AUTHOR=Böse Maren , Hutchison Allie A. , Manighetti Isabelle , Li Jiawei , Massin Frédérick , Clinton John Francis 

TITLE=FinDerS(+): Real-Time Earthquake Slip Profiles and Magnitudes Estimated from Backprojected Displacement with Consideration of Fault Source Maturity Gradient

JOURNAL=Frontiers in Earth Science

VOLUME=Volume 9 - 2021

YEAR=2021

URL=https://www.frontiersin.org/journals/earth-science/articles/10.3389/feart.2021.685879

DOI=10.3389/feart.2021.685879

ISSN=2296-6463

ABSTRACT=The Finite-Fault Rupture Detector (FinDer) algorithm computes rapid line-source models from high-frequency seismic acceleration amplitudes and magnitudes from empirical rupture length-magnitude relations as are needed for earthquake early warning. In this paper, we add several algorithms to the FinDer suite. First, we calculate slip profiles in real time (FinDerS) by employing a simple empirical relation between long-period dynamic peak ground displacement (PGD) and the closest distance from the FinDer line-source to the receiver. The algorithm is further expanded by using a second empirical relationship based on the long-term fault maturity properties that influence along-strike coseismic slip behavior. We use this relationship for an “intelligent” interpolation and smoothing of the estimated slip profiles. Using the mean slip of these profiles provides an alternative estimate of earthquake magnitude compared to FinDer. The second expansion of FinDer (FinDerS+) allows for additional future rupture growth of the FinDer line-source and provides predicted values for final rupture length and alternative magnitude estimates. FinDerS and FinDerS+ can incorporate both seismic and geodetic data. In order to recover dynamic PGD, we double-integrate and high-pass filter seismic acceleration records. Although less accurate compared to PGD amplitudes derived from geodetic GNSS records, this provides information from which first-order slip distributions in a dense seismic network can be recovered. Due to smaller packet sizes (usually 1 sample per second), GNSS data may in some cases be transmitted faster to the datacenter, but requires processing to convert position to displacement time series. The FinDer algorithms are tested for the 2019 Mw7.1 Ridgecrest (California), 2016 Mw7.0 Kumamoto (Japan), and 2008 Mw7.9 Wenchuan (China) earthquakes. Conclusively, low-frequency PGD data does not speed-up calculations for these events, but provides additional information on slip distributions and alternative estimates of magnitudes that can be useful to check for consistency. Finally, the slip distributions can be useful for improved ground-motion prediction given our understanding of the relationship between seismic radiation and fault maturity.