This dataset contains the location (epicentre), origin time (UTC), magnitude, depth [km] and type of the last 100 earthquakes in and around the Netherlands. The location of the epicentre has an uncertainty of approx. 1 kilometer (horizontal). The type of earthquake distinguishes between natural (tectonic) earthquakes and induced earthquakes. The seismic network of the KNMI consists of geophones in boreholes (up to 300 m depth), accelerometers, and "broadband" seismometers.

In 2024, a total of 1,374 earthquakes with magnitude of five or more were recorded worldwide as of December that year. The Ring of Fire Large earthquakes generally result in higher death tolls in developing countries or countries where building codes are less stringent. China has suffered from a number of strong earthquakes that have resulted in extremely high death tolls. While earthquakes occur around the globe along the various tectonic plate boundaries, a significant proportion occur around the basin of the Pacific Ocean, in what is referred to as the Ring of Fire due to the high degree of tectonic activity. Many of the countries in the Ring of Fire, including Japan, Chile, the United States and New Zealand, led the way in earthquake policy and science as a result. The impacts of earthquakes The tragic loss of life is not the only major negative effect of earthquakes, a number of earthquakes have caused billions of dollars worth of damage to infrastructure and private property. The high cost of damage in the 2011 Fukushima and Christchurch earthquakes in Japan and New Zealand respectively demonstrates that even wealthy, developed countries who are experienced in dealing with earthquakes are ill-equipped when the large earthquakes hit.

The Cape Egmont Fault Zone in the southern Taranaki Basin, New Zealand, is a complex series of synthetic and antithetic dip-slip normal faults accommodating present-day extension. The fault zone comprises new and reactivated faults developed over multiple phases of plate boundary deformation during the last 100 Myrs. The fault zone is well imaged on petroleum industry seismic reflection data, with a number of faults exposed and studied onshore. The Cape Egmont Fault Zone is seismically active, with damaging historic earthquakes of up to MW 5.4. Most earthquakes occur beneath the Late Cretaceous to Holocene sedimentary sequence at depths greater than 5–8 km. The maximum depth of fault rupture is c. 20 km, above which 90% of recorded earthquakes occur. Focal mechanisms from these earthquakes generally indicate strike-slip to oblique-normal faulting, which contrasts with the predominantly dip-slip faulting observed in the overlying sedimentary sequence and surface fault traces. Data from regional earthquake studies and petroleum well deformation show faults imaged in the sedimentary sequence to be preferentially oriented for slip in the present-day stress field. The greatest earthquake risk is on major basement-penetrating crustal-scale faults greater than 20 km in length. Fault lengths and maximum vertical offsets of the sedimentary sequence, determined from a three-dimensional structural model, are consistent with global displacement-length scaling relationships. This validation permits fault lengths to be used to determine potential future earthquake magnitudes using global fault length-magnitude relationships. Fault lengths of post-Pliocene normal faults are typically ≤21 km, resulting in maximum predicted magnitudes MW 6.3. The most likely earthquake magnitude from the fault population sampled is MW 5.4 ± 0.5. The largest and most mature fault – the Cape Egmont Fault – is at least 53 km long and, depending on the number of segments ruptured during a future event, is capable of generating an earthquake between MW 7 and 7.3. DOI: https://doi.org/10.21420/ED9K-EP20 Cite data as: Seebeck H, Thrasher GP, Viskovic GPD, Macklin C, Bull S, Wang X, Nicol A, Holden C, Kaneko Y, Mouslopoulou V, Begg JG. 2021. Geologic, earthquake and tsunami modelling of the active Cape Egmont Fault Zone. Lower Hutt (NZ): GNS Science. 370 p. (GNS Science report; 2021/06). doi:10.21420/100K-VW73. (with data available at DOI: https://doi.org/10.21420/ED9K-EP20)

Epicenters of known M≥5 earthquakes from 1769 to 2016 are shown for California and a 100 km area bordering the state. Earthquakes are grouped by: M = 5-6; M = 6-7; M = 7+.

