Home / News and Events / Events / StatSei10 / CORSSA tutorials

CORSSA tutorials

StatSei10 Pre-workshop Tutorials presented by the Community Online Resource of Statistical Seismicity Analysis (CORSSA)

On Monday, 20th February, CORSSA organised two two-hour talks , which were held at Victoria University of Wellington. The speakers have kindly made available their slides. They can be downloaded from the link below each abstract.

Talk 1: The Poisson Assumption: Applications in Spite of Clustering 

Dr. Andrew J. Michael

Research Geophysicist

US Geological Survey, Monlo Park, CA, USA


The Poisson distribution is frequently used to describe the temporal behavior of earthquakes.  It is the basis of Probabilistic Seismic Hazard Analysis and a frequent null hypothesis for testing earthquake prediction and forecasting methods.  But the clustering of seismicity, which has been noted from our earliest earthquake records,  violates the Poisson assumption that earthquakes are independent events.  To get from our highly clustered data to a valid application of the Poisson distribution, we can decluster the catalog by removing foreshocks and aftershocks or transform time such that the events appear independent.  In this tutorial, we will consider the uncertainties associated with these approaches; tests for determining if the Poisson  distribution is a good description of the resulting data; applications to global seismicity, earthquake prediction tests, and hazards assessment; non-stationary Poisson models for Operational Earthquake Forecasting; and routes to move beyond the limitations of the Poisson assumption.

You can download the slides of the presentation here: CORSSA-Michael-2017.pdf 7.79 MB

Talk 2: Statistical Modelling of Induced Seismicity

Dr. Sebastian Hainzl

Senior Scientist

GFZ German Research Centre for Geosciences, Potsdam, Germany


Induced seismicity has become an emergent topic of seismology because the increased exploitation of natural resources has locally led to a dramatic increase of the seismicity level and some strong earthquakes. An estimation of the corresponding probabilistic seismic hazard requires a proper seismicity model for the induced events. For this purpose, one can exploit some physical knowledge about the triggering mechanisms and the fact that induced seismicity shares some common features with natural seismicity. This tutorial will introduce this research topic; in particular, it will cover the following points:

  • types of induced seismicity
  • characteristics of induced versus tectonic seismicity
  • empirical models of induced seismicity
  • physics-based seismicity models based on Coulomb stress calculations.

You can download the slides of the presentation here: CORSSA-Hainzl-2017.pdf 2.26 MB