Item

Abstract

Distributed Acoustic Sensing (DAS) interrogators have become commercially available, enabling long term measurements in diverse environments. Collection of permanent data is facilitating new applications such as earthquake detection. Pre-installed standard communication fibres can be used for the detection of seismic activities. However, the signal to noise ratio (SNR) is rather low at earthquake frequencies and therefore it is difficult to detect seismic waves at high ambient noise. Although standardized laboratory tests reveal a higher noise floor at low frequencies for DAS interrogators, the detection of earthquakes is still possible through filtering and signal enhancement methods. Furthermore, efficient real-time data evaluation is necessary for early-warning mechanisms. This article compares data of 3 different fibre lines with different characteristics and ambient noise for earthquake detection. Tests are performed with an inner-city line, which is exposed to high noise by day due to traffic loads and low noise at night. Additionally, a fibre line in a rural area with lower noise behaviour is investigated. Furthermore, DAS measurements at a fibre line at the national underground seismic observatory were performed and compared with the close by seismometers. In this paper we focus on methods for detection and evaluation of seismic waves and discuss the technical challenges of detecting small signals in big data. The paper also addresses matters of performance related to real-time data evaluation.