1: %% LyX 1.3 created this file. For more info, see http://www.lyx.org/.
2: %% Do not edit unless you really know what you are doing.
3: \documentclass[a4paper,english,prl,showpacs]{revtex4}
4: \usepackage[T1]{fontenc}
5: \usepackage{amsmath}
6: \usepackage[dvips]{graphicx}
7: \usepackage{amssymb}
8:
9: \makeatletter
10: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%% User specified LaTeX commands.
11: % Uncomment to print out only slides and overlays
12: %
13: %\onlyslides{\slides}
14:
15: % Uncomment to print out only notes
16: %
17: %\onlynotes{\notes}
18:
19: \usepackage{babel}
20: \makeatother
21: \begin{document}
22:
23: \title{Universal Scaling Laws for Large Events in Driven Nonequilibrium
24: Systems}
25:
26:
27: \author{M. K. Verma}
28:
29:
30: \affiliation{Department of Physics, Indian Institute of Technology Kanpur, Kanpur
31: 208016, India}
32:
33:
34: \author{S. Manna}
35:
36:
37: \affiliation{Department of Biophysics, University of Delhi South Campus, Benito
38: Juanez Road, New Delhi110021, India}
39:
40:
41: \author{J. Banerjee}
42:
43:
44: \affiliation{Department of Biophysics, University of Delhi South Campus, Benito
45: Juanez Road, New Delhi 110021, India}
46:
47:
48: \author{S. Ghosh}
49:
50:
51: \affiliation{Department of Animal Sciences, School of Life Sciences, University
52: of Hyderabad, Hyderabad 500046, India}
53:
54: \begin{abstract}
55: For many driven-nonequilibrium systems, the probability distribution
56: functions of magnitude and recurrence-time of large events follow
57: a powerlaw indicating a strong temporal correlation. In this paper
58: we argue why these probability distribution functions are ubiquitous
59: in driven nonequilibrium systems, and we derive universal scaling
60: laws connecting the magnitudes, recurrence-time, and spatial intervals
61: of large events. The relationships between the scaling exponents have
62: also been studied. We show that the ion-channel current in Voltage-dependent
63: Anion Channels obeys the universal scaling law.
64: \end{abstract}
65:
66: \pacs{05.65.+b, 05.45.Tp, 05.40.Ca}
67:
68: \maketitle
69: The physics of nonequilibrium systems has been of major interest in
70: last several decades. Recently large events have become a major resource
71: for characterizing nonequilibrium system. Bak et al. \cite{Bak:SOC_PRL,Bak:SOC_PRA}
72: constructed a model called {}``sandpile'' to abstract some of the
73: essential features of driven-nonequilibrium systems. Original model
74: of Bak et al. (BTW) as well as its variants exhibit powerlaw behaviour
75: for the magnitude and duration of avalanches, which play the role
76: of large events. Later scientists started investigating the temporal
77: correlations in various systems using recurrence-time pdf of large
78: events. In the BTW model, the pdf of recurrence time ($\tau$) of
79: large events is $\exp(-\tau/\tau_{c})/\tau_{c}$, which indicates
80: that the triggering of large events in this system are uncorrelated
81: \cite{Boff:SolarflarePRL,Sanc:SOCorNot}. This is a generic feature
82: of Poisson process. It was however soon realized that the events are
83: correlated in many nonequilibrium systems. Boffatta et al. \cite{Boff:SolarflarePRL}
84: studied solar flares, and found that the pdf of recurrence time for
85: the solar flares is a powerlaw, i. e., \begin{equation}
86: P(\tau)=A\tau^{-\beta},\label{eq:1}\end{equation}
87: with $\beta\approx2.4$. These authors also studied the recurrence-time
88: distribution for the dissipation-bursts in magnetohydrodynamic turbulence,
89: and again found a powerlaw with $\beta\sim2.7$. These results show
90: that the large events in solar flares and magnetohydrodynamic turbulence
91: are temporally correlated. Note that the pdf of solar-flare magnitudes
92: is a powerlaw just like sandpile model \cite{Lu:Solarflare,Pacz:EQ_Intesity}.
93: The difference between the sandpile model and solar flare is in the
94: temporal correlations between the large events \cite{Boff:SolarflarePRL,Sanc:SOCorNot}.
95:
96: Temporal correlations described above have been observed for many
97: other nonequilibrium systems. Corral \cite{Corr:PRL_Cluster,Corr:PRL_RG}
98: studied the pdf of recurrence time for large earthquakes (beyond 2.5
99: in Richter scale), and found it to be a powerlaw with $\beta\approx0.3$.
100: Bak et al. \cite{Bak:EQ} had done similar analysis. Earlier Guttenburg
101: and Richter \cite{GuteRich} showed that the magnitude of earthquakes
102: follows a powerlaw. {}``Extremal models'' (see Paczuski et al. \cite{Pacz:AvalacheDyna}
103: for review of these models) which are used to describe interface depinning,
104: biological evolution etc. also have similar temporal correlations.
105: Recently Banerjee et al. \cite{MKV:Ion} studied ion-channel currents
106: in Voltage Dependent Anion Channel (VDAC). Here too, the pdf of magnitude
107: as well as that of recurrence-time follow a powerlaw with $\beta\approx1.5$.
108: In thermal convection of Helium, Sreenivasan et al. \cite{KRS:Wind}
109: observed large-scale wind which switches directions; the pdf of inter-switch
110: interval is given by $(1/\tau)\exp(-\tau/\tau_{M})$. This feature
111: appears to indicate {}``release of convective stresses'' during
112: switching, just like energy release in earthquakes. The pdf