Editor's Note:

The following is one of the most exciting scientific descriptions of the Nisqually earthquake that I have read. It recounts the strong shaking produced by the earthquake, and most interestingly, describes the formation of a sand boil in the Port of Seattle on Harbor Island. Sand boils result from the liquefaction of sand layers that contain abundant water. This description by Bob Norris, a seismologist with the U.S. Geological Survey, is one of the few known observations of the formation of a sand boil by a scientist. Bob wrote his account about 6 days after the earthquake on March 6, 2001.

This is a narrative of my observations of strong ground motion and a sand blow I observed on Harbor Island from the Nisqually earthquake.

Harbor Island is located on the south shore of Elliot Bay, south of downtown Seattle. The island consists largely of artificial fill and overlies former tidal flats of the Duwamish River delta. In common with other sites on artificial fill, Harbor Island shows high site response during earthquakes. (I can now verify that from personal experience.) For this reason, the USGS maintains a portable digitally recording seismograph on the island.

At 10:54 AM on February 28, I was driving along 11th Ave. SW on Harbor Island enroute to the seismograph site to download seismic data and perform routine maintenance. The instrument is located in a small outbuilding that contains fire control equipment for a complex of oil tanks nearby. I had just entered the gravel driveway that gives access to the site when the truck started yawing from side to side as if I'd just driven diagonally over a large speed bump. I thought I had driven over something I hadn't seen, and went through several seconds of confusion because the truck was still rocking sharply after I had stopped. It wasn't until I looked up and saw some electrical wires overhead swaying and their support poles leaning back and forth that I realized this was not only an earthquake, but an unholy BIG earthquake! I was utterly amazed that all this ground motion could go on so quietly-- I could hear things creaking and clanging from all the swaying, but the ground itself was silent.

When I stopped gawking and resumed thinking, I remembered the wires overhead and gunned the truck into an open area about 60 feet ahead, which looked like a safe area to wait it out. At this point, about 15-20 seconds into the strong ground motion, its amplitude seemed relatively constant and although the ride was bumpy I had no problem steering the truck into the open area. It seemed the worst was over, but as I stopped the truck again the amplitude of shaking abruptly increased. In less than a second the truck was rocking so violently I lost sight of everything outside and could do nothing but hold on and hope my flying head didn't hit anything. This violent phase was brief, perhaps 5-7 seconds, but long enough to give me a mild strain in my neck. If it went on much longer, this could turn from an exciting professional experience to a survival situation. I remembered I was next to an oil tank farm and had a visualization of the huge oil fire in Valdez, Alaska after the 1964 quake, which I could have done without.

When it eased enough for me to be able to look around again, perhaps 25-30 seconds after the strong shaking began, I saw the dozen or so 200 foot tall cargo cranes that line the waterways of Harbor Island quivering and flexing in place, resembling huge steel giraffes trying to dance. I remember hoping no one was in them. It was at least another minute before I felt safe enought to get out of the truck. Outside, there was a pervasive background din of car and industrial alarms going off all over the city. As I walked over to a crew of ARCO people in hardhats about 100 yards away to see if they were all right, I could plainly feel the asphalt under my feet gently moving back and forth with about a 2-3 second period. We traded our stories for a short time, perhaps 2 to 3 minutes, then I walked back to the truck. At this time I could still feel subtle ground motion if I stopped walking.

Since my cell phone was out and I no longer felt in danger there, I thought I might as well get the earthquake data from the seismograph. I estimate that what happened next occurred at least 5 minutes after the onset of strong shaking at that point, perhaps as long as 10 minutes, but that's probably an upper limit.

I had just opened the door of my truck to get my laptop and notebook when I was distracted by a wet swishing sound coming from the ground nearby. I looked over to its source and saw a smooth dome of brown fluid, perhaps a foot and a half wide and high, issuing from the ground a few yards away from the southeast corner of the fire control house where the seismograph was located. This dome lasted perhaps two seconds, then grew and burst into a muddy geyser. This geyser issued three or four very fluid splashes over the next few seconds, about a yard high each, then it widened and collapsed into a column about half that wide that discharged a tremendous volume of muddy water. This flood emerged much faster than it could spread, so that within a few seconds the flow front had become a surge several inches high, like a small wave travelling up a dry beach. Its velocity was near 3ft/second as far as I could tell. Within an estimated 30 seconds, the surge had grown into a shallow rotating pool about 25 feet across with bits of suds floating on it, still vigorously fed by the column of water at the original breakout site.

I confess I didn't think it was a liquefaction feature at all; the delayed onset, the limited amount of sediment in the water, and the high flow rate convinced me it was a water main break-- particularly as it occurred near a building containing fire control equipment. In fact, I was annoyed because I thought the growing pool might engulf the driveway and strand me there, or prevent me from getting to the seismograph and downloading data! The feeder column remained centralized at the breakout site but began to gradually wane after a couple of minutes. I walked over to get a closer look and was surprised to find the water was relatively clear; I could see to a depth of a couple of inches in the pond.

Unfortunately, I paid no further attention to it and focused on getting data from the seismograph. When I left the site about 90 minutes later, I noticed that the column had dwindled to a disturbed patch of water in the now-quiet pool, which had approximately doubled in size.

After learning that this was indeed a sand blow, I returned to the site as soon as time allowed- about 3 days after the quake. Its deposit consisted mostly of dark sand-sized material, much coarser than the fine muds emitted by similar features along First Avenue South. This may explain why the eruption was so fluid; the sand added relatively little viscosity to the water and quickly settled out once the water had surfaced. The vent that I had witnessed forming had been filled in with gravel by the property owners, who had cordoned off the area. The area covered by emitted sand was approximately 45-50 feet in maximum diameter. I was surprised to see several other vents in the sand (closer to where my truck had been!); these may have erupted shortly after the initial vent and were submerged before I could see them.

This was only one of many sand blows that occurred on Harbor Island. I was luckier to get out of there than I realized; subsurface piping had opened an oblique collapse pit about 3 feet wide and of uncertain depth, only a few feet from where my truck had been parked.

Bob Norris