AdriaArray - Data analysis and research

Collaborative Research Groups (CRGs)

According to the Memorandum of Collaboration (MoC), the scientific work is accomplished by so-called Collaborative Research Groups (CRGs).

  • CRGs are suggested by participants,

  • CRGs work independently,

  • Themes of CRGs are discussed and agreed upon by the AdASG Steering Committee,

  • CRGs report on activities at AdASG Workshops,

  • CRGs must consist of researchers from at least two AdASG members and from different countries,

  • Participants may join one or several CRGs,

  • If possible, the individual CRGs are coordinated by an experienced researcher and an early career scientist,

  • Participants in CRGs may change any time upon notice to the coordinators of the CRGs.

Working Group 5 of the AdASG entitled ‘Scientific co-operation’ will coordinate and support the establishment of CRGs. First suggestions of the CRGs have been proposed during the second international workshop in Potsdam, third international workshop in Dubrovnik & EGU 2023 AdriaArray Splinter meeting.

Now, AdriaArray participants are invited to contact the initial contacts of CRGs (given below) to join one or several Collaborative Research Groups. Initial contacts will organize the reviews of the topic, the elections of chairs and co-chairs, and maintain a list of interested participants.

Initial contacts will also report the outcome of these discussions to Working Group 5 of the AdASG. Suggestions of CRGs will be reviewed by Working Group 5 before presentation to and approval by the Steering Committee.

In the following we summarize the discussion on potential topics of CRGs.

CRG ‘Seismicity & Seismic Sources’

  • Chairs: G. Petersen. & E. Sokos.

  • Membership secretary: F. Tilmann maintains a register of participants and a mailing list, which will reach all members of the CRG.

Description of topic

Main topics of the CRG are the detection and location of earthquakes, the determination of source mechanisms, either in real-time or offline during detailed in-depth studies, as well as approaches for statistical analysis and interpretations of seismotectonics.

The CRG Seismicity & Sources includes multiple subgroups that were established to coordinate the collaborative work. These subgroups are temporary and subject to changes based on needs and ideas for collaborations.

The CRG Seismicity & Seismic Sources is split into four subgroups to coordinate the collaborative work:

  1. Survey of monitoring practices at national agencies.

  2. Parametric data exchange with EMSC, station naming and reporting: C. Papazachos & C. Evangelidis.

  3. Machine learning for seismicity detection and location & crowd processing: D. Sindija & F. Tilmann.

  4. Moment tensors and seismic sources: G. Petersen & E. Sokos. The AdriaArray Seismic Network produced a huge increase in available real time data for source studies in local and regional distances. The subgroup deals with the usage of this data in real time moment tensor or focal mechanism retrieval applications, as well as in detailed source studies of earthquakes across different magnitude scales, in the Adria area (e.g. extended source studies, slip inversions etc). Goals are the exchange of knowledge and methods, fostering discussions of best practices, exchanging results and collaborating in research projects.

CRG ‘Body wave tomography’

Description of topic and first tasks to be tackled

The large station array and huge high-quality data stream, produced earlier by the AlpArray and now by AdriaArray Seismic Networks, provide a rich and unique dataset for body wave tomography studies.

Body wave arrival times determination

To optimize the exploitation of the size and quality of these dataset we need to employ semi- or fully automated picking procedures able to reproduce in the best way the careful, handy re-picked tasks of the experienced seismologists.

This activity could include:

  1. surveying and benchmarking of the existing automated picking tools for regional and teleseismic phases;

  2. surveying, checking the consistency and sharing the available carefully picked datasets that can be used for comparison and/or learning phases with automated picking procedures;

  3. comparing picks for a number of selected earthquakes at all AdriaArray stations;

  4. creating a common dataset for one or more AdA sub regions to compare different picking and different inversion codes;

  5. optimizing automatic picking algorithms and possibly providing training and application on the use of software tools.

Traveltime tomography and synthetic tests to assess the model resolution

Studying the crust and upper mantle velocity structure beneath the AdriaArray target region will involve various inversion approaches to process the extensive arrivals dataset (including both linearized and non-linearized methods). An essential final stage is evaluating the model resolution. We propose relying on our collective expertise to ensure a uniform evaluation of model performance.

This involves:

  1. Creating and testing synthetic models: collaboratively develop simplistic or realistic synthetic models reflecting expected geological structures (i.e. standardised benchmarks for testing and comparison).

  2. Establishing shared testing guidelines: work together to devise guidelines that standardised testing procedures across the board.

  3. Comparing model performance: assess the ability of different models to reproduce observed waveforms accurately.

This task could be addressed in cooperation with other CRGs (surface waves, waveform inversion, bodywaves from ambient noise, linking geophysical observables).

CRG ‘Modeling of seismic wave propagation and full waveform inversion’

Description of topic and first tasks to be tackled

AdriaArray offers the unique opportunity to measure wavefields at regional scale in a tectonically highly variable region encompassing multiple subduction zones, various orogenic belts, back-arc basins, and volcanic fields. Accurate forward modelling of wave propagation in strongly heterogenous and anisotropic structures is needed to understand the influence of slabs, Moho and lithosphere-asthenosphere topography and mantle flow on seismic wave propagation. Accurate forward modelling stands at the basis for advanced seismic analysis methods including full waveform inversion. At first, we will review existing forward modelling methods. Moreover, we aim at benchmarking existing codes and at the development of advanced inversion methods to invert for crustal and mantle structure in the AdriaArray region. In addition, we will provide training for the use of existing and new modelling and inversion codes.

