EGU General Assembly 2020
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Stress Drop Mapping in the Northern Chilean Subduction Zone

Jonas Folesky, Joern Kummerow, and Serge A. Shapiro
Jonas Folesky et al.
  • Freie Universität Berlin, Geowissenschaften, Geophyisk, Berlin, Germany (

The Northern Chilean subduction zone has been monitored by the IPOC network for more than ten years. During this time period two very large earthquakes occurred, the 2007 MW7.7  Tocopilla earthquake and the 2014 MW8.1 Iquique earthquake. Over the entire subduction zone a vast amount of seismic activity has been recorded and a huge catalog was compiled including over 100000 events (Sippl et al. 2018). With this exceptional data base we attempt a systematic analysis of the stress drops of as many events from the catalog as possible. We apply different estimation techniques, namely the spectral ratio type, the spectral stacking approach, and the lower bound method. A goal of our research is a comparison and possibly a combination of the techniques to obtain reliable and well constrained results.

The data set covers events at the interface, within the subducting plate, crustal events, and intermediate depth events. It therefore bears a great potential to better understand the stress drop distribution within a subduction zone. Also, the long observation interval allows to analyze temporal variations according to pre-, inter-, and post-seismic phases of megathrust earthquakes.   

We present preliminary results where a subset of 730 events with a magnitude range of ML2.7 - ML4.8  was used for analysis with the spectral ratio technique. For these events we show maps of spatial stress drop variation, and we analyze the time dependent stress drop variance. 

How to cite: Folesky, J., Kummerow, J., and Shapiro, S. A.: Stress Drop Mapping in the Northern Chilean Subduction Zone, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-5019,, 2020


Display file

Comments on the display

AC: Author Comment | CC: Community Comment | Report abuse

displays version 2 – uploaded on 05 May 2020
A statement has been added to the Acknowledgements section and a citation has been added.
  • CC1: This is a summary of Qs and As from the live chat, Henriette Sudhaus, 07 May 2020

    Q: Thanks for sharing your results! Your 3. conclusion is that there are "slight spatial dependencies of stress drop e.g. with distance to the interface". Is the significance tested? Would we expect increased stress drop with larger distance as indicated?
    A: 1) significane has not been tested explicitly, but numbers of events are high supporting the finding. 2)yes, a possible explanation might be maturity of existing fractures when receding from the interface

    Q: Do you have an interpretation for the lack of depth dependence? Does this imply similar amounts of stress building up at all depths?
    A: My perception would be the other way around. Why should stress drop depend on depth explicitly. It is not a measure of the in situ stress level but of stress level difference.
    Also this has been described before by multiple authors. A key point is the exact removal of attenuation which would contribute to artifical depth dependece

    Q: I'd be interested if your data might be consistent/inconsistent with the model I proposed a few presentations back, particularly if there might be a dependence of inferred stress drop on roughness (higher roughness farther from interface). I'd be interested in following up with you on this later
    A: Yes please.


    Q: you mention magnitude dependency of stress drop; how large is this dependency ?
    A: We quantified this. I must look it up. It is significant something like log(sd)~0.5*log(M0), and yes it does violate self similarity assumption

    Q: great display - Can you comment on the temporal variation you mention for the Mw8.1 Iquique event? and just a quick clarification question further distance from interface does not imply larger depths?

    A: 1) we observa a jump to higher sd values immediately before the April 1, 2014 MW8.1 Iquique mainshock, a continuous reduction thereafter and a subsequent recovery to normal values after a few weeks
    2) yes, thats different. 
    Christos Evangelidis (NOA) (09:48) It is a typo: The HF radiation starts NE of the hypocenter


    Q: nice study. Could there still be an effect of spatial variations of shear-wave speed that are not accounted for in the final stress-drop values shown?
    A: As this is a EGF technique, I would suppose there should not.
    Q: with the EGF technique you estimate corner frequency; but then fc is used to compute stress drop according to Brune / Boatwright or a similar model, which includes \beta, the shear-wave speed at the source. So some local-structure effects should be present — unless I am misunderstanding your approach.
    A: that is correct. we use a pseude 3d velocty model here, the best we could find. If shear wave speed at the source changes, there would be an effect.  

    Q: What is distribution of stress drop in deepest and shallow events

    Q: Just below 60 km there seems to be a decrease in stress drop - did you find this to be not significant - if it's interpreted to be transition to stable sliding it might not be unreasonable

    A: No we did not yet consider this explicitly. But it fits to our considerations and possibly to Christians observation that interface seismicity in this depth vanishes

displays version 1 – uploaded on 02 May 2020, no comments