Stochastic finite-fault ground-motion prediction equations (GMPEs) are developed for the stable continental region of southeastern Australia (SEA). The models are based on reinterpreted source and attenuation parameters for small-to-moderate magnitude local earthquakes and a dataset augmented with ground-motion records from recent well-recorded moderate-magnitude earthquakes relative to those used in prior studies (Allen et al., 2007). The models are applicable for median horizontal-component ground-motions for earthquakes 4.0 <= MW <= 7.5 and at rupture distances less than 400 km. Careful analysis of well-constrained Brune stress drops indicates a dependence on hypocentral depth. It is speculated that this is the effect of an increasing crustal stress profile with depth. However, rather than a continuous increase, the change in stress drop appears to indicate a discrete step near 10 km depth. Average Brune stress drops for SEA earthquakes shallower and deeper than 10 km are estimated to be 23 MPa and 50 MPa, respectively. These stress parameters are subsequently input into the stochastic ground-motion simulations for the development of two discrete GMPEs for shallow and deep events. The GMPEs developed estimate response spectral accelerations similar to the Atkinson and Boore (2006) GMPE for eastern North America (ENA) at short rupture distances (less than approximately 100 km). However, owing to higher attenuation observed in the SEA crust (Allen and Atkinson, 2007), the SEA GMPEs estimate lower ground-motions than ENA models at larger distances. These differences become most obvious at distances greater than 200 km. A correlation between measured near-surface shear-wave velocity (VS30) and the site-dependent diminution term (K0) was developed from the limited data available to determine the average site condition to which the GMPEs are applicable. Assuming the correlation holds, a VS30 of approximately 820 m/s is obtained assuming an average path-independent diminution term K0 of 0.006 s from SEA seismic stations. Consequently, the GMPE presented herein can be assumed to be appropriate for rock sites of B to BC site class in the modified National Earthquake Hazards Reduction Program (Wills et al., 2000; Building Seismic Safety Council, 2003) site classification scheme. The response spectral models are compared against moderate-magnitude (4.0 <= MW <= 5.3) earthquakes from eastern Australia. Overall the SEA GMPEs show low median residuals across the full range of spectral period and distance. In contrast, ENA models tend to overestimate response spectra at larger distances. Because of these differences, the present analysis justifies the need to develop Australian-specific GMPEs where ground-motion hazard from a distant seismic source may become important.

Measurements of maximum trace amplitudes from 181 short- period vertical seismograms recorded at hypocentral distances of 3-1500 km from 36 earthquakes in the magnitude range 0.8 - 4.3 were used to derive a new preliminary ML scale for southeastern Australia ML = log A + (1.34±0.09)log(R/100) + (0.00055±0.00012)(R-100) + 3.13 + S where ML is local magnitude, A (mm) is equivalent Wood-Anderson trace amplitude not corrected for the measurement having been made on a vertical component, R (km) the hypocentral distance and S the station correction.

Epicenters of known M≥5 earthquakes from 1769 to 2016 are shown for California and a 100 km area bordering the state. Earthquakes are grouped by: M = 5-6; M = 6-7; M = 7+.

This data set provides Modified Mercalli Intensity maps for the Hayward earthquake of October 21, 1868. To construct the Modified Mercalli Intensity (MMI) ShakeMap for the 1868 Hayward earthquake, we started with two sets of damage descriptions and felt reports. The first set of 100 sites was compiled by A.A. Bullock in the Lawson (1908) report on the 1906 San Francisco earthquake. The second set of 45 sites was compiled by Toppozada et al. (1981) from an extensive search of newspaper archives. We supplemented these two sets of reports with new observations from 30 sites using surveys of cemetery damage, reports of damage to historic adobe structures, pioneer narratives, and reports from newspapers that Toppozada et al. (1981) did not retrieve.

[Summary provided by the USGS.]

