six weeks, the number of daily earthquakes was relatively stable as shown by the nearly straight line.

Text version of graph:

A graph is shown of the cumulative number of earthquakes (y axis) versus time (x axis) between March 18 and late May. On March 25 the relatively flat curve turns sharply upward through March 27, the day of the first steam-blast eruption (see next slide). Between March 27 and May 18, the curve maintains a relatively uniform slope, which indicates a steady number of earthquakes during this period.

Page 3

Steam-blast eruption from the summit in early April. The dark color in the eruption column is volcanic ash (rock fragments less than 2 mm in diameter); the upper white part of the column is steam. These eruptions declined in frequency from an average of about one per hour in late March to about one per day in mid April; they temporarily stopped on April 22.

Text version of image:

A billowing column of volcanic ash and steam is shown rising from the top of Mount St. Helens.

Page 4

Summit of Mount St. Helens in early April. The eruptions created two craters at the summit; dark patches on snow are volcanic ash. Note the double crack system cutting across the summit on both sides of the craters. For the next 6 weeks, the area between the cracks subsided as the north flank (left side of volcano) moved outward.

Text version of image:

A photograph showing a close view of the summit of Mount St. Helens is shown. Two craters are visible at the summit of the volcano. A double set of crack systems cut across the snow and ice covered summit on both sides of the craters; at the very top, the cracks form a series of cliffs. Dark patches on the volcano represent areas covered with volcanic ash.

Page 5

Scientist surveys a series of targets placed on the north flank of the volcano. In April and May, the north flank moved mainly outward (nearly horizontally) at a constant 1.5 m per day! Such steady and rapid movement formed a very noticeable bulge on the north flank. Move mouse over image for details.

Text version of image:

A scientist is shown using a surveying instrument (theodolite) for measuring targets placed on the volcano's north flank, which is visible in the distance. An umbrella shields the sensitive instrument from the sun. The scientist is looking through the theodolite towards the volcano. Gray ash covers much of the volcano, and a ?bulge? can be seen on the upper part of the north side of the volcano.

Page 6

On May 18 at 8:32 A.M., without additional warning, a magnitude 5.1 earthquake shook loose the bulge. This triggered a big landslide down the volcano's north flank (green in drawing). The sliding mass 'uncorked' the magma system inside the cone, causing powerful explosions to rip through the sliding debris (red in drawing). The exploding material quickly overtook the landslide.

Text version of images:

Three sketches of Mount St. Helens are shown to illustrate the landslide and series of explosions that occurred during the first few minutes of the May 18th eruption. In the first sketch a landslide (shown in green) extends from the top of the volcano down to the base of its north side; the landslide consists of two large pieces or blocks. Between the two landslide blocks, the beginning of a very large explosion is drawn in red. In the second image the landslide has moved 1-2 km beyond the base of the volcano and the explosion is larger. In the third image the landslide has moved another 2 km and the top of the volcano is now missing (removed by the landslide). The red cloud of the explosion has grown rapidly and nearly covers the underlying landslide, which is still moving.

Page 7

The explosions sent rocks, ash, volcanic gas, and steam as fast as 500 kilometers per hour across the forest west, north, and east of the volcano. Trees 2 meters in diameter were easily knocked over by the fast-moving debris. Temperatures inside the blast reached as high as 350 degrees C.

Text version of image:

A hillside is shown covered with trees knocked down with roots attached and snapped off above the ground. The trees fell in one direction (left to right in the photograph). Two people standing at the base of the hillside provide a rough scale for the size of the trees. Most of the trees are 1-2 m in diameter and tens of meters long. Some of the standing tree trunks are 10-15 m tall.

Page 8

The blast lasted only about 10 minutes. It reached as far as 27 km northwest of the volcano, knocking down about 4 billion board feet of timber. The blast killed 57 people.

Text version of image:

A pick-up truck that has been battered and smashed by the rock debris in the explosion is shown. in the center of the picture. The truck appears as if it was rolled by the laterally-directed blast to its current position. The doors are twisted and open, and the windows smashed. The ground is covered with rock debris from the explosion, and in the distance a collection of tree trunks and can be seen.

Page 9

The landslide moved 22 km down the North Fork Toutle River valley between 110 and 240 kilometers per hour. It buried the valley to an average depth of 50 m! Move mouse over the image for details.

Text version of image:

The top of Mount St. Helens (upper left) and the North Fork Toutle River valley (center foreground) are shown. The two edges (east and west rims) of the new crater are visible but much of the volcano's north flank is hidden behind a ridge in the middle of the image. The river valley sweeps from the middle left of the image to the center foreground. The valley is filled with rock debris, which is the landslide deposit that formed the top of Mount St. Helens before the eruption.

Page 10

The hot rock debris erupted during the initial eruption melted snow and ice on the volcano, creating surges of water that eroded and mixed with loose rock debris to form lahars (volcanic mudflows). This bridge across the Toutle River was destroyed by the largest lahar that formed by water seeping from inside the landslide deposit through most of the day.

Text version of image:

A section of the Toutle River valley is shown. A green bridge frame that spanned the Toutle River for vehicles is partly buried by rock debris deposited here by a lahar on May 18. The bridge was ripped from its foundation by the lahar, then carried and deposited here several kilometers downstream from its original location. The surface of the lahar deposit is littered with boulders as large as 1 m in diameter.

Page 11

Even in hindsight, the specific time of the May 18 eruption could not have been predicted. However, development of the two crack systems across the summit of the volcano and the constant but rapid growth of the bulge are now recognized by scientists across the world as warning signs for large volcanic landslides and laterally-directed explosions.

Text version of image:

A view of Mount St. Helens volcano soon after the May 18, 1980 eruption is shown. From the top of a ridge where a scientist is examining deposits from the lateral explosions, the new horse-shoe shaped crater is clearly visible. A lake at the base of the volcano, Spirit Lake, is partly covered with tree trunks. The trees were carried into the lake by the landslide and force of the explosions. Looking directly south towards the volcano the crater and flood deposits can be seen. The volcano is completely treeless, and volcanic fume and steam is visible in the crater.

(end of Flash page 11, Mount St. Helens Slide show)
| back to top |