Emmy-Noether-Projekt "Brückenschlag zwischen Geodäsie und Seismologie"

Source model for the 1997 Zirkuh earthquake (MW=7.2) in Iran derived from JERS and ERS InSAR observations

Sudhaus, H. and S. Jónsson (2011).

Geophysical Journal International, 185(2), 676-692.

DOI: 10.1111/j.1365-246X.2011.04973.x

Wiley online library


We present the first detailed source model of the 1997 M7.2 Zirkuh earthquake that ruptured the entire Abiz fault in East Iran producing a 125 km long, bended and segmented fault trace.

Using SAR data from the ERS and JERS-1 satellites we first determined a multisegment fault model for this predominately strike-slip earthquake by estimating fault-segment dip, slip, and rake values using an evolutionary optimization algorithm.

We then inverted the InSAR data for variable slip and rake in more detail along the multisegment fault plane. We complement our optimization with importance sampling of the model parameter space to ensure that the derived optimum model has a high likelihood, to detect correlations or trade-offs between model parameters, and to image the model resolution.

Our results are in an agreement with field observations showing that this predominantly strike-slip earthquake had a clear change in style of faulting along its rupture. In the north we find that thrust faulting on a westerly dipping fault is accompanied with the strike-slip that changes to thrust faulting on an eastward dipping fault plane in the south. The centre part of the fault is vertical and has almost pure dextral strike-slip.

The heterogeneous fault slip distribution shows two regions of low slip near significant fault step-overs of the Abiz fault and therefore these fault complexities appear to reduce the fault slip. Furthermore, shallow fault slip is generally reduced with respect to slip at depth.

This shallow slip deficit varies along the Zirkuh fault from a small deficit in the North to a much larger deficit along the central part of the fault, a variation that is possibly related to different interseismic repose times.


SRCMOD - Database

    • Inversion modelling of geodetic (InSAR) and seismological data
    • earthquake slip complexity and co-seismic rupture history
    • Connecting earthquake models to observations
    • Kinematic earthquake source inversion


    In my research i am interested in how earthquakes ruptures behave and how and why earthquakes develop complex ruptures in space and time. Complex means that the earthquake ruptures e.g. across multiple fault planes with different geometries or slows down/accelerates in different areas. We know that earthquakes rupture with different degrees of complexity and we believe that larger earthquake rupture in more complex ways. This would however violate the common assumption of self-similarity of earthquakes across magnitudes. Often the choice of the modeled degree of complexity is however dependent on expert knowledge. Therefore i am looking for data driven ways to help us evaluate possibly rupture segmentation. Also I focus on small to medium sized earthquakes to investigate if we can resolve any complex ruptures from them or if they do not exhibit such behavior. I am using InSAR, GPS and seismological data.

    To asses the evolution of an earthquake rupture in time i have developed a multi-array backprojection code, which is available on github: Palantiri