Exploring vertical component dominance in seismic bedload transport signals: Horizontal-to-Vertical Spectral Ratio (HVSR) analysis in the River Feshie, Scotland

Abstract

Bedload transport plays a crucial role in shaping landscapes, yet monitoring it is challenging. Seismic sensors have emerged as valuable tools for continuous and non-invasive bedload transport monitoring. However, isolating the seismic signal of bedload transport from other environmental signals such as flow turbulence remains a challenge. While seismic waves propagate both vertically and horizontally, previous seismic bedload transport studies focused solely on the vertical component. This was based on the assumption that the bedload transport signal was mainly contained in Rayleigh waves which propagate with both vertical and horizontal motion, as opposed to Love waves which propagate with only horizontal motion. We hypothesise that there may be a significant signal from horizontally-propagating waves that characterises the interactions of coarse bedload impacts, and that this signal will be strongest in a flow-parallel orientation.   This study employs the Horizontal-to-Vertical Spectral Ratio (HVSR) which is a passive method, commonly used in engineering seismology, that determines the ratio between horizontal and vertical seismic signal components. In this study, we explore the potential of the HVSR method to isolate the dominant component in seismic bedload transport signals and its applicability for monitoring fluvial processes within rivers. Using seismic, hydroacoustic, and hydrological measurements from the River Feshie in Scotland, our findings challenge prior belief that the seismic signal of bedload transport predominantly resides in the vertical component; instead, the horizontal component contains significant fluvial and bedload transport information. Due to differences in seismic wave characteristics, the HVSR method acts as a tool to isolate signals of bedload transport and water turbulence. Additionally, the HVSR method demonstrates promise in effectively filtering out meteorological signals that may contaminate raw river-induced seismic signals, enabling more accurate monitoring of bedload transport occurrences. However, we acknowledge that the contributions of horizontal and vertical signals greatly depend on sensor location and site characteristics. This study emphasises the significance of utilising horizontal seismic signals for comprehensive bedload transport monitoring, presenting an opportunity for this method to enhance our understanding of complex fluvial processes within river systems.