Ultralow Viscosity of Earth’s Near Surface Material Inferred from PBO Borehole Strainmeter Data

Abstract:

We perform an analysis of tidal signals recorded in the borehole strainmeters of the Plate Boundary Observatory (PBO) in the United States of America. We decompose the original strain data into various components of tide signal and other terms including a trend, air pressure response and random noise, using BAYTAP, a program that applies a Bayesian modeling procedure to analyze the strainmeter data. We find an anomalous strong S1 strain tide (with a frequency of 1 cycle/day) recorded in the strainmeters with three distinct characteristics: 1) the extracted S1 strain tide has abnormally strong amplitudes (50-200 times larger than that of the normal standard S1 tide); 2) the S1 strain amplitudes exhibit a linear relationship with air pressure S1 amplitude; and 3) the S1 strain tide has a phase lag of 0-90° from the air pressure S1 tide. Various quantitative analyses suggest that the observed anomalous S1 strain tide cannot be explained by loading of ocean water or thermoelastic response of rock to daily temperature change. Neither can it be explained by elastic loading of air pressure, as elastic response would result in no phase lag between them. We suggest that the observed anomalous strain S1 tide can be best explained by viscous response to the loading of air pressure S1 tide, resulting in a linear correlation between the amplitudes of the two tides but with a phase lag. Using a Maxwell viscoelastic model, we estimate the viscosity of the Earth’s top materials (between the surface and strainmeter buried depth) is at the order of 10¹⁵ Pas. Such a viscosity value is much smaller than the normal viscosity of the rock (10²¹ Pas) and suggests that the physical behavior of the top thin layer of the Earth’s surface is very different than we have assumed.