Analisi della quiescenza sismica che ha preceduto la sequenza dell'Italia centrale
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Analisi della quiescenza sismica che ha
preceduto la sequenza dell’Italia centrale
S. Gentili1, R. Di Giovambattista2, A. Peresan1
1 Istituto Nazionale di Oceanografia e di Geofisica Sperimentale
Centro Ricerche Sismologiche, Udine, Italy.
2 Istituto Nazionale di Geofisica e Vulcanologia
Sezione di Roma, Roma, Italy.
Istituto Nazionale
di Oceanografia e di Geofisica Sperimentale 1
Central Italy earthquake
• The devastating 2016-2017 central Italy
seismic sequence is the largest recorded
in Italy since the 1980 M 6.9 Irpinia
earthquake.
• It fell in a seismic gap between northern
and central Apennines, between the
1997-1998 Umbria-Marche and the 2009
L’Aquila earthquakes
• We applied the Region Time Length
(RTL) (Sobolev et al., 1997, 1996)
algorithm to study detailed property of
the quiescence. (works on declustered
catalogue) Gentili et al., PEPI, 272, 27‐33 (2017)
Istituto Nazionale
di Oceanografia e di Geofisica Sperimentale 2
Extended seismicity decrease
Number of earthquakes
observed in the analyzed area
Gentili et al., PEPI, 272, 27‐33 (2017)
Earthquakes from Earthquakes from
for two subsequent time windows
8/23/2014 to 8/23/2015 8/24/2015 to 8/23/2016
of one year: (a) From 8/24/2014
to 8/24/2015 (b) from 8/25/2015
to 8/24/2016. The area is
sampled by a grid 10x10 km
spaced, and, for every node, the
number of earthquakes within
100 km radius in the previous
year is shown.
A clear decrease of seismicity
can be seen in the last year
with respect to the previous
one
Istituto Nazionale
di Oceanografia e di Geofisica Sperimentale 3
Catalogue declustering
• Declustering is particularly difficult in the Apennines, due to the
closeness in space and time of different clusters => we used a
statistical method called “nearest‐neighbours”.
• As shown by Zaliapin and Ben‐Zion (2013), the removal of clusters
identified by Nearest‐Neighbors method does not alter the
features of inhomogeneous and possibly non‐stationary
background seismicity, which are relevant for this study.
• The method requires only two input parameters: the b‐value, b
and the fractal dimension of epicenters, d
Istituto Nazionale
di Oceanografia e di Geofisica Sperimentale 4
Rescaled space and time distances
Catalogue declustering
Fractal dimension of epicenters
Zaliapin et al., PRL, 101, 018501 (2008)
Gutenberg-Richter law, b
ηij =tij rijd 10-bmi /2 with tij>0
Spatial distance
Interevent time
The Nearest-Neighbor Method (Zaliapin et al., 2008;
Zaliapin and Ben-Zion, 2013) expresses the space-
time distance η between two earthquakes j and i in
terms of rescaled time T and rescaled distance R:
Istituto Nazionale
di Oceanografia e di Geofisica Sperimentale 5
Nearest‐Neighbors analysis: Central Italy
The 2D density map of rescaled time and
space components of the nearest-neighbor
distance in the observed catalogues is
prominently bimodal bimodality is used to
Gentili et al., PEPI, 272, 27‐33 (2017)
separate the earthquakes into Poissonian
background and cluster populations
(Zaliapin and Ben-Zion, 2013).
Earthquake data for Central Italy (Lat=[40, 46]°N,
Lon=[10, 15]°E), 2005-2017: Italian
Seismological Instrumental and Parametric
Data-Base (till March 2017)
http://iside.rm.ingv.it/iside/ Mc=2.2
Scaling parameters: based on the Unified Scaling
Law for Earthquakes (USLE) (Nekrasova et al.,
2011). Robust average estimates are:
B-value b = 0.8-1.0
Fractal dimension d = 1.3-1.4
Istituto Nazionale
di Oceanografia e di Geofisica Sperimentale
6
RTL algorithm
• The RTL algorithm measures the level of seismic activity in moving time windows by
counting the number of earthquakes, weighted by their size, and inversely weighted
by their distance, in time and space from the point of observation.