This data publication contains (i) a slab model of the Cascadia subduction zone, derived from receiver functions, parameterized as depth to the three interfaces: t (top), c (central) and m (Moho), in NetCDF format; (ii) the station measurements of all parameters in the model in tabular and Raysum model file format; (iii) the raw receiver functions in SAC format; and (iv) auxiliary scripts for loading and plotting the data. A total of 45,601 individual receiver functions recorded at 298 seismic stations distributed across the Cascadia forearc contributed to the slab model. For each station, 100 s recordings symmetric about the P -wave arrival (i.e. 50 s noise and 50 s signal) of earthquakes with magnitudes between 5.5 and 8, in the distance range between 30 and 100 degree, were downloaded from the Incorporated Research Institutions for Seismology (IRIS) data center, the Northern California Earthquake Data Center (NCEDC), and the Natural Resources Canada Data Center (NRCAN). After quality control, radial and transverse receiver functions were computed through frequency-domain simultaneous deconvolution, with an optimal damping factor found through generalized cross validation. The continental forearc and subducting slab were parameterized as three layers over a mantle half-space, with the subduction stratigraphy bounding interfaces labeled as t (top), c (central) and m (Moho). Synthetic receiver functions were calculated through ray-theoretical modeling of plane-wave scattering at the model interfaces. The thickness, S -wave velocity (VS) and P - to S -wave velocity ratio (VP/VS) of each layer, as well as the common strike and dip of the bottom two layers and the top of the half space (in total 11 parameters) were optimized simultaneously through a simulated annealing global parameter search scheme. The misfit was defined as the anti-correlation (1 minus the cross-correlation coefficient) between the observed and predicted receiver functions, bandpass filtered between 2 and 20 s period duration. In total, 171, 143 and 137 quality A nodes were determined to constrain the t, c and m interfaces, respectively. At the trench, 105 nodes at 3 km below the local bathymetry were inserted to constrain the t and c interfaces, and at 6.5 km deeper to constrain the m interface, representing typical sediment and igneous crustal thicknesses. A spline surface was fitted to these nodes to yield margin-wide depth models. The spline coefficients were found using singular value decomposition, with the nominal depth uncertainties supplied as weights. The solution was damped by retaining the 116, 117, and 116 largest singular values for the t, c and m interfaces, respectively, based on analysis of L-curves and the Akaike information criterion. The data set is the supplemental material to Bloch, W., Bostock, M. G., Audet, P. (2023) A Cascadia Slab Model from Receiver Functions. Geochemistry, Geophysics, Geosystems.

New earthquake risk maps of southwest Western Australia including continental margins have been prepared. The risk is depicted as contours of peak ground velocity, acceleration, and ground intensity with a 10 per cent probability of being exceeded in 50 years. The maps are based on the Cornell-McGuire methodology. Ten earthquake source zones have been thus defined and corresponding recurrence relations derived. The relation obtained, using a maximum-likelihood fit, for the primary zone to the east of Perth, is log N = 3.66-0.90 ML, where N is the number of events greater than or equal to the Richter magnitude, ML. Local intensity attenuation constants, a, b, and c, are derived for the expression I = aebMLR-c, where I is the estimated Modified Mercalli intensity at a hypocentral distance R km from an earthquake of magnitude ML. Using the relation log A= I/3.1-2.3 to convert the intensity to peak ground acceleration, A in m.s-2, the adopted constants were 0.025, 1.10 and 1.03 respectively. Similarly, using the empirical formula 21 = 7v/5 to convert intensity to peak ground velocity, v in mm.s-1, the corresponding values were 3.30, 1.04 and 0.96, respectively. The contour expressing the greatest risk in the area of interest is that of a peak ground velocity of 160 mm.s-1, and it encloses an area of about 2000 km2 centred on the most active source zone east of Perth. The value for Perth city is 48 mm.s-1. Increasing (i) the maximum magnitude from ML 7.5 to ML 8.5; (ii) the depth of earthquake foci from 5 km to 15 km; (iii) the b value from 0.90 to 0.94; and (iv) the attenuation constants to their estimated maximum value, in the primary source zone, changes the Perth velocity contour from 48 mm.s-1 to 56 mm.s-1 , 48 mm.s-1 , 42 mm.s-1 , and 58 mm.s-1, respectively. The omission of a suspected seismic gap 100 km east of Perth from the primary source zone changes the velocity contour from 48 mm.s-1 to 46 mm.s-1. Sensitivity to adopting another empirical relation between peak ground acceleration and intensity has been examined. This increases the risk at Perth from 0.44 m.s-2 to 0.65 m.s-2. We recommend a microzonation study of Perth and installation of more strong-motion instruments to improve our risk estimates, which should be updated in 5-10 years.

In plate boundary regions moderate to large earthquakes are often sufficiently frequent that fundamental seismic parameters such as the recurrence intervals of large earthquakes and maximum credible earthquake (Mmax) can be estimated with some degree of confidence. The same is not true for the Stable Continental Regions (SCRs) of the world. Large earthquakes are so infrequent that the data distributions upon which recurrence and Mmax estimates are based are heavily skewed towards magnitudes below Mw5.0, and so require significant extrapolation up to magnitudes for which the most damaging ground-shaking might be expected. The rarity of validating evidence from surface rupturing palaeo-earthquakes typically limits the confidence with which these extrapolated statistical parameters may be applied. Herein we present a new earthquake catalogue containing, in addition to the historic record of seismicity, 150 palaeo-earthquakes derived from 60 palaeo-earthquake features spanning the last > 100 ka of the history of the Precambrian shield and fringing extended margin of southwest Western Australia. From this combined dataset we show that Mmax in non-extended-SCR is M7.25 ± 0.1 and in extended-SCR is M7.65 ± 0.1. We also demonstrate that in the 230,000 km2 area of non-extended-SCR crust, the rate of seismic activity required to build these scarps is one tenth of the contemporary seismicity in the area, consistent with episodic or clustered models describing SCR earthquake recurrence. A dominance in the landscape of earthquake scarps reflecting multiple events suggests that the largest earthquakes are likely to occur on pre-existing faults. We expect these results might apply to most areas of non-extended-SCR worldwide.

