Broadband, short-period or geophone nodes? Quality assessment of Passive Seismic signals acquired during the Maupasacq experiment

Abstract

Passive Seismic is a broad term, incorporating various techniques and methodologies, which all exploit some part of the seismic signal that naturally exists or occurs in the Earth’s subsurface. This signal may differ significantly in the form and/or the provenance (e.g. earthquakes, ambient seismic noise, etc.), as well as the frequency content and, subsequently, the part of the subspace on which it may carry useful information. People involved in Passive Seismic often encounter the question: ‘What type of instrument is suitable for a passive seismic survey?’. Passive Seismic instrumentation usually consists of three-component seismic sensors, which mainly differ in the frequency range they are able to record (broad-band, short-period or geophone nodes). Having its roots in seismology, where traditionally broadband stations have been used for decades, but heading towards exploration, where instrumentation has to be cost-efficient and easy to handle in order to permit the adaptation at a reservoir scale, Passive Seismic instrumentation still tries to strike a balance between cost and bandwidth. Having in mind the variability of Passive Seismic methodologies and instruments, the Maupasacq experiment, a large passive seismic survey, has been launched in the Mauleon basin, SW France. The scope of the experiment was to image the area of interest by jointly applying a number of passive seismic methodologies, each one contributing to the final image with a different piece of useful information. The area of interest was carefully selected as, on one hand, the Mauleon basin consists of a former Cretaceous hyper-extended rift, inverted during pyrenean orogeny and, on the other hand, it provides a means of evaluating the results acquired, as an abundance of geological and geophysical data already exist in the area. In this context, a dense seismic network of 417 three-component (3C) sensors was deployed in an area of approximately 1500 km2. In addition to those, 24 peripheral stations have also been installed in an outer ring, extending the survey area to 3500 km2. The network was continuously operating for a recording period of six months (from April to October 2017) and consisted of three different types of seismic stations: 190 geophone nodes (SG- 10 3C SERCEL), 197 3C Seismotech short-period stations and 54 broadband stations (Guralp CMG40, Trillium Compact and Trillium 120). This fact, apart from imposing the difficulty of jointly processing data recorded by different types of instruments, having to deal with different instrument responses and data formats, it also permitted an evaluation of the suitability of each instrument type for each one of the passive seismic methodologies applied. This evaluation was performed in the course of an initial quality control (QC) procedure of the acquired dataset, permitting the extraction of valuable conclusions on the performance of different types of instruments, operating in the same area, during the same period of time. The main aspects that were evaluated were the acquired signal itself, as well as the frequency content of the recordings, in various circumstances (i.e. the occurrence of a local earthquake or a teleseismic event). Besides, it has been observed that the energy of a passive seismic source seems to play a big role in the definition of the ‘real’ recording limits of each type of instrument