• It represents the deviations from the background seismicity:
seismicity decreases ‐> RTL decreases; seismicity increases‐> RTL increases
r0 and t0= free parameters
Rs,Ls, Ts=linear trend
R ( x, y , t ) T ( x, y , t ) L ( x, y , t )
RTL ( x, y , t ) corrections
R T L σR, σR, σR=standard
deviation
n
t ti
T ( x, y , t ) exp Ts ( x, y , t ) n li
n
r t 0 Ls ( x, y , t ) if ri
R( x, y, t ) exp i Rs ( x, y, t )
i 1
i 1 r
L( x , y , t ) i
r0 li L ( x, y , t ) if
n
i 1 ri
i 1
s
Distance from x, y Time of occurrence Source size
Istituto Nazionale
di Oceanografia e di Geofisica Sperimentale
7
Parameters estimation
• An ad-hoc choice of the parameters may cause artifacts
• Huang and Ding [2012] proposed a formal procedure for the
identification of the most stable values, based on the correlation
coefficient over pairs of the RTL functions estimated using different
parameters which leads to a map of the quality of parameters.
• They select an area into the r0-t0 space where at least 70% of the
correlation coefficients over pairs exceeds a specified threshold for r0
and t0. Based on this map, they propose a method to find the optimal
values of r0 and t0.
Istituto Nazionale
di Oceanografia e di Geofisica Sperimentale 8
Parameters estimation
Gentili et al., PEPI, 272, 27‐33 (2017)
RTL vs time at the epicenter of the
Amatrice earthquake until earthquake r0 and t0 parameters quality map the
occurrence. Red line: RTL using epicenter of the Amatrice earthquake
r0=50 km and t0=1 year.
Istituto Nazionale
di Oceanografia e di Geofisica Sperimentale 9
Results
Gentili et al., PEPI, 272, 27‐33 (2017)
Grid 10x10 km spaced.
24, 2015, to August 23, 2016.
Map of mean RTL from August
• The area around the epicenter exhibits significant seismic quiescence anomaly
quantified by the mean RTL values.
• It started from the beginning of September 2015 and lasted for about 1 year
• It extended throughout a broad region north of the epicenter
Istituto Nazionale
di Oceanografia e di Geofisica Sperimentale 10Results
Gentili et al., PEPI, 272, 27‐33 (2017)
24, 2015, to August 23, 2016.
Map of mean RTL from August
• Contour of the quiescence area has the shape of an irregular elliptical form with
lengths of major and minor axes of about 40-45 and 30-35 km, respectively.
• All the aftershocks overlap eastern part of the quiescence area that to the east is
nearly coterminous with the Sibillini thrust and the main causative fault of the
earthquake.
Istituto Nazionale
di Oceanografia e di Geofisica Sperimentale 11Comparison with previous results
• Gentili [2010], using the RTL method, analyzed the seismicity
preceding the earthquakes with magnitude ≥5 in Italy from 1994
to 2004.
• She found a quiescence in 92% of the cases, with a duration D
varying from 0.6 to 3 years and an interval s from the
quiescence end to the earthquake of 0 to 2.9 years.
• Huang [2004] in a review on 8 earthquakes with magnitude >7
in Japan, Russia and Turkey, indicates a duration of 1-2.5 years
and a start of the quiescence few years before the mainshock.
• The Amatrice RTL, with D=1 and s=0 years, is therefore
compatible with previous results.
Istituto Nazionale
di Oceanografia e di Geofisica Sperimentale 12Comparison with previous results
• The RTL function has been previously applied to moderate/large earthquakes
in Italy [e.g., Di Giovambattista and Tyupkin, 2000, Di Giovambattista and
Tyupkin, 2004, Gentili and Bressan, 2007, Mignan and Di Giovambattista, 2008,
Gentili, 2010, Gambino et al., 2014, Gentili et al., 2017] and in other parts of
the world [e.g. Sobolev et al., 2002, Huang, 2006, Chen and Wu, 2006, Huang,
2008, Nagao et al. 2011, Huang and Ding, 2012, Wen et al., 2016].
• Several studies in Italy using RTL [Di Giovambattista and Tyupkin 2000 and
2004, Gentili and Bressan, 2007, Gentili 2010, Gambino et al., 2014] Z [Wyss et
al., 1997, Console et al., 2000] and beta statistics [Bragato 2014] have shown a
decreased seismicity before strong events in Italy.
Istituto Nazionale
di Oceanografia e di Geofisica Sperimentale 13RTL in Italy
5.6 Kobarid 1998 4.9 Sernio 2002 10
5
0
-5
-10
-15
-20
-25
1988 1990 1992 1994 1996 1998
Date
6.0 Umbria‐Marche
1997
5.2 Kobarid 2004
5.8 Palermo 2002
Istituto Nazionale
di Oceanografia e di Geofisica Sperimentale 14RTL in the world
1999 Chi-Chi Mw 7.6 2000 Tottori M 6.7
Chen and Wu 2006 Huang 2006
Kamchatka 1992-1993
Sobolev and Tyupkin1997
1999 Izmit Mw 7.4 1995 Kobe M 7.2
Huang 2004 Huang 2001
Istituto Nazionale
di Oceanografia e di Geofisica Sperimentale 15You can also read