This dataset comprises data from three seismic stations deployed across the Bransfield Strait, Antarctica, at Livingston Island, Deception Island, and Cierva Cove, from August 2014 to May 2015. Stations were originally deployed in February 2008 to assess the local seismic activity at the selected sites, and to investigate the crustal structure across the Bransfield Strait. More details can be found in Carmona et al (2014).

The selected period contains earthquake data from a seismic swarm that occurred in the Bransfield Strait, near Livingston Island (Almendros et al. 2015, 2018). This swarm comprises several thousand earthquakes with magnitudes up to 4.6 and event rates up to 180 earthquakes per day. More information can be found in Almendros et al. (2018).

DATA FORMAT Raw seismograms in miniseed format (e.g. http://ds.iris.edu/ds/nodes/dmc/data/formats)

DIRECTORY STRUCTURE Data are single-channel, one-day-long seismograms. Files are named and stored in a directory structure following the SEISCOMP3 SDS convention (http://www.seiscomp3.org/wiki/doc/applications/slarchive/SDS): ROOTDIR/YYYY/NET/STAT/CHAN.TYPE/NET.STA.LOC.CHAN.TYPE.YYYY.DDD ROOTDIR is the base folder where the data are stored, YYYY is the year, NN is the network code, STAT is the station name, CHAN is the code to identify the station channels, TYPE is a single character indicating the data type (set to D for waveform data), LOC is the location name (left empty), and DDD is the day of year.

STATION NAMES CCV Cierva Cove; DCP Deception Island; LVN Livinston Island

STATION COORDINATES CCV 60.9667W 64.1500S; DCP 60.6833W 62.9833S; LVN 60.3833W 62.6667S

INSTRUMENTS All three stations are equipped with a 16-s, three-component EENTEC 4000 seismometer, and a 24-bit EENTEC DR-400 data acquisition system sampling at 100 sps. More information can be found in Carmona et al (2014).

RECORDING PERIODS CCV 2014-08-01 to 2015-05-10; DCP 2014-08-01 to 2015-05-10; LVN 2014-08-01 to 2015-04-02

SUPPORTING PROJECTS POL2006-08663 (CORSHET project, Spanish Ministry of Education); CTM2009-07705, CTM2009-08085, CTM2010-11740, CTM2011-16049 (Spanish Ministry of Science and Education); Research Grant IGME1198 2014-2015, 2015-2016, 2016-2017, 2017-2018 (Spanish Polar Committee, Spanish Ministry of Economy); CTM2016-77315 (BRAVOSEIS project, Spanish Ministry of Economy)

DATA QUALITY Timing errors have been occasionally detected at these stations due to GPS malfunction. We have corrected them using waveform correlations, although time has to be considered uncertain (within a fraction of a second) for the following station and period: LVN 2014-08-01 to 2014-11-30

REFERENCES Carmona, E., Almendros, J., Martín, R., Cortes, G., Alguacil, G., Moreno, J., Martin, B., Martos, A., Serrano, I., Stich, D., Ibanez, J. M. , Advances in seismic monitoring at Deception Island volcano (Antarctica) since the International Polar Year , Annals of Geophysics 57(3), SS0321, doi:10.4401/ag-6378 (2014) Almendros, J., Carmona, E., Jimenez, V., Diaz-Moreno, A., Lorenzo, F., Berrocoso, M., De Gil, A., Fernandez-Ros, A., Rosado, B. , Deception Island: Sustained deformation and large increase in seismic activity during the 2014-2015 survey , Bulletin of the Global Volcanism Network, 40(7) (2015) Almendros, J., Carmona, E., Jimenez, V., Diaz-Moreno, A., Lorenzo, F., Volcano-tectonic activity at Deception Island volcano following a seismic swarm in the Bransfield Rift (2014-2015), submitted for publication (2018).

