Earthquake Physics
Determining causal relationships behind earthquake occurrence on temporal and spatial scales relevant for informing on earthquake hazard is a challenging problem.
In a time independent way, we can predict all major earthquakes by drawing wiggly lines around the major plate boundaries, but this information is not precise enough to be very useful except to the first order.
On a regional scale, we can stack modern, historic and paleo events to map a time independent hazard, but we know that we have not sampled all possible earthquakes in the modern digital era (since ~1970s) and historic records are inherently incomplete. Such maps form the basis of modern hazard assessment.
More ambitious is to try to map the spatial-temporal evolution of earthquakes. We know that when large earthquakes occur, other events will be triggered in the vicinity, but models which forecast where they are most likely to occur have proven difficult to validate statistically.
On the statistical side, I am particularly interested in quantify limits of predictability, how we interpret complex emergent datasets and how we deal with poorly sampled datasets.
Edinburgh GeoHazards Research GroupProjects
Origin and non-universality of earthquake inter-event time distributionsTesting evidence for Characteristic earthquakes in frequency-magnitude plots
Earthquakes as a slowly driven complex system
CORSSA - Community Online Resource for Statistical Seismology Analysis