Geotechnical Trainspotting: Early Observations From the New Seattle Liquefaction Array

The Seattle Liquefaction Array (SLA) is a geotechnical monitoring array established by the US Geological Survey earlier this year in industrialized Seattle, Washington. Funding for the array was provided by the Advanced National Seismic System, at the behest of the Pacific Northwest Seismic Network's regional advisory committee. The SLA aims to further the understanding of earthquake-induced liquefaction, particularly the processes associated with repeatedly liquefied soils and the liquefaction of deeply buried deposits. The SLA occupies a site at which shaking-induced liquefaction was observed during earthquakes in 1949, 1965, and 2001. The SLA site is seismically noisy but important as it is similar to sites that host many structures in Seattle. The site is comprised chiefly by loose-to-dense interbedded coastal and river outwash sands. Instrumentation at the site includes four 3-component accelerometers at the surface and at depths of 5.4, 44.9, and 56.4 meters, a surface barometer, and six piezometers at depths of 6.9, 22.9, 28.9, 43.1, 46.9, and 51.9 meters. Emplacement depths were selected to sample a variety of liquefaction susceptibilities. Continuous data from all sensors are sampled at 200 samples per second, and are available from the IRIS DMC archive, with a buffer of data stored on site in the event of telemetry failure. To date, only a handful of earthquakes have produced shaking strong enough at the SLA to be observed within the high levels of background noise. However, the noise itself provides data useful to constrain the low-strain seismic and pressure response of the site. Notably, the array is within a few meters of a set of busy railroad tracks. Passing and parked trains expose the site to a broad bandwidth of deformations, including seismic frequencies, albeit with a source at the surface. Many times each day the site experiences both high levels of shaking, and step changes in the pressure field of a variety of amplitudes that may last from seconds to hours. The pore pressure response varies between sensors and can be used to elucidate the drained and undrained pressure response within the different strata. Spectral ratios of the ground motions from trains amongst the accelerometers confirm that the site is still responding linearly to these ground motions.