Contact: Sandra Hines
University of Washington

A small earthquake off the coast of Washington that caused hydrothermal vent systems miles away to pump out substantially warmer water at 10 times the rate and in an unexpected pulsing pattern has seafloor geologists questioning long-held assumptions about how fluid circulates within oceanic crust.

Hydrothermal circulation of seawater within the rocks of the seafloor has been studied for more then 20 years and conventional models assume that fluid flow is stable and relatively unchanging for long periods of time. Scientists were therefore astonished that hydrothermal vent systems nearly five miles from the epicenter of a June 1999 earthquake were affected at all, much less in such complex ways, according to Paul Johnson, University of Washington oceanographer and lead author of a paper about the observations in the Sept. 14 issue of Nature.

“The oscillations of temperature and flow, in particular, are telling us something really fundamental about fluid flow in the crust that we don’t understand,” Johnson says. He will be chief scientist on a National Science Foundation expedition Sept. 27 to Oct. 8 to deploy additional instruments to investigate these phenomena.

The effects reported in the Nature paper followed a tectonic earthquake June 8, 1999, about 180 miles off the Washington coast on the Endeavour Segment of the Juan de Fuca Ridge. Scientists have never detected volcanic eruptions on the Endeavour Segment and it is assumed to be “quiescent.” The segment has several large, stable hydrothermal fields with a small number of vents issuing water as hot as 350 degrees C but with most areas producing diffuse flows of water only 4 to 10 degrees warmer than the surrounding ocean.

National Oceanic and Atmospheric Administration scientists in Newport, Ore., recorded a magnitude 4.3 earthquake at the Endeavour Segment June 8, 1999 using the Navy’s SOSUS (SOund SUrveillance System) array. Four to 11 days later, seafloor instruments at hydrothermal fields 4.7 miles away unexpectedly began recording major changes in the hydrothermal vents. Preliminary analyses indicate that flow increased tenfold, vent fluids became substantially warmer and, most surprising, fluid temperatures began oscillating up and down (+/- 5 degrees C) with periods of 8 and 12 days.

While the majority of the aftershocks from the earthquake ceased within five days of the primary shock, the seafloor measurements showed that the changes in vent temperatures continued for up to 80 days, when the instruments were recovered.

“The conventional wisdom for nearly 20 years has been that hydrothermal systems are long-lived, are relatively stable in flow, temperature and chemistry, and there isn’t much that can change them,” Johnson says. Now what Johnson describes as a “relatively modest earthquake” must have changed the entire circulation pattern of crustal fluid over a wide area of the ridge in order to affect the vents being monitored by researchers’ instruments nearly five miles away. The instruments recording these changes were more than a mile from each other and were located within distinct hydrothermal vent fields believed to have fundamentally different sub-surface circulation paths.

“This was not just a local event. We’re continuing to try to determine how much of the ridge system may have changed in response to the earthquake,” Johnson says. It could be that the fluids over hundreds of miles of mid-ocean ridges are influenced by even moderate quakes, he says.

These earthquake-induced changes would most strongly affect the populations of animals and bacteria that live on chemicals in vent fluids. Warmer, more abundant vent fluids, cloudy with nutrient-rich particulates, would cause bacteria to flourish and the number of animals that feed on the bacteria to increase. These favorable conditions produced a “flush of life” around the Juan de Fuca hydrothermal systems observed by scientists following the June 1999 earthquake.

The increased biological activity might be the cause of the oscillations in vent temperature and flows that were detected, the scientists say in the Nature paper. Perhaps the oscillations are caused by the periodic plugging and unblocking of the sub-surface “plumbing” by bacterial mats, material that was shaken lose by the earthquake.

A second hypothesis for the source of the oscillations is that the earthquake could have caused cracks in the basaltic seafloor to propagate downward into deeper, hotter crustal rocks, periodically supplying new sources of heat to the hydrothermal system, Johnson says. A third proposed model is that the temperature variations may simply represent a temporary, unstable convection mode as the fluids come to new circulation patterns after the earthquake, similar to the changes observed in geysers on land that have been disturbed by nearby earthquakes.

A final model, not discussed in the Nature paper, compares the variations in the seafloor hydrothermal flow to oscillations observed in oil fields on land. Oil flowing from the porous rock has been observed to oscillate when the fields are disturbed by explosives or over-pumping in order to extract the remaining oil in the reservoir. The oil beds, however, oscillate much faster than the variations measured in the sea floor hydrothermal systems, Johnson says.


Other authors on the paper are Michael Hutnak, formerly a University of Washington oceanographer, now pursuing graduate studies at University of California, Santa Cruz; Robert Dziak and Christopher Fox of NOAA’s Pacific Marine Environmental Laboratory and Oregon State University; Charles Fisher of Pennsylvania State University; Ivan Urcuyo formerly of Pennsylvania State University, now at Gettysburg College; Jim Cowen of University of Hawaii; and John Nabelek of Oregon State University.

For more information:
Paul Johnson, (206) 206-543-8474, (206) 543-0912,

IMAGES AVAILABLE: Contact Sandra Hines for a locator map and charts showing the oscillations in temperature following the earthquake.

Sidebar: Quake jars assumptions
Researchers almost tossed data

“The damn thing broke.”

That’s what University of Washington oceanographers Paul Johnson and Mike Hutnak thought after retrieving a year’s worth of temperature data that behaved as expected for 10 months and “then went nuts.” The “nutty” data appears in the Sept. 14 issue of Nature and flies in the face of some long-held assumptions.

“Scientists have prejudices just like everyone else and we had this mental view in our heads that things on the seafloor just aren’t supposed to change like that. We thought, ‘Well, we did something wrong, our instrument broke,'” Johnson says.

It was during a chance conversation as Johnson walked his bicycle to his office one morning that he first learned from a UW colleague of a deep-sea earthquake that coincided closely with the start of the strange readings.

Johnson then sought quake information from Robert Dziak and Christopher Fox with the National Oceanic and Atmospheric Administration in Newport, Ore., who monitor earthquakes on the Juan de Fuca Ridge. They confirmed the mid-June event. He also contacted Charles Fisher at the Pennsylvania State University knowing that group also had instruments on the ridge. Johnson learned that Fisher also was considering discarding the summer months of temperature data because the wild variations just didn’t make sense.

Plotted out with the seismic information, the data went from suspect to dramatic: two distinct vent systems were rhythmically pumping out substantially warmer water at ten times the usual volumes within days of the quake.

Related Links

Earthquake-induced changes in a hydrothermal system
on the Juan de Fuca mid-ocean ridge
, [abstract] Nature, 14 Sep 2000

“Hydrothermal vents on mid-ocean ridges of the northeast Pacific Ocean are known to respond to seismic disturbances, with observed
changes in vent temperature. But these disturbances resulted from submarine volcanic activity; until now, there have been no
observations of the response of a vent system to non-magmatic, tectonic events. Here we report measurements of hydrothermal vent
temperature from several vents on the Juan de Fuca ridge in June 1999, before, during and after an earthquake swarm of apparent
tectonic origin. Such a large, segment-wide
thermal response to relatively modest tectonic activity is surprising, and raises questions about the sources of excess heat and fluid,
and the possible effect on vent biological communities.”