EGU General Assembly 2022 - Presentation & OSPP
During the 2022 EGU General Assembly I was able to (virtually…) present the results from my first year (and a half…) of PhD.
For the presentation (available here) I was awarded the Outstanding Student and PhD candidate Presentation (OSPP) Award 2022.
The work I talked about revolves around the idea of reconstructing the ground shaking maps for an area in real-time using sparse data (i.e., the values of the ground motion parameter recorded at each seismic station).
This work can have real applicability in a monitoring environment, complementing traditional methods which rely on ground motion predictions equations and thus require information about the magnitude and location of the event, which are slow to compute: the Italian Department of Civil Protection gets these information (manually revised by INGV operators) on average 12 minutes after an event; on the other hand, the data from the stations are available with a delay of just a few seconds since the data have been recorded.
The station information available in real-time are exploited using a machine learning approach that provides robust results in real-time: despite the hardware at DMG seeming to be the bottleneck, an updated ground shaking map is available every few seconds using the data from the whole Italian accelerometric network (RAN)!
The abstract of the presentation is available here.
A paper about this work has been published on BSSA:
Simone Francesco Fornasari, Veronica Pazzi, Giovanni Costa; A Machine‐Learning Approach for the Reconstruction of Ground‐Shaking Fields in Real Time. Bulletin of the Seismological Society of America 2022; doi: https://doi.org/10.1785/0120220034