The site of the 1998 eruption was Axial volcano, along the Juan de Fuca Ridge seafloor, located about 300 miles off of Cannon Beach, Oregon. This volcano was the focus of a long-term National Oceanic and Atmospheric Administration (NOAA) research effort called the Vents Program, which studies the mechanisms by which the earth's interior exchanges heat and chemicals with the earth's surface through seafloor spreading centers.

"We began monitoring Axial in 1987, using simple bottom pressure recorders to measure the long-term vertical movements of the seafloor associated with magma transport within the volcano," says Christopher G. Fox, of the NOAA facility in Newport, Oregon. "We never expected to get this close a look at the eruptive process."

The seafloor instrument, a Volcanic System Monitor (VSM), had been installed on the summit of Axial in October 1997 to continue a research effort begun by scientists in 1987. The precise location of the VSM, also affectionately known as the "Rumbleometer," was based on geological and geophysical measurements and the instrument was thought to overlie the magma center.

Boy, did it ever. Axial blew its top in January 1998.

"When we originally designed the instrument, we field tested it on the big island of Hawaii near some active volcanoes," says NOAA's Christian Meinig, the lead engineer at Pacific Marine Environmental Laboratory (PMEL) who devised the apparatus to recover the instrument. "But we never imagined it would get stuck in a lava flow. To my knowledge, this is the only time an instrument has become engulfed in lava and recovered from the seafloor."

Recovery Operations
Following the eruption, NOAA vessels visited the site and attempted to recover the instrument. Although the VSM responded, it would not release to the surface. The ROPOS group in Sidney, British Columbia, Canada, was brought in to help locate and recover the Rumbleometer.

ROPOS stands for the Remotely Operated Platform for Ocean Science, an electro-hydraulic ROV designed and built by International Submarine Engineering. In deepwater mode, the ROPOS vehicle is a component of a cage/vehicle system rated to 16,400-foot (5,000m) depths. In this configuration, the vehicle and cage are deployed as a unit to the target depth. At depth, the vehicle operates independent of the cage on up to 300m of flying tether. In this mode, ROPOS is a 30Hp vehicle, with an additional 10Hp available to cage systems. The ROV is equipped with two video cameras, two manipulators, sonar, a variety of custom sampling tools and several digital data channels. The vehicle and cage are normally navigated with an acoustic long baseline tracking system that is calibrated with differential GPS.

Keith Shepherd headed the ROPOS group on the mission. The first recovery attempt was launched from the NOAA ship Ron Brown in August 1998.

It did not go well.

"The rigging was designed and built by both Chris Meinig and his shop, as well as our shop," said Shepherd. "The design was arrived at with discussions between us all. The first attempt was made with material that we had on hand ­ we used a 2,000-pound weak link."

The team took the winter to regroup. They designed an improved rigging apparatus, with a 10,000-pound weak link. They went at it again in June 1999, launching off the University of Washington's RV Thomas Thompson. The second time proved the be the charm.

"The Rumbleometer, which is self-buoyant, had released its anchor, but it did not surface because the legs were encased in lava. We knew that, once it was torn free, it should float on its own. We needed to attach an anchor, so that when it was free, we could control when it surfaced ­ we did not want it floating up into the vehicle or cage. The pull needed to be made from the cage, which meant attaching the cage to the Rumbleometer."

A spool of high strength Spectron line was mounted on the vehicle. The ROPOS then swam to the Rumbleometer and attached the line, which was paid out as the vehicle swam back to the cage. Once in the cage, the ROV transferred the other end of the line to a lifting bridle on the cage itself, which was hauled up with the winch.

"At this point, the Rumbleometer broke free and was floating in the water column on the end of a line to the new anchor," says Shepherd. "We triggered the hydraulic releases on the cage lifting bridle, which released the lifting line from the cage. The Rumbleometer was now free of the cage and ROV. We stood the ship and cage off of the site, swam the ROV to the Rumbleometer, and cut the anchor line. The Rumbleometer floated to the surface and was recovered with the small boat, which towed it to the ship's side for lifting."

While successful, it was a delicate operation. "The Rumbleometer was not designed to be recovered this way," Shepherd says. "It is much easier to design an instrument for ROV recovery before it is deployed! The heave of the cage, caused by the surface waves, caused some interesting and nail-biting motions of the buoyant Spectron line. The small boat had some lines tangled around the electrical connectors during recovery, and this tore off the connectors and allowed some water to enter the housings. Quick action saved the wet electronics, and more importantly, the data!"

What Does the Data Tell Us?
To everyone's surprise, the instrument was recovered with very little damage. Bill Chadwick, Jr., of Oregon State University's Cooperative Institute for Marine Resources Studies, says, "Perhaps the most surprising thing was that the temperature recorded in a pressure case on the instrument only rose from three degrees Celsius up to seven degrees Celsius, even though the lava erupted at about 1,200 degress Celsius and completely flowed under the instrument. This is apparently because a solid crust forms on the surface of the lava almost immediately and insulates the molten interior of the flow. Also there was an almost unlimited supply of frigid seawater to dissipate the heat."

"Much of the data were intact, in particular the pressure and temperature data," said Fox.

The data give a detailed view of the dynamics of a deep ocean volcanic eruption. In addition to the information on the flow itself, the long-term pressure record, in conjunction with other instruments deployed around the volcano by NOAA's Vents Program, provided a picture of what happened to the magma in the subsurface, making the 1998 Axial event the first deep submarine eruption ever recorded.

Fox says little is known about deepsea eruptions because only in the last decade have we been able to detect them, and none has ever been witnessed. "The data we report here, recorded by the VSM instrument caught in the 1998 lava flow at Axial volcano, were obtained by fortuitous circumstance. The instrument was simply in the right place at the right time, with the right sensors, and happened to survive the eruption," he said. "It is doubtful that we will ever be clever enough to intentionally place an instrument in an active submarine lava flow, so this serendipitous recording becomes a benchmark in our understanding of submarine volcanism."