Horizontal-to-Vertical Spectral Ratio (HVSR) and Industrial Components: Enhanced Characterization and Semi-Automated Removal to Avoid Interpretation Pitfalls

Over the past decades, the Horizontal-to-Vertical Spectral Ratio (HVSR) from ambient microtremors has been widely used for seismic microzonation and constraining near-surface shear-wave velocity profiles. However, industrial vibrations may introduce artificial spectral components that distort HVSR curves and lead to misleading interpretations. This study presents a semi-automated workflow to detect and mitigate industrial influences by combining coherence analysis, a spectral derivative-envelope indicator, and particle-motion analysis. Affected spectral intervals are removed and reconstructed by interpolation, yielding HVSR curves that better represent the natural wavefield. Field datasets demonstrate reliable identification and removal of anthropogenic components, even when overlapping with natural peaks.