Currently, spatially correlated sampling of conditional (and "unconditional") IM values requires the evaluation of the full covariance matrix. This poses memory and performance problems, when we consider a large number of sites (for example more than 6000).
Describe the solution you'd like
The shakemap algorithm proposed by Engler et al. (2022) is currently implemented via the computation of the full conditional covariance matrix. However, the authors propose a probabilistically equivalent solution to separate between-event and within-event residuals. Using that procedure, allows to easily benefit from efficient sampling techniques such as the one proposed in Verros (2016).
Steps:
Implement alternative updating procedure for Shakemap_EnglerEtAl2022
Implement efficient sampling for Shakemap_EnglerEtAl2022
References:
Engler T., Worden B., Thompson E., and Jaiswal K. (2022): Partitioning Ground Motion Uncertainty When Conditioned on Station Data. Bulletin of the Seismological Society of America. doi: : 10.1785/0120210177
Verros S. A. 2016. A class of efficient algorithms for stochastic seismic ground motions, Master’s thesis, Colorado School of Mines, Golden, Colorado. Link to Mines repository
Is your proposal related to a problem?
Currently, spatially correlated sampling of conditional (and "unconditional") IM values requires the evaluation of the full covariance matrix. This poses memory and performance problems, when we consider a large number of sites (for example more than 6000).
Describe the solution you'd like
The shakemap algorithm proposed by Engler et al. (2022) is currently implemented via the computation of the full conditional covariance matrix. However, the authors propose a probabilistically equivalent solution to separate between-event and within-event residuals. Using that procedure, allows to easily benefit from efficient sampling techniques such as the one proposed in Verros (2016).
Steps:
Shakemap_EnglerEtAl2022Shakemap_EnglerEtAl2022References:
Describe alternatives you've considered
na
Additional context
na