Work package number: WP8

Work package title: Monitoring seismicity and fluid activity near the fault using existing cabled and autonomous multiparameter seafloor instrumentation          

WP Leader: IFREMER                 


The objective of WP8 is to implement an integrated approach based on multiparameter seafloor observatories, to continuously monitor the micro-seismicity along with the fluid expulsion activity within the submerged fault zone. The discovery of very characteristic tremors, 44 minutes prior to the Izmit earthquake in 1999 [Bouchon et al, 2011], and the finding that gas reservoirs are connected to the Main Marmara fault zone [Bourry et al, 2009, Géli et al, 2008, Gasperini et al., 2012] opens new perspectives that were not imaginable a few years ago: if seismic tremors are found to be associated with clear anomalies in gas emission activity, then we will have more criteria for characterizing and identifying the recorded signals as indicators that the probability of occurrence of an impending earthquake is increasing. WP8 will hence contribute to find robust and multiparametric methods for implementing early warning strategies in the near future. KOERI presently operates a unique network of 5 cabled, seafloor observatories (SBOs) in the SoM, which provides seismological data from Broad-Band OBSs, along with accelerometric data, sea bottom temperature and pressure gauge data. This network needs to be complemented with an additional network of multiparameter, seafloor observatories to continuously collect geochemical and geophysical data from the immediate vicinity of the fault zone. However, the implementation of such observatories cannot be fulfilled within the time span and budgetary frame of MARsite. To get a step forward, additional data with autonomous[1] seafloor instrumentation (such as SN4, that was deployed for almost 12 months in the Gulf of Izmit) will be collected within Task 1 and compared with periodical sampling collection of the water column in coincidence of the faulted/not faulted areas of the Marmara sea. The data from autonomous stations will be integrated and analyzed jointly with the KOERI network under Tasks 2 and innovative methods for multi-parameter data interpretation will be developed within Task 3 in order: i) to detect low magnitude earthquakes and improve the characterization of the near-fault micro-seismicity, particularly along the central part of the SoM; ii) to search for seismic tremors similar to those that were documented prior to the Izmit, 1999 earthquake; iv) to detect gas-bursts related events; v) to establish correlations between fluids and seismicity. The next generation of seafloor observatories for geo-hazard monitoring will be prepared within Task 4.

Description of work

Task 1. Collect multi-parameter time-series through three repeated (every 6 months) sea-based cruises, using existing autonomous seafloor observatories (SN-4, piezometers, OBSs, acoustic gas bubble detectors

The evidence that the seafloor of the sea of Marmara shows clear clues of recent faulting related to the last destructive Izmit event and areas where fluids vents into the sea water, supports the possibility to gain a better insight of new faulting phenomena based on a monitoring activity carried out by both seafloor observatories and periodical observations of the venting fluids.

Task 1 will take care of the collection of geophysical and geochemical data using existing submarine equipments (already used in the same area in previous projects) and oceanographic cruises to collect water samples aimed to constrain the origin of the vented fluids, their mixing ratios and possible interactions processes. This data set will be the reference for a correct interpretation of the multiparameter long-term time series.

Task 2. Seabed, continuous seismological monitoring and integration of multi-parameter datasets from cabled (KOERI’s) and autonomous systems (T, p, seis)

The geophysical dataset (including micro-seismicity from Broad-Band OBSs, accelerometry, sea-bottom water pressure and temperature variations) collected by the submarine, cabled KOERI network will be analyzed within Task 2. Specific methods will be implemented to detect earthquake doublets, e.g. similar earthquakes coming from the same location. These methods will also include the detection of slow slip events and seismic tremors, supposedly associated to the earthquake preparatory phase, based on laboratory and modeling results. This requires improvement of the techniques used for the determination of earthquake parameters (hypocenter location, moment magnitude, etc).

