New Perspectives on the Shallow Environment in Alaska from co-located seismic, pressure, temperature, and wind sensors

Abstract

Listen

Translate article

The EarthScope Transportable Array (TA) in Alaska has been a unique seismic network since about 2014 because most stations are equipped with environmental sensors to record pressure, temperature, and wind (speed and direction). We will summarize some physical insights of near-surface properties in Alaska that can be gained from the combined analysis of seismic and environmental sensors. We also point out a possible effect of the thick sea ice on the climate in the North Slope region that faces the polar ocean. First, the combined analysis of seismic data and pressure data allows us to separate two distinct types of seismic noise; one is the ordinary seismic noise, consisting of propagating body and surface waves, and the other is the deformation caused by the local pressure loading. This loading effect is observed at many stations when surface pressure becomes high. It can be confirmed based on two pieces of evidence; one from high coherence between seismic and pressure data and the other from the phase difference between pressure and vertical seismic displacement. By selecting data from a high-pressure range, we can apply the compliance method, similar to the compliance method applied to ocean bottom observations (e.g., Webb and Crawford, 1998). We will show a map of shallow rigidity variations for the depth range of 50-100m. Second, the combined analysis of temperature and seismic noise allows us to identify the major effects caused by near-surface melting, primarily in the permafrost area. Some stations show a thousand-fold increase of horizontal noise in summer at 0.01-0.03 Hz in comparison to the frozen state. This anomalous horizontal noise can be seen at low frequency (< 0.1 Hz) and is undoubtedly related to tilt effects as its amplitude increases towards lower frequency. Third, seasonal variation in horizontal noise shows a rapid increase in summer due to melting but the way the noise level returns to the frozen (low-noise) state varies from station to station. For most stations, this return occurs well after the surface temperature becomes negative in September or October. But some stations require time until March of next year to return to the low noise level. These data suggest that the melt layer remains at depth for a long time even after temperature drops below freezing, perhaps developing a sandwiched molten layer between the developing ice from the surface and the underlying permafrost ice.