 |
| Titel |
Tipping point analysis of seismological data |
| VerfasserIn |
Valerie N. Livina, Elena Tolkova |
| Konferenz |
EGU General Assembly 2014
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 16 (2014) |
| Datensatznummer |
250093366
|
| Publikation (Nr.) |
 EGU/EGU2014-8014.pdf |
|
|
|
|
|
| Zusammenfassung |
| We apply the tipping point toolbox [1-7] to study sensor data of pressure variations
and vertical velocity of the sea floor after two seismic events: 21 October 2010,
M6.9, D10km (California) and 11 March 2011, M9.0, D30km (Japan). One type of
datasets was measured by nano-resolution pressure sensor [8], while the other, for
comparison, by a co-located ocean bottom seismometer. Both sensors registered the
seismic wave, and we investigated the early warning and detection signals of the
wave arrival for possible application with a remote and cabled tsunami warning
detector network (NOAA DART system and Japan Trench Tsunami Observation
System).
We study the early warning and detection signals of the wave arrival using methodology
that combines degenerate fingerprinting and potential analysis techniques for anticipation,
detection and forecast of tipping points in a dynamical system. Degenerate fingerprinting
indicator is a dynamically derived lag-1 autocorrelation, ACF (or, alternatively, short-range
scaling exponent of Detrended Fluctuation Analysis, DFA [1]), which shows short-term
memory in a series. When such values rise monotonically, this indicates an upcoming
transition or bifurcation in a series and can be used for early warning signals analysis. The
potential analysis detects a transition or bifurcation in a series at the time when it happens,
which is illustrated in a special contour plot mapping the potential dynamics of
the system [2-6]. The methodology has been extensively tested on artificial data
and on various geophysical, ecological and industrial sensor datasets [2-5,7], and
proved to be applicable to trajectories of dynamical systems of arbitrary origin
[9].
In this seismological application, we have obtained early warning signals in the described
series using ACF- and DFA-indicators and detected the Rayleigh wave arrival in the potential
contour plots. In the case of the event in 2010, the early warning signal starts appearing about
2 min before the first peak of the Rayleigh train is detected by the sensor, whereas in the case
of event of 2011, the early warning signal appears closer to the peak arrival, within 1 min.
The different strength of early warning signals of the Rayleigh trains may be due
to different depths of the events (10 and 30 km), which we plan to test in further
analysis.
References: [1]Livina and Lenton, GRL 2007; [2]Livina et al, Climate of the Past
2010; [3]Livina et al, Climate Dynamics 2011; [4]Livina et al, Physica A 2012; [5]Livina
and Lenton, Cryosphere 2013; [6]Livina et al, Physica A 2013; [7]Livina et al, Journal of
Civil Structural Health Monitoring, in press; [8]Tolkova and Schaad, arXiv:1401.0096v1;
[9]Vaz Martins et al, PRE 2010. |
|
|
|
|
|
|