CRG ‘Receiver functions’

Description of topic and first tasks to be tackled

This CRG will provide support to AdriaArray members who work on receiver function studies as well as those who want to use receiver function results in their research. In terms of methodologies, the CRG will provide access to harmonized receiver function analysis tools and training on how to use these. It will also serve as a platform to share software and tips on data analysis, and to carry out community-based benchmarking tests. In terms of imaging targets, the CRG will help coordinate efforts to avoid potential overlaps and foster collaborations. At the end of the project, the CRG will produce a harmonized set of receiver function products for the entire AdriaArray network.

CRG ‘Ambient noise’

Description of topic and first tasks to be tackled

In this overarching research group, we deal with various aspects of ambient noise, starting from an analysis of the various sources contributing to the noise field, over the compilation of a cross-correlation database to the application of various imaging methods using these cross-correlations.

Noise sources - We will work on the characterization and localization of natural noise variation, resulting from e.g., local weather effects like wind or air pressure and anthropogenic noise sources, e.g., power plants, rotating machines. We also study variations of primary, secondary, and local microseism over time.

Database - We will deconvolve and downsample the continuous seismic data from the various subnetworks participating in AdriaArray. This data (around 4-6 TB) will be made available to anybody interested in computing cross-correlation functions for various purposes from mapping ambient noise sources to tomography or time lapse monitoring. If requested, we might also offer a cross-correlation database following a standard preprocessing scheme.

Imaging - Besides the obvious surface wave imaging methods, recent advancements in passive noise interferometry have shown promising results in the retrieval of body waves from noise correlations. This primarily includes various body-wave reflection phases generated by dominant subsurface discontinuities, such as the Basin sediment/bedrock interface, the Moho interface, and the 410-km and 660-km mantle transition zone interfaces. The information that provided is particularly valuable for determining the depth undulations of subsurface discontinuities and can be seamlessly integrated into seismic tomography for a more comprehensive understanding of the deep earth structure. Given its remarkable data quality and data coverage, the AdriaArray seismic network presents an ideal setting for the implementation and development of such cutting-edge seismic techniques.

CRG ‘Surface wave tomography’

Description of topic and first tasks to be tackled

This CRG aims at gathering multiple methodologies that use surface waves (Rayleigh and/or Love) for 2D and 3D imaging of the crustal and upper mantle structure in the region.

Following traditional frameworks of surface wave tomography (SWT), the work handled within this CRG targets four main collaborative tasks:

  1. the construction of reliable surface wave datasets from earthquake data and/or ambient noise;

  2. the measurement of the surface wave phase/group travel times and amplitudes for constructing 2D maps of isotropic (and anisotropic) Rayleigh/Love velocities;

  3. the depth-inversion of phase (and group) velocity dispersion curves;

  4. the joint inversion with other observables.

This group provides a platform for setting up and sharing databases, method testing and benchmarking, resolution test analysis, and further discussions on methodological developments. It includes linearized inversion method, Bayesian approaches, Eikonal/Helmholtz tomography, 2D phase velocity map inversions or direct 3D approaches, radial and azimuthal anisotropy and many others.

CRG ‘Shear-wave splitting and anisotropy’

Description of topic and first tasks to be tackled

After we collected all available seismic anisotropy measurements for the study region, we identified gaps to be filled with new data. New measurements can give the opportunity to have benchmarking methods & training. We expect to produce new shear wave splitting measurements and splitting intensity values, to be obtained using multiple seismic phases and by analysis methods including joint inversions. We will also provide interpretations of the underlying seismic anisotropy structure within the geodynamic context of the Adria subduction system and eastern Alps. An important objective is also to interact with other CRGs focusing on body and surface wave tomography as well as receiver functions to share and integrate any new data set and results. The work done by each member of the CRG is an advantage in adding one or more pieces of the puzzle of the anisotropy structure of the tectonic environment of the study area. Therefore, active communication between CRG members is encouraged and appreciated. In this regard, updated feedback from CRG members on their progress in collecting and processing shear wave splitting data would be appreciated.

CRG ‘Engineering Seismology’

CRG ‘Linking geophysical observations and geodynamics’

Description of topic and first tasks to be tackled

Using AdriaArray data, the interior of the Adriatic plate and its margins, slabs and slab windows as well as upper mantle flow will be imaged to clarify open questions regarding the driving forces of plate deformation and kinematics. To test the resolution capabilities of imaging methods and to design input models for numerical geodynamic experiments, at first existing hypotheses of lithospheric and upper mantle structure are to be described in digital form. Hypotheses of slabs and slab windows along the margins of the Adriatic plate (Alps, Apennines, Calabrian Arc, Carpathians, Dinarides, Hellenides) are to be reviewed and discussed based on available observables like seismicity, Moho maps, tomographic models, and receiver function images. Different hypotheses for the slab interface are to be provided in digital format. For these hypotheses, consistent 3D models of various parameters (seismic velocities, temperatures, densities, composition, viscosity) are to be set up using thermomechanical modelling. They form the basis for numerical geodynamic modelling of quantities like plate kinematics, stress, strain fields or exhumation rates that are to be compared with field observations and the geological record. Experts from various relevant disciplines are invited to join the CRG.