1:
2: \documentclass[prl,floatfix,twocolumn]{revtex4}
3: \usepackage{graphicx}
4: \usepackage{psfrag}
5:
6: %\hoffset=-0.35cm
7: %\voffset=0.3cm
8:
9: \def\au{autocorrelation function }
10: \def\sy{system }
11: \def\syn{system}
12: \def\dn{detector}
13: \def\d{detector }
14: \def\ds{detectors }
15: \def\dn{detector}
16: \def\dsn{detectors}
17: \def\sc{superconductor }
18: \def\scn{superconductor}
19: \def\scg{superconducting }
20: \def\sa{sapphire }
21: \def\san{sapphire}
22: \def\c{crystal }
23: \def\cn{crystal}
24: \def\w{$\delta t$ }
25: \def\eq{earthquake }
26: \def\eqn{earthquake}
27:
28: \begin{document}
29:
30: \title{Comment on ``Universal Distribution of Interearthquake
31: Times''}
32:
33: \author{
34: J. {\AA}str\"om $^2$,
35: P.C.F.~Di Stefano$^{4
36: }$,
37: F.~Pr\"obst$^1$,
38: L.~Stodolsky$^{1 *}$,
39: J.~Timonen$^3$,
40: }
41:
42: \affiliation{
43: $^1$ Max-Planck-Institut f\"ur Physik, F\"ohringer Ring
44: 6, D-80805 Munich, Germany;
45: $^2$ CSC - IT Center for Science, P.O.Box 405, FIN-02101 Esbo,
46: Finland;
47: $^3$ Department of Physics, P.O. Box 35 (YFL), FIN-40014 University
48: of Jyv\"askyl\"a, Finland;
49: $^4$ Institut de Physique Nucl\'eaire de Lyon, Universit\'e Claude
50: Bernard Lyon
51: I, 4 rue Enrico Fermi, 69622 Villeurbanne Cedex, France;
52: $^*$ Corresponding author, {\it email address:} les@mppmu.mpg.de.}
53:
54:
55:
56:
57:
58: \begin{abstract}
59: \end{abstract}
60:
61:
62: \maketitle
63:
64: In a {\it Letter} earlier this year ~\cite{sorn} and in a
65: number of preceeding publications~\cite{cor}\cite{sca}\cite{bak},
66: the probability
67: distributions for the ``waiting time'' between earthquake events
68: have been discussed. In particular it appears that
69: the probability distribution for the number of events with waiting
70: time $w$, when expressed in terms of a suitably scaled
71: variable $(w/w_0)$ with $w_0$ some characteristic time constant,
72: follows a universal
73: function~\cite{bak}. In this {\it Comment} we would like to draw
74: attention to the
75: fact that recently published data \cite{crst} of the CRESST
76: collaboration on microfractures in sapphire show the same
77: features. Indeed there is a great similarity, if not a
78: remarkable complete identity, of the probability distributions
79: expressed in this manner between the \eqn s and the microfractures.
80:
81:
82:
83: \begin{figure}[h]\label{ww}
84: {{\includegraphics[width=\hsize]
85: {fig.ps}}}
86: \caption{ CRESST waiting time distributions. Upper curve:
87: microfractures,
88: fit to $\propto w^{-\alpha}e^{-w/w_0}$. Lower curve:
89: photon-induced events from a calibration run, fit to
90: $\propto e^{-w/w_0}$.}
91: \end{figure}
92:
93: In Fig 1 we reproduce Fig 2 of ref~\cite{crst}. The upper curve is
94: the data on microfractures, fit to
95: \begin{equation}\label{one}
96: dN/dw\propto w^{-\alpha}e^{-w/w_0}
97: \end{equation}
98: with $\alpha=0.33$ and $w_0=0.0014$ hrs. It will
99: be seen there
100: is an excellent fit. The lower curve
101: represents a test of the
102: apparatus and analysis, using photon-induced
103: events from an external radioactive source. These should
104: follow the Poissonian $e^{-w/w_0}$ and there is also a good
105: fit.
106:
107: According to Corral ( Physica A)
108: the form Eq~\ref{one} describes the waiting times for \eqn s, and
109: with the same power, $\alpha=0.33$.
110: Concerning the time
111: scale parameter $w_0$, it is essentially the inverse of the
112: observational or experimental event rate R since from Eq~\ref{one}
113:
114: \begin{equation}\label{two}
115: 1/R={\bar w}=(1-\alpha)w_0\;.
116: \end{equation}
117:
118: The data used in the Figure satisfy this relation to within a few
119: percent, as would be expected from the good fit. We find that
120: raising the energy threshold in a data set, and so reducing R,
121: leads to a linear relation between the fit $w_0$ and R, as would be
122: expected from Eq~\ref{two} with a constant $\alpha$.
123:
124: Alternatively one
125: could renounce fitting $w_0$, and simply substitute
126: $w_0^{-1}=(1-\alpha)R$ into Eq~\ref{one}, use the
127: experimental R (=28 000 events/28.5 hrs), and fit for $\alpha$.
128: This essentially one parameter fit is satisfactory and yields
129: $\alpha=0.26$.
130:
131:
132: Although the CRESST values for $\alpha$ thus vary somewhat
133: according to the analysis and from run to run, the parallelism
134: between the two kinds of phenomena is striking.
135: These considerations, involving such widely disparate time scales,
136: energies, and material properties, raise the question as to whether
137: Eqs \ref{one} and \ref{two} do not represent a general law,
138: applicable to many kinds of fracture processes.
139:
140:
141:
142:
143: \begin{thebibliography}{00}
144:
145: \bibitem{sorn} A. Saichev and D. Sornette, Phys. Rev. Lett. {\bf
146: 97}, 078501, (2006).
147:
148: \bibitem{cor} A. Corral, Phys Rev. {\bf E 68}, 035102
149: (2003); A. Corral, Physica (Amsterdam) {\bf A 340}, 590 (2004);
150:
151: \bibitem{sca} N. Scafetta and B. J. West, Phys. Rev. Lett. {\bf
152: 92}, 138501, (2004).
153:
154: \bibitem{bak} P. Bak, K. Christensen, L. Danon, and T. Scanlon,
155: Phys. Rev. Lett. {\bf 88}, 178501, (2002);
156:
157: \bibitem{crst} J. {\AA}str\"om et al., Phys. Lett. {\bf A356} 262
158: (2006),
159: (arXiv.org: physics/0504151); Nucl. Inst. Methods {\bf A559}, 754
160: (2006).
161:
162:
163: \end{thebibliography}
164:
165: \end{document}
166: