Statistical and predictive interpretation of the relationship between frequency and magnitude underlying a recently tested 'earthquake traffic light' on a California seismic sequence
Successfully performed the first test on the "earthquake traffic light", a model developed last October by Laura Gulia and Stefan Wiemer of the Swiss Seismological Service (SED), capable of statistically interpreting the variation of the b-value, i.e. the parameter that expresses the relationship between frequency and magnitude of earthquakes. This model is based on systematic analyzes of seismic sequences that have occurred all over the world to provide a real-time response based on an alert code (characterised by the colors green, orange and red).
The test, just published in the journal Seismological Research Letters in the study 'Pseudoprospective Evaluation of the Foreshock Traffic-Light System in Ridgecrest and Implications for Aftershock Hazard Assessment', and conducted by a team of researchers from the SED and the University of Bologna, Gulia and Wiemer, in collaboration with Gianfranco Vannucci from the National Institute of Geophysics and Volcanology (INGV), was carried out through the implementation of the model in near real time, i.e. retrospectively but using preliminary data from the Californian Ridgecrest seismic sequence of July 2019 which were available at the time in real-time.
The task of INGV in the study was to reconstruct, in the initial stages of the analysis of the sequence, the possible seismogenic sources of the two main events: the earthquake of magnitude 6.4 of 4 July 2019 and the subsequent mainshock, or the main seismic event, of magnitude 7.1 of 6 July 2019.
"We analyzed the relationship between distribution over the territory and magnitude of earthquakes, known as Gutenberg-Richter, whose slope provides a parameter known in seismology as b-value, and its spatio-temporal variations on the fault plane", explains Gianfranco Vannucci, an INGV researcher who collaborated on the study. “A decrease in the b-value after an earthquake with a magnitude greater than 6 corresponds to an alert that varies from 'orange' to 'red'”.
In the reproduction of real-time conditions “we reconstructed the fault plane of the magnitude 6.4 earthquake after the first 24 hours”, continues Vannucci, “and then used the preliminary data to start evaluating the temporal changes in the b-value. Same procedure after the event of magnitude 7.1. The model successfully identified the two aftershocks: the first as a foreshock (red alert), the second as a mainshock (green alert)."
In the model was also an analysis on the sensitivity of b-value as both the completeness magnitude, the minimum threshold value for data selection, and the “no-alert-time”, i.e. the time interval, immediately following the events of greater magnitude, in which it is not possible to clearly distinguish the numerous earthquakes that follow one another on the seismograms.
Through a mapping of the variations of the b-value, the authors also successfully identified the future enucleation area mainshock of 6 July. After this shock, in fact, the b-value increased giving a 'green' traffic light signal, which indicated that it had already taken place mainshock.
The analysis conducted using the preliminary data from the California seismic network was also repeated with the high-resolution data published a few months later. The initial results have been confirmed allowing to reduce both the no-alert-time is the value of the magnitude of completeness.
This demonstrated, for the purposes of applying the model, the importance of a dense seismic network which makes available high-resolution data collections, ie with well-localized and small-magnitude events. This model, at present, is not yet applicable everywhere in the world or for earthquakes that occur at sea.
#ingv #sed #unibo #earthquakes
Figure 1 – Results of the Gulia and Wiemer model (Nature, 2019) with preliminary data: time series of the b-value of the entire seismic sequence (A) calculated on the fault planes of the events of 4 (M64) and 6 July (M71) with the threshold values of the 3 alerts (green-orange-red). The reference value (background) of the b-value is indicated in blue. The black vertical dotted lines indicate the origin times of the two events of 4 (M6.4) and 6 July (M7.1) while the gray area indicates the standard deviation of 1 sigma with the Shi and Bolt method , (1982). The black rectangle shows a zoom of the time course of the b-value between the two aftershocks. Images B and C indicate the distribution of the frequency-magnitude of the earthquakes, in the different time intervals, the related b-values, the percentage variations with respect to the reference value and the corresponding alert (from Gulia et a., SRL, 2020).
Figure 2 - Map of the percentage variations of the b-value with respect to the background value between the shocks of 4 (M6.4) and 6 July (M7.1) (image A) and in the week following the shock of M7.1 (image B) . The negative variations of the b-value indicated by the red circle indicate the enucleation area of the Magnitude 7.1 event (red star) (from Gulia et a., SRL, 2020).