Work package number: WP5

Work package title: Real- and quasi-real-time Earthquake &Tsunami Hazard Monitoring

WP Leader: KOERI


The main aim of this WP is to implement an integrated approach by harmonizing geodetic and seismic data to be used in early warning applications, such as fast centroid moment tensor inversion and rapid slip inversion, so that in addition a quick determination of the rupture characteristics could also assist the identification of the tsunamigenic potential of an earthquake in combination with a tectonic origin tsunami scenario database. This integrated approach will provide a unique performance compared with only seismic or geodetic networks and false alarm will be minimized for Earthquake and Tsunami warnings. Together with a finite source description and calibration of available geodetic and seismic data, a rapid and quantitative Shake map generation scheme will be provided. For this purpose, data processing techniques and computing algorithms will be investigated and designed by making full use of the monitoring system of the MARsite. Rapid source information (not only hypocenter, magnitude but also, for example, rupture directivity) is vital, particularly for damage estimation in the configuration of the Marmara region and Istanbul with respect to the NAFZ, and especially the expected large earthquake. From this point of view inversion of geodetic data immediately enhances the speed and accuracy of the preliminary damage maps.  It is intended to create high-resolution geodetic/seismic infrastructure, at a test site, to receive real-time GPS and Seismic data and is to develop and improve analysis techniques and methods to develop PGV shake maps that is obtained with automated finite-source inversion results. Taking into account the continuous geodetic (GPS) data will be a key issue. High resolution geodetic/seismic infrastructure will be installed (update of 26 GPS stations by cGPS with co-deployment of strong-motion sensors) in this area to provide real-time data necessary for the finite-source inversion (Task 1), rapid finite source inversion tools will be improved (Task 2), ground-motion simulation tools will be calibrated (Task 3), a scenario database for both for seismic and tsunami hazard evaluation will be created (Task 4), the final hazard map by taking into account uncertainties  and ground-motion variability will be improved (Task 5)  and finally long and short-term earthquake forecast maps will be developed (Task 6).

Description of work

Task 1. Establishment of appropriate infrastructure (particularly for GPS and strong motion stations in Marmara Region) to obtain real time data

Two objectives of this task are to update 16 of 26 GPS sites in order to establish real-time data transmission and installation of strong ground motion instruments.  The acquisition and harmonization of real time GPS and Strong Motion time series will provide excellent time resolution of real time earthquake monitoring and also provide to measurements of tectonic strain accumulation across the Marmara Fault zone. These measurements will also enable a quick determination of the rupture characteristics to assist the identification of the tsunamigenic potential of an earthquake. In other words, this refined and newly established infrastructure and combined GPS and Seismic real time data will contribute to develop real-time applications that allow to closely and rapid monitor earthquake processes and tsunami assessment.

The next-generation geodetic and seismic data can be used for EEW applications, fast centroid moment tensor inversion and rapid slip inversion. This infrastructure will also opportunity to research unknown fault parameters and decisive contribute to refinement of the seismic hazard map for this important region. This integrated approach will provide a unique performance compared with only seismic or geodetic networks and false alarm will be minimized for Earthquake and Tsunami warnings. Currently, the existing GPS monitoring arrays do not provide real-time data because of the absence of proper continuous power and communication infrastructure. All of the GPS sites (26) in Marmara Region will be improved upon by the installation of power source unit, (GPRS or Satellite). The output of this Task will be real time GPS and Seismic time series and will be an input of the Task 2, 3 and 4.

Task 2. Near real-time determination of the earthquake finite-fault source parameters and models, based on GPS and strong motion data

Information about the extended source properties are needed for performing the ground motion simulation associated to the earthquake rupture on the causative fault. The main goal of this task is the fast determination of the earthquake source, with special focus on its finite-fault characteristics.

Analyzing geodetic and seismic data together using a Kalman filter will provide precise and true broadband record of displacements across the entire frequency range, including the static component. These analyses can be done in near real time and are particularly suited for capturing near-source large earthquakes.

In order to improve rapid ground-motion simulations in case of large earthquakes in the Marmara region, a new tool for rapid reconstruction of the rupture process of large earthquakes using near-field strong-motion and high-rate GPS data will be developed. The array-seismological method will be extended by taking into account empirical or synthetic Green’s functions. Aim of such method will be providing a fast and reliable estimation of most relevant source parameters (e.g. moment magnitude, fault size, rupture duration, slip centroid) rather than achieving a high spatio-temporal resolution.

Furthermore, two finite-fault inversion programs will be developed, tested and compared. The first code implements a linear technique to invert strong motion and GPS data: it is very fast and produces a model of the earthquake rupture process in term of heterogeneous slip distribution, uniform rise time and constant rupture velocity. The second code is based on a simulated annealing technique: it is slower than the former, but it may handle very complicated rupture model with heterogeneous slip, rise time and rupture velocity, other than several kind of source time function. The performance of the above codes will be assessed in term of accuracy of the solution and quickness of the execution run through several synthetic tests, specifically designed for the Marmara Sea tectonic setting and observational network.

