Validation of Peak Ground Velocities Recorded on Very-high rate GNSS Against NGA-West2 Ground Motion Models

Abstract

Observations of strong ground motion during large earthquakes are generally made with strong-motion accelerometers. These observations have a critical role in early warning systems, seismic engineering, source physics studies, basin and site amplification, and macroseismic intensity estimation. In this manuscript, we present a new observation of strong ground motion made with very high rate (>= 5 Hz) Global Navigation Satellite System (GNSS) derived velocities. We demonstrate that velocity observations recorded on GNSS instruments are consistent with existing ground motion models and macroseismic intensity observations. We find that the ground motion predictions using existing NGA-West2 models match our observed peak ground velocities with a median log total residual of 0.03-0.33 and standard deviation of 0.72-0.79, and are statistically significant following normality testing. We finish by deriving a Ground Motion Model for peak ground velocity from GNSS and find a total residual standard deviation 0.58, which can be improved by ~2% when considering a simple correction for Vs30.

Characterizing High Rate GNSS Velocity Noise for Synthesizing a GNSS Strong Motion Learning Catalog

“…Crowell et al (2023) also finds that the lowest noise power in the frequency domain exists in the 1-10s periods of the highest sample rate observations (20Hz in their study), notable given this intersects the spectral region of the seismic ground motion waveforms of interest. Given the spectrum at higher sampling rates, there is likely potential for improved screening of TDCP velocities for our signals of interest to reduce temporal aliasing (Hohensinn et al, 2020;Crowell et al, 2023).…”

“…Such a classifier will be embedded in enhanced network operations and hazard monitoring for automated, stand-alone event detection. The subsequent benefit is an expanded training catalog (Dittmann et al, 2023) and framework that supports deeper learning models that are "data hungry" (Mousavi and Beroza, 2022). This includes expanding functional learning outputs, such as denoising, regression for magnitude inversion, and forecasting.…”

“…At approximately 4-6s period the noise spectrum inflects and begins increasing at a mirrored power law exponential to the lower frequencies. In TDCP at higher rates (>1Hz), Crowell et al (2023) observes in multiple sample rates from a single receiver that TDCP velocities have increased noise in the time domain, roughly a factor of 7 of standard deviation from 1 Hz to 10 Hz veloci- ties. In the frequency domain these velocities present as a reverse power law of increasing noise as frequency increases, flattening at a corner around 0.2s period (5Hz).…”

“…If available in real-time, GNSS-derived high rate time differenced carrier phase (TDCP) velocities (GRAAS and SOLOVIEV, 2004) applied to seismology (Colosimo et al, 2011) are an additional source of these intrinsic measurements (Parameswaran et al, 2023) that are traditionally sourced from dedicated inertial sensor networks. If available in near-real time or post processing, GNSS * Corresponding author: stdi2687@colorado.edu velocities can contribute to catalogs of ground motion measurements used for empirical regional and local ground motion models (Crowell et al, 2023). GNSS spatially complements or substitutes existing inertial ground motion observations (Crowell, 2021), especially valuable in sparse networks (Grapenthin et al, 2017).…”

“…However, ambient GNSS velocity noise remains well above the noise floor of inertial sensors, largely due to sources of uncertainty related to ranging of space-based weak radio frequency signals. Analysis of high rate positioning noise (Genrich and Bock, 2006), carrier phase noise (Wang et al, 2021), and TDCP velocities (Shu et al, 2018;Crowell et al, 2023) has shed valuable insight into the factors that influence the ambient noise floor of these GNSS velocities. To date, the GNSS velocity noise frequency spectrum has not been evaluated across sufficiently large temporal and spatial scales to statistically report on the ambient noise across a network.…”