1989-90 Eruption of Redoubt Volcano, Alaska, and the First Test Case of a USGS Lahar-Detection System

Eruption column at Redoubt Volcano, Alaska

Redoubt Volcano, Alaska
Photograph by R.J. Clucas on April 21, 1990

Nearly every one of the volcanic events during the 1989-90 eruption of Redoubt Volcano generated lahars in the Drift River Valley. The lahars were caused by sudden melting of snow and ice from hot pyroclastic flows and dome collapses that swept down the volcano's north flank. Many of the lahars swept all the way to Cook Inlet (about 35 km), which raised concern about the safety of an oil-storage facility built on the bank of Drift River.

The repeated collapse of Redoubt's lava dome and and resulting lahars in Drift River created an ideal situation for testing a new experimental system designed to track lahars and other debris flows down river valleys. The system monitored in-the-ground vibrations caused by rock fragments crashing together and into the channel bottom as a lahar moved downstream. The system was removed from the Drift River Valley after the eruption ended.

Dome Collapses Trigger Lahars

Summit of Redoubt Volcano and lava dome, Alaska

Photograph by S.R. Brantley on April 25, 1990

Lava dome on north flank.

The last of a series of 14 lava domes erupted in 1989 to 1990 is precariously perched high on Redoubt Volcano's steep north flank. Each of the previous domes were destroyed as they collapsed down the north flank and across a glacier. The hot lava rocks carried by the pyroclastic flows melted snow and ice quickly and generated lahars that swept down Drift River valley. This dome (dark mound in center) began erupting four days earlier on April 21, and is about 100 m in diameter.

Lahars in Drift River Valley

Redoubt Volcano and Drift River valley, Alaska

Photograph by J. Major on March 10, 1990

Drift River valley below Redoubt Volcano.

Dark pathways of a lahar streak the snow-covered Drift River valley after a lava dome collapsed down the steep north flank of Redoubt Volcano on March 9, 1990. This relatively small lahar swept 35 km down Drift River valley to Cook Inlet. A few of the lahars were large enough to inundate the entire valley floor, which is about 2 km wide in this photograph.

Lahars Threaten Oil Terminal

Mouth of Drift River and oil terminal near Cook Inlet,

Photograph by J. Major on March 10, 1990

Mouth of Drift River.

Dark pathways of the lahar on March 9 move past an oil terminal marked by the white rectangle in lower center of photograph. As the lahar moved down Drift River, it became diluted with additional snow melt and passed on both sides of the facility. Lahars triggered by the growth and subsequent collapse of Redoubt's many lava domes generated concern about the safety of the oil tanks and people working at the site.

Oil Storage Tanks

Storage tanks of Drift River oil terminal

Photograph by J. Major in 1990

Oil storage tanks.

Oil tanks of the Drift River terminal temporarily store oil from about 10 oil platforms in Cook Inlet before the oil is pumped into tankers just offshore. The tanks can store up to 1.9 million barrels of oil. The largest lahar from Redoubt Volcano on January 2, 1990, flooded part of the terminal with muddy water. Lahars on March 22-23 and April 4, 2009 were about as large and also flooded the terminal with muddy water, ice, logs, and sediment (mainly silt and sand).

By the time Redoubt Volcano first erupted on December 13, 1989, a USGS scientist had designed a prototype system for detecting a lahar as it moved down a river valley. The system relied on seismometers sensitive to high-frequency ground vibrations (10-300 Hertz); seismometers used in recording earthquakes are sensitive to much lower frequencies, usually less than 20 Hertz. The only problem was that the system hadn't been field tested yet. Would it be useful in providing an immediate real-time warning to the people working at the oil terminal?

Scientists Install Acoustic-Flow Sensors

Storage tanks of Drift River oil terminal

Photograph by S.R. Brantley on April 23, 1990

Drift River valley.

In hopes of detecting and tracking lahars as they moved down the Drift River scientists installed 3 seismometers senstive to ground vibration at relatively high frequencies along the edge of the valley. The sensors are called acoustic-flow monitors (AFM) to distinguish them from traditional seismometers. The AFM's were installed at increasing distances from the volcano -- the nearest station at the base of the volcano was installed about 9 km from the dome. AFM #2 was installed 13 km from the dome and AFM #3 was installed about 20 km from the dome.

Sensors Send Data by Radio

Acoustic-flow monitor site

Photograph by S.R. Brantley on April 25, 1990

Acoustic-flow monitor site.

Thick layers of snow and ash in Drift River valley made it difficult for scientists to inspect the experimental acoustic-flow monitors in winter. Here at AFM site #1, Drift River heads east toward Cook Inlet; AFM #2 is about 5 km downstream. Radios at each stations sent data to the Alaska Volcano Observatory in Anchorage where an inexpensive laptop computer stored the data and received emergency alert transmissions when ground vibrations above a certain threshold were detected the AFM's.

Sensors Detect Vibrations

Graph showing data from the three acoustic-flow sensors

A lahar triggered by the dome collapse on April 6, 1990, was detected by all 3 AFMs. The top graph shows the amplitude of a seismic signal that was triggered directly by the dome collapse at 5:23 p.m. Note the delay in time before the AFM's recorded an increase in ground vibration caused not by the eruption, but a lahar that moved progressively downstream. Note the change in scale for each sensor. Did the seismometer used in detecting earthquakes and volcanic explosions detect the lahar? Did the flow sensors detect the explosion? The peak ground vibration for AFM #1 occurred at 5:30 p.m., AFM #2 at 5:35 p.m., and AFM #3 at 5:59 p.m.

It Works!

The field test at Redoubt Volcano was a tremendous success. Clearly, the system could detect and track lahars moving down a valley in real time. Calculating the flow velocities of the lahars was relatively simple, but estimating the relative size of the lahars proved to be much more difficult. However, experience at other volcanoes with other lahars since 1990 has shown that such estimates are indeed possible.


Brantley, S.R. (ed.), 1990, The eruption of Redoubt Volcano, Alaska, December 14, 1989 - August 31, 1990: U.S. Geological Survey Circular 1061, p. 33.

Dorava, J.M., and Meyer, D.F., 1994, Hydrologic hazards in the lower Drift River basin associated with the 1989-1990 eruptions of Redoubt Volcano, Alaska: Journal of Volcanology and Geothermal Research, v. 62, p. 387-407.