Ambient Noise Characterisation and Detection Threshold of the DAS System at the Irpinia Near Fault Observatory

Distributed Acoustic Sensing (DAS) is an emerging technology that turns optical fibers into dense seismic arrays, offering dense spatial coverage for ground motion monitoring. However, DAS recordings are limited by instrumental noise, which can obscure weak seismic signals and ambient noise, and complicate interpretation. Understanding the characteristics of this noise, its variability, and its potential informational content is therefore a crucial step toward establishing DAS as a robust tool for seismology.  

We systematically analyse ambient seismic noise recorded by a ~22 km DAS array deployed in the Irpinia region of Southern Italy, under the Italian project MEET – Monitoring Earth Evolution and Tectonics. Using Power Spectral Density (PSD) analysis of continuous strain rate data, we quantify the spatial and temporal variability of noise along the fibre. A primary objective of this work is to characterise the noise floor of our DAS system and its dependence on frequency and channel positions. Our analysis reveals that noise levels even in the quietest fiber sections approach the converted Peterson High Noise Model, indicating that the DAS array along commercial cables exhibits higher noise levels than traditional seismometers, usually deployed in remote areas. Spatial variability in PSD shows that some sections exhibit a flat, low noise response (down to -200 dB), dominated by instrumental noise, whereas others are more sensitive to ambient noise. This substantial variation in noise levels, governed by cable deployment conditions, directly controls the sensitivity of individual DAS channels. 

Based on the statistical characterization of ambient noise, we estimate the detection threshold of the DAS array by comparing the observed noise spectra with the theoretical strain rate spectra. From this channel level analysis, we derived an integrated detection threshold curve for the entire array. We found that the DAS fiber can detect events with a minimum magnitude of Mw 1.75 at a hypocentral distance of 25 km, with the threshold increasing to Mw 2.5 at 60 km. This empirically derived threshold is validated by real seismicity.