← Back to Sessions

Real-Time Ground Shaking Maps Reconstructions With a Hybrid ShakeMap Implementation

Description: 

Ground shaking maps have been applied to a variety of tasks but their most useful application is arguably for seismic monitoring and civil defence operations as they provide information about the area and amplitude of the ground motion relative to a seismic event. Shaking maps, expressed in terms of a ground motion parameter, are reconstructed constraining the theoretical values obtained from ground motion prediction equations (GMPEs), computed from the magnitude and location of the earthquake, with the ground motion parameters values recorded at the stations, accounting also for local site effects. The need for source parameters to evaluate the GMPEs prevents a real-time implementation of such a method. One possible solution to the problem is to develop algorithms that can constrain the interpolation process using only the ground motion parameters recorded at the stations.
We propose a hybrid model combining the conditioned multivariate normal distribution (MVN) technique adopted by ShakeMap and a neural network replacing the GMPE. The neural network provides a purely data-driven approximation of the GMPE results based only on the spatially sparse ground motion parameters recorded at the stations, with possible correction for the site effects. Moreover, by limiting the use of a neural network to a specific task we improve its explainability with respect to end-to-end models. The proposed implementation has been trained and tested with data from the Italian territory and its results compared to the ones obtained with ShakeMap. This approach is easily integrable into the existing workflow, combines well-studied interpolation techniques and neural networks in an explainable structure, and provides high-resolution estimates of the ground-shaking fields with real-time capabilities and potential relevance in the context of early warning.

Session: ShakeMap-related Research, Development, Operations, Applications and Uses

Type: Oral

Date: 4/19/2023

Presentation Time: 05:00 PM (local time)

Presenting Author: Simone Francesco Fornasari

Student Presenter: Yes

Invited Presentation: 

Authors