The land-based seismic network indicates that the central SoM is characterized by a relative absence of seismic activity, most particularly across the Kumburgas Basin and Central High. How much silent this segment actually is an open question of fundamental importance that will be addressed by combining the data from the KOERI permanent network and the data from the autonomous OBSs (see Task 1). Earthquake clusters and micro-earthquakes will also be specifically investigated to get information about future large earthquakes. Cluster earthquakes are small and impossible to locate with a limited number of stations. After locating a moderate earthquake, we will look for small earthquakes of similar size, by waveform correlation with the larger one, in order to list earthquakes coming from the same site.

Specific methods will also developed to detect non-seismic micro-events that have recently been interpreted as the result of gas outbursts from the Marmara seafloor. Recent, unpublished data indicate a relation between these non-seismic tremors and temperature variations. An upward displacement has also been observed on the vertical OBS component just before the gas bursts. The relations between non-seismic micro-events and gas outbursts will be fully documented within Task 2, with the view of finding indicators of seafloor deformation.

Task 3. Multi-parameter data analysis

Monitoring of fluid emissions through the seafloor near an active fault is an innovative experiment, which requires the development of new methods for data analysis. On land around the Sea of Marmara, statistical properties of radon time series were shown to vary with seismic activity (Inan et al., 2008) and similar relationships will be looked for in the marine data set. Fluids migration is influenced by stress and strain rate variations, and in turn, fluids migration through the crust can be a source of induced seismicity (e.g.. Miller et al., 2004; Daniel et al., 2011). The primary objective of this task is thus to analyze time series of fluid parameters (fluid pressure in the sediment and at the seafloor, bubble flux, fluid velocity, fluid chemical composition), and examine possible correlations with micro-seismicity and strain. In addition, the influence of water level oscillations on fluxes should be taken into account, although these oscillations are relatively small in the Sea of Marmara. They display a complex spectrum with tidal as well as longer and shorter period oscillations (Alpar and Yüce, 1998). The modification in the flow regime and fluids composition can be also induced by tectonic events (Heinicke et al., 2009). Permanent seafloor seismological stations equipped with environmental sensors (pressure, temperature, salinity, current velocity) will bring records unaffected by coastal site effects and will be used to calibrate models of normal mode oscillations that have been developed at ITU. Characterizing normal oscillations in the Sea of Marmara will also have important outcome for the mitigation of tsunami hazard and for understanding the deposition of seismo-turbidites in the deep basins (Beck et al., 2004). Work planned thus comprises: (1) Analysis of fluid emission parameters and correlation with water column parameters, and micro-seismicity and strain; (2) modeling of water column oscillations and their consequences.

Task 4. Prepare the next generation of seafloor observatories for geo-hazard monitoring

The experience got after many years investigating fluids and their interactions with both the seafloor and tectonic structures, allows to better focus the attention on improvements of seafloor observatories devoted to long-term continuous monitoring of geohazards. Volcanic and tectonic-related fluids have the common feature that they can suddenly change in composition, flow rate, physical properties as a function of the intensity of the temporal development of the natural processes. Stress accumulation in an active fault is a slow process inducing normally changes in the circulating fluids over a similar temporal scale, while the closer is the rupture time, the faster are the related modifications of the fluids (including waters and gases with changes of some parameters over a time-scale from hours to seconds). A new generation of seafloor observatories should support the observation of both slow and quick variations, thus besides the collection of long-term data set it should enable a data collection in a real-time or, at least a near-real-time mode.

An improvement of the sea-floor equipments is planned in terms of the amount of contemporary active instruments, their interlink with “smart sensor” capacities (threshold detection, triggering) and the quality of the collected data. If the logistic conditions will allow it, a real-time data collection system is planned to be designed, built and deployed for a while (in cooperation with DAIMAR SME and Ifremer) in key locations of the SoM (Gulf of Izmit, Western High, Central High). The activities of planning, building, testing and deploying a new seafloor observatory may represent a reference for the future technology of submarine equipments, seafloor data collection, storage and transmission.

[1] Autonomous observatories are powered by batteries that need to be serviced on a regular basis, while permanent, cabled observatories are connected to a shore station by means of a cable, for power and real-time, data transmission.