Task 3. Generation of a routine for simulation of strong ground motion based on integrated data

This task aims to establish the rapid PGV Shake maps through numerical simulations by integrating the various data (GPS and Strong Motion). Most of the automated Shake Map applications are primarily based on point source approximations; however finite source effects are significant for major earthquakes. First of all, in order to understand the variability of the ground motions, we intend to introduce a deterministic-stochastic finite source description (INGV). In parallel, we are going to optimize the existing numerical codes (e.g. finite difference, spectral element) in order to adjust the parameters requested from the expected PGV map resolution and rapidity (BRGM-KOERI). The numerical tools are to be available for further use. Finite source models are obtained from the geodetic and seismic data rapidly for major earthquakes (Tasks 1 & 2). Such information (hypocentre, fault dimension, rupture directivity and velocity, and some more) should be integrated in the simulations, and we examine the rapidity and the precisions of such rapid numerical PGV maps, which are useful for further earthquake and tsunami rapid information infrastructure (Tasks 4, 5 & 6). Multiple windows simulation techniques improved with site correction make it possible to simulate strong ground motion for generation of PGV ShakeMaps.  The suitability and sensitivity of the inversion and simulation scheme for producing rapid Shake Maps will be tested by well-recorded earthquakes; however there are limited numbers of medium to large earthquakes that are recorded with GPS receivers. Among them 2004 Parkfield earthquake data can be utilized as test data. This earthquake was recorded at thirteen 1-Hz GPS receivers and several strong motion instruments. Simulated models will also be compared with distribution of ground shaking intensity provided by available Earthquake Early Warning Algorithms.

Task 4. Creating a scenario database for earthquake triggered tsunamis and Testing of the routine with well-studied events

The goal of the present task is to build up a detailed scenario database for all possible earthquakes in the Marmara Sea with a tsunamigenic potential. Due to the very short travel times in Marmara Sea, a Tsunami Early Warning System (TEWS) cannot rely on real-time calculations and has to be based on a pre-computed tsunami scenario database to be queried in real-time, basing on the initial determination of earthquake hypocentre and Magnitude, but also on dislocation models calculated from real-time inversion of geodetic and seismic data (from Tasks 2 and 4), similarly to e.g. the GI-TEWS in Indonesia. Such a database could be inspired to that implemented in the Japanese TEWS, which nonetheless will be adapted to Marmara region. The Marmara region will be divided in grid areas of 0.1°x0.1° and tsunami scenarios will be created for each bin, where the bin centre will be characterized as the epicentre location. Earthquake source parameters will be defined based on a study of characteristic source parameters in the region, supported with historical and statistical studies. A decision support system for the TEWS should also be supported with offshore tsunameters and or hydrophone/pressure meters, if possible. In the presence of these tsunami data, together with seismic and geodetic data, the database might be also used as a set of Green’s functions for recovering the tsunami source via real time inversion with the aim of constraining the tsunami forecast. Moreover, these Green’s functions could be used for long-term probabilistic tsunami hazard assessment (PTHA) if earthquake recurrence times will be provided by Task 6 in this WP.

Task 5. Improvement of the probabilistic seismic hazard assessment by taking into account uncertainties and ground-motion variability

Probabilistic seismic hazard analysis (PSHA) is de­fined as evaluation of the probability or likelihood that there will be ground motion in excess of certain levels during a specific time period. The basic analytical procedure used in present-day PSHA was originally proposed by Cornell (1968). Since that time there has been significant progress in scientific understanding of the earthquake process and in the technique for evaluation of the relevant seismology, geo­logical, and geophysical data.  Several studies with various degree of sophistication are conducted for the assessment of seismic hazard in the Marmara Region (Atakan et al., 2002; Erdik et al., 2004; Kalkan et al., 2008).

The betterment of the knowledge on the seismotectonic regime of the Marmara region will pave the path for development of alternative source models for the improvement of existing probabilistic hazard maps. In this connection, the most recent findings and outputs of different work packages of the project, in terms of seismicity, fault segmentation, slip rate data and association of past earthquakes with individual segments will be utilized. Various renewal-type stochastic models and characteristic earthquake occurrence will be utilized for the earthquake rupture forecasting. For near fault quantification of hazard at long period spectral accelerations, the directivity affects will be considered in the analysis. The epistemic and aleatory uncertainties will be rigorously treated respectively, through the use of a comprehensive logic tree analysis and the consideration of inter-event and intra-event variabilities.

The assessment of the inter-event correlation of earthquake ground motion will be facilitated through use the data obtained from the dense accelerometric network in Istanbul.

Task 6. Develop short-term earthquake forecast maps

The aim of this task is to focus on the development of short-term earthquake forecast maps for the Marmara region. The long term probabilistic hazard models are currently the most crucial forecasting tool against earthquake damage, because they are used as guidelines for earthquake safety provisions of building codes, whereas the short-term forecasting of the earthquake aftershocks is used for time-dependent seismic hazards to help communities prepare for potentially destructive aftershocks.

The improvement of the existing long-term earthquake hazard assessments in the Marmara region will be carried out under WP5-Task 5. The possibility of developing models for the short-term forecasts in the Marmara region will be investigated in this task based on the analysis of aftershock sequences and seismic clustering. The short-term model of Gerstenberger et al. (2007), used in EU FP6 NERIES Project, will also be considered.  The possibility of integrating the studies with the on-going research in EU FP7 projects of REAKT and NERA as well as with the Collaboratory for the Study of Earthquake Prediction (CSEP, project will be explored.