On August 17, 1999, at 3:01 A.M. local time (00:01:39.8 UTC) a magnitude (Mw) 7.4 earthquake occurred along the westernmost North Anatolian fault. The earthquake epicenter was 11 km (7 mi) southeast of the City of Izmit, in the sub-province of Kocaeli, a densely populated area in the industrial heartland of Turkey, and less than 80 km southeast of Istanbul.The Anatolian fault, a right lateral strike-slip fault, has a history of earthquakes. (A right-lateral, strike-slip fault is one in which the motion of the opposite side of the fault, as one looks across the fault, is to the right.) In the last sixty years, there have been eleven earthquakes with magnitudes (Ms) larger than 6.7 along this fault. The August 1999 quake was located in a seismic gap between areas of the fault that had broken in 1967 and in 1963. The maximum displacement along the fault was more than five meters. The total rupture length was nearly 170 km. Accelerograms showed that the earthquake consisted of two major events located about 30 km apart.The earthquake damaged buildings across seven provinces for a distance of 250 km from Istanbul to Bolu. As many as seventy percent of the buildings in portions of the cities of Adapazari, Golcuk, Izmit, Topcular, and Kular were severely damaged or collapsed. Nearly all the fatalities and injuries can be attributed to building collapse. An estimated 60,000-115,000 buildings were damaged or destroyed. Damage - estimates range from $10 billion to $40 billion. The fault crossed some of the most densely populated regions of Turkey. The affected population numbered 15 million people. Casualties totaled 17,000 and additional thousands were missing and presumed dead. Injuries numbered 24,000 and 500,000 people were left homeless with 200,000 living on the streets. The economics of the damaged region presented ten percent of the Gross National Product of Turkey.Structural damage occurred in several ways. Many structures were deformed or destroyed by the lateral and vertical offsets of the fault itself. Several apartment buildings were torn apart by the fault rupture and collapsed. There was great variation of response even among similarly constructed buildings in the same neighborhood, with some collapsing and others having moderate or little apparent damage. Residential buildings were usually three to seven stories in height. The predominant structural type in Turkey consists of reinforced concrete frames with unreinforced masonry infill (brick walls filling the gaps between concrete frames). These infill walls tended to fall out with the earthquake shaking, adding stress to the beams and columns.Failures occurred in foundations, and in soft stories. These are stories that have few supporting walls and are often found in the first story of a structure. They may be open stories in order to accommodate shops or a garage, and may have few walls supporting the second story floor. Failures also occurred in weak columns paired with strong beams, and in columns with lack of detailing and column confinement. Buildings having four or more stories were much more likely to be damaged or to collapse since the buildings with greater height incurred a greater amount of displacement at the top relative to the bottom of the buildings. In some areas, buildings were subjected to damaging ground settlement, liquefaction, or subsidence and inundation from lake waters. Hundreds of buildings in the city of Adapazari settled, tipped, or toppled as liquefaction weakened lakebed sediments. In the severe liquefaction areas, more than sixty percent of multi-story buildings suffered partial or total collapse due to structural failure. With the exception of the most heavily damaged areas, the water system was functional in six days or less. There was no significant damage reported to dams or reservoirs. Wastewater pipelines were heavily damaged. Electric power was disrupted when buildings collapsed onto electrical lines, electrical poles tilted or collapsed, and transformers tilted due to support failure. However, the electric power service was nearly restored in two weeks. Some highway bridges were damaged when fault rupture occurred beneath them. Only a few residential fires were reported, since most building materials were fire-resistant, and there are no natural gas pipelines in the region. Following the earthquake, fire broke out at the large Tupas Oil Refinery in Korfez. One fire at the refinery resulted when a collapsed 90-meter reinforced concrete stack knocked down equipment and pipeways. An oil spill from the refinery contaminated the waters of Izmit Bay. Loss of electrical power, debris on the roads, and lack of water hampered firefighting efforts. A number of damaging quakes have occurred in this area in the past. In 1754, the area near Izmit experienced an earthquake that resulted in 2,000 deaths. In 1967, a 7.1 magnitude earthquake occurred near Izmit. In 1970, an earthquake near Gediz, 160 km (100 mi) to the south, killed 1,000 people. This area of Turkey will continue to experience seismic activity. Appropriate steps need to be taken now to minimize the effects and fatalities of the next earthquake.

