Synthetic Ground Motion Data for Finite Source Inversion Blind Tests in the Marmara Area for FP7 MARSite projectThis page is for the purpose of sharing the synthetic data for various earthquake scenarios in the Sea of Marmara (MARSite project - WP5). We are working on the ground motion simulations around the Sea of Marmara, including the region of Istanbul, over a dimension of 200 km (East-West) x 120 km (North-South).
This work is progressing with the GFZ (Germany) team, in particular. If someone else is interested, please contact me.
Figure 1: Map of model area (black frame) with a dimension of 200 km (EW) x 120 km (NS) x 40 km (depth). The receiver positions implemented in the simulations (open circle : Broadband statoin operated by KOERI, triangles : Strong Motion stations operated by KOERI). Small dots show the seismicity (Mw > 3) from the KOERI catalogue. Besiders these receiver points, the synthetic seismograms are recorded at thousands structured grid points (top-right panel) in order to provide the ground motion map.
Common Information for Blind Tests of Finite Source Inversion
Synthetic data set from ground motion simulationsGround motion is calculated using a 3D finite difference method in the 3D geological model. A layer of the Sea (depth down to 1-1.5 km) is taken into account. The ground surface is flat (z = 0 km). The origin time of the simulations is taken at t = 0. However the source model and the structure model are kept unknown for the purpose of blind test.
- Real-time GPU simulation-visualisation. See You Tube (updated by Thomas Ulrich in October 2014).
Fault segmentsSuppose the earthquake takes place along the North Anatolian fault. For the convenence of the inversion, the following is the two main segments apprximating the fault traces.
Central Marmara segment 100 km length, strike N86°E, dip 90°
(27.6862°E, 40.8183°N) to (28.87°E, 40.88°N)
Cinarcik segment 35 km length, strike N118°E, dip 90°
(28.87°E, 40.88°N) to (29.2367°E, 40.7324°N)
Receiver positions (StationName)Synthetic data are given by separated file for each receiver. Some receiver points may be located outside of the model area. Stations on land is always set on the ground surface. Station at depth (including OBS) is set to the nearest depth of the simulation grids.
KOERI BB stations Turkey_BB.sta (ID, longitude, latitude, height, name) KOERI SM stations Turkey_SM.sta (ID, longitude, latitude, height, name) SM stations requested by GFZ sm_station_name.dat GPS stations requested by GFZ magnet_gps_location_name.dat
Data format (StationName_SourceModel_StructureModel.dat)
1st line Coordinate in simulation (X, Y, Z) in m.
The grid number in simulation (I, J, K).
Vs in m/s at the receiver point in simulation
2nd line Number of time step (NT).
Time step in second (dt)
3rd - (NT+2)th line Ground Velocity (Vx=EW, Vy=NS, Vz=UD) in m/s
3D Green function (ground motion example from a point source)
Date 21st July 2015 Source information Hypocenter (28.5274°E, 40.8717°N, 7 km).
Seismic moment 10^16 N.m (Mw4.6).
Source time function: square of half sine with a duratio of 1 sec.
Ground motion 47 receivers. 8.2 Mb
Scenario 1 for Blind Test
Date Original in July 2014.
Revised in August 2014.
Updated on 20th November 2014 (data for 120 seconds, additional receivers)
Hypocenter 28.5274°E, 40.8717°N, 9.75 km depth Data 47 receivers. 8.2 Mb Note Dynamically simulated source, probably the easiest case to identify the rupture area, rupture directivity, and magnitude.
Scenario 2 for Blind Test
Date Updated on 24th November 2014. Hypocenter 28.7991°E, 40.8730°N, 9.8 km depth Data 47 receivers. 8.1 Mb Note Dynamically simulated source, more complex than the previous one.
Scenario 3 for Blind Test
Date 28th Nocember 2014. Hypocenter 28.455°E, 40.859°N, 14 km depth Data 47 receivers. 8.2 Mb Note Kinematically constructed based on the characteristic source model.
Scenario 4 for Blind Test
Date 7 January 2015. Hypocenter 28.317°E, 40.852°N, 17 km depth Data 47 receivers. 8.1 Mb Note Source model based on a kinematic inversion of another event.
Scenario 5 for Blind Test
Date 16 January 2015.
Updated also on the 16th January 2015.
Hypocenter 28.2305°E, 40.8482°N, 7 km depth Data 47 receivers. 8.1 Mb (revised) Note Checker-board like inversion test (original slip given by GFZ).
Scenario 6 for Blind Test
Date 20 January 2015. Hypocenter 27.7572°E, 40.8223°N, 11 km depth Data 47 receivers. 7.8 Mb Note Kinematic model based on a slip distibution of another inversion (given by GFZ).
Scenario 7 for Open Resolution Tests
Date 1D simulation: 3 March 2015.
3D simulation: 5 March 2015.
Setting Click image to enlarge. Data (1D simulation) 47 receivers. 7.9 Mb
Ground motion is calculated in the given 1D layered model.
Data (3D simulation) 47 receivers. 8.0 Mb
Ground motion is calculated in the 3D structure for the same source model.
Figure 2: Synthetic ground motion at station CTKS for three scenarios. BLACK: Scenario 1, RED: Scenario 2, and BLUE: Scenario 3. No filter is applied.
Marmara Finite Source Inversion Blind Test / Marmara Earthquake Scenario Database / Marmara Synthetic Ground Motion Database
Created on the 20th November 2014. Contact to Hideo Aochi.