Search

Characterizing and monitoring the highly variable near-surface underneath urban areas is crucial for understanding phenomena such as resource distribution and storage potential, changes in hydrological systems, and seismic hazard at the sub-kilometer scale. Our current understanding of the urban subsurface is limited by our ability to image its structure and temporal variations at high resolution using classical geophysical approaches. Distributed Acoustic Sensing (DAS) deployed on unused telecommunication fiber-optic cables (dark fibers) is an attractive alternative to classic seismological studies in urban areas, as it enables continuous recording of the seismic wavefield across long distances (10’s of km) at high spatial (~1 m) and temporal (down to seconds) resolution across a broad frequency range (mHz to kHz). Here, we explore the potential and challenges of using DAS deployed on dark fiber networks for characterizing and monitoring the near-surface beneath urban areas. We will discuss dark fiber experiments conducted in a variety of settings, including several km-long profiles crossing agricultural areas and densely populated centers in California, and experiments conducted within the city of Potsdam (Germany). We evaluate the usability of anthropogenic signals as passive sources for near-surface imaging and monitoring, as well as the potential for active source seismic imaging in complex noise environments. Challenges associated with poorly documented deployment and coupling conditions of dark fiber networks and the large data volumes generated in these studies will also be discussed. Ultimately, we aim at providing insights into the challenges and opportunities of using dark fiber DAS for understanding the urban subsurface., The 28th IUGG General Assembly (IUGG2023) (Berlin 2023)

The here referenced dataset provides eventbased Distributed Acoustic Sensing (DAS) recordings made with an approximately 22 km long dark telecommunication fiber lying in urban Potsdam and surroundings. For each of 164 Mgt;=5 earthquakes occurring in February 2023 and listed by the USGS, one hour of data is provided starting with the event's origin time. Additionally, the whole day of February 14 is provided in hourly files. The data was recorded in the frame of the global DAS month, an initiative to collaboratively record and share simultaneously recorded DAS data from all over the world (https://www.norsar.no/in-focus/global-das-monitoring-month-february-2023). DAS is an emerging technology increasingly used by seismologists to convert kilometer long optical fibers into seismic sensors.

Reliable well completion and monitoring technologies are a prerequisite to achieve safe, sustainable and efficient use of geothermal reservoirs. Since borehole completion is a high-risk process, we tested how distributed fiber optic sensing technologies can aid monitor the completion process continuously in space and real time. To evaluate the success of the operation conventional well logging tools and other methods can only be used after the completion and show snapshots of discrete points in time.We present results from a case study in Berlin, where a 450 m deep exploration well was completed for Aquifer Thermal Energy Storage (ATES) purpose. To simultaneously measure distributed temperature (DTS) and distributed elastic and static strain (DAS and DSS) a fiber optic cable was installed behind the production tubing. The combined evaluation of the different fiber optic technologies evidences the potential to observe previously untraceable processes, such as changes in subsurface flow paths due to clogging. Further, the unwanted process of caving can be identified and the rise of gravel or cement observed. This allows to react on site, and subsurface risks will be considerably mitigated., The 28th IUGG General Assembly (IUGG2023) (Berlin 2023)

Seismic methods have been widely used to image the subsurface for geological characterization and exploration. Due to the limitation of active experiments in populated areas, methods tapping the existing ambient noise as a source of excitation have gained increasing attention. The combination of this approach with Distributed Acoustic Sensing (DAS) is particularly appealing as vast telecommunication networks can potentially be used as sensors. However, seismic noise in urban areas is complex. Usually it is of anthropogenic nature and dominated by traffic. To extract coherent signals from recorded noise, the acquisition of long time series associated with high computational costs is required. In the study presented, we analyzed the cultural noise along an urban road in Berlin. DAS data was collected at a spatial period of 8 m and a temporal frequency of 1000 Hz along a 3-km long segment of a dark telecommunication fiber. The recorded noise contains transient signals with frequencies between 1 and 40 Hz attributable to the passage of trains. As these signals inhibit the extraction of coherent Rayleigh Waves, we applied a denoising method based on the harmonic-percussive separation technique prior to the standard ambient noise pre-processing (data decimation, band pass filtering, temporal and spectral normalization, cross-correlation) . We compare processing results with and without denoising and conclude that our approach leverages the recorded data better and yields improved virtual shot gathers . This implies the possibility to conduct shorter campaigns for acquiring equivalent results with respect to ambient noise analysis., The 28th IUGG General Assembly (IUGG2023) (Berlin 2023)

The dataset presented here is an earthquake catalog for the Central Sea of Marmara (Turkey) obtained by applying a matched-fliter technic to continuous waveforms. The magnitude of completeness of this catalog is Mc=1.1. We use as templates events published by national agencies (KOERI and AFAD). The matched-fliter technic is described in Bentz et al. (2020). The column of the data file are: event ID, Year, Month, Day, Hour, Minute, Seconds, Matlab time (serial time), Latitude (dec.degrees), Longitude (dec.degrees), Depth (km), Magnitude, Cross-correlation coefficient (CC), Template ID, MAD(ratio between CC and median absolution of daily correlogram), Quality flag The ZIP files contains configuration files for ph2dt and HypoDD applications together with input phase and seismic network data.

In January 2020, a swarm of earthquakes started under Thorbjorn volcano, Reykjanes, SW Iceland, associated to the uplift of up to 0.5 cm per day. Concern in Iceland is growing and the Iceland Meteorological Office suggested that possibly magma was intruded in the crust at shallow depth (3 to 9 km). The GFZ is starting a seismological Hazard and Risk Team (HART) in cooperation with IMO, ISOR and the University of Iceland. This dataset comprise a selection of waveforms recorded along a fibre length of 21 km. The interrogator is located in Grindavik and is connected to a standard telecom cable. Waveform data are available from the GEOFON data centre, under network code 5J.

This paper was selected for presentation at the symposium by an ARMA Technical Program Committee based on a technical and critical review.