This digital database comprises seismological data recorded during the EC NEAREST project by a 3 component broadband seismometer. The sensor was installed on the GEOSTAR deep-sea multidisciplinary observatory, in the Gulf of Cadiz (Portugal). In the same vessel together with the seismometer there was an accelerometer. Both sensors were connected to a 24 bit digitizer, and their time reference came from an external rubidium clock connected to the digitizer.

This award supports a project to strengthen collaborations between the various research groups working on iceberg calving. Relatively little is known about the calving process, especially the physics that governs the initiation and propagation of fractures within the ice. This knowledge gap exists in part because of the diverse range in spatial and temporal scales associated with calving (ranging from less than one meter to over a hundred kilometers in length scale). It is becoming increasingly clear that to predict the future behavior of the Antarctic Ice Sheet and its contribution to sea level rise, it is necessary to improve our understanding of iceberg calving processes. Further challenges stem from difficulties in monitoring and quantifying short-time and spatial-scale processes associated with ice fracture, including increased fracturing events in ice shelves or outlet glaciers that may be a precursor to disintegration, retreat or increased calving rates. Coupled, these fundamental problems currently prohibit the inclusion of iceberg calving into numerical ice sheet models and hinder our ability to accurately forecast changes in sea level in response to climate change. Seismic data from four markedly different environmental regimes forms the basis of the proposed research, and researchers most familiar with the datasets will perform all analyses. Extracting the similarities and differences across the full breadth of calving processes embodies the core of the proposed work, combining and improving methods previously developed by each group. Techniques derived from solid Earth seismology, including waveform cross-correlation and clustering will be applied to each data set allowing quantitative process comparisons on a significantly higher level than previously possible. This project will derive catalogues of glaciologically produced seismic events; the events will then be located and categorized based on their location, waveform and waveform spectra both within individual environments and between regions.

The Indonesian earthquake of 19 August 1977, with a magnitude (M), 8 was felt in Western Australia up to distances of 2600 km. The maximum ground intensity felt was MM V in northwest towns up to 1100 km from the epicentre. The ground intensity in Perth, 2300 km from the epic entre, was MM III or less. Resonance of multi-storey buildings resulted in an eight-fold amplification of peak ground acceleration on the upper floors. Only minor damage occurred in Perth. Seismic sea waves up to six metres in height were reported several hours after the earthquake at towns along the northwest coast. There were no reports of damage associated with these waves. They arrived along the coast near low tide, otherwise there could have been some flooding. An examination of data from the Sunda arc region, for the period 1900-1977, suggests that about eight earthquakes of magnitude greater than 7.5 could occur in this region every hundred years. Of these, two or three may be felt in Western Australia. The 1977 earthquake was the largest and closest to Western Australia since 1900 and its effects are the maximum likely to be experienced from this region. Consideration should be given to installing accelerographs in selected buildings in Perth.

This digital database comprises seismological data recorded continuously during the MARMARA-DM 1 project by a 3 component broadband seismometer. The sensor was installed on a GEOSTAR-class deep-sea multidisciplinary observatory, deployed in the Marmara Sea at a depth of 167 m. The seismometer is installed in a dedicated vessel integrated in a separate structure connected to the observatory via a special mechanical release. To guarantee a good coupling with the sea bottom, the structure is disconnected just after the observatory touch-down and kept linked to the frame by a slack rope (Favali et al., 2006). The sensor was connected to a 24 bit digitizer, and its time reference came from an external rubidium clock connected to the digitizer.

This digital database comprises seismological data recorded during the GNDT 1 project by a 3 component broadband seismometer, together with a hydrophone. The sensor was installed on the GEOSTAR deep-sea multidisciplinary observatory, offshore of the Eastern coast of Sicily (Southern Italy) at a depth of 2071 m. The seismometer is installed in a dedicated vessel integrated in a separate structure connected to the observatory via a special mechanical release. To guarantee a good coupling with the sea bottom, the structure is disconnected just after the observatory touch-down and kept linked to the frame by a slack rope (Favali et al., 2006). The sensor was connected to a 24 bit digitizer, and its time reference came from an external rubidium clock connected to the digitizer.

This digital database comprises data recorded from the hydrophone during the GNDT 2 project. The hydrophone was installed on the frame of the GEOSTAR deep-sea multidisciplinary observatory, deployed offshore of the Eastern coast of Sicily (Southern Italy) at a depth of 2065 m.The sensor was connected to a 24 bit digitizer, and its time reference came from an external rubidium clock connected to the digitizer, and synchronized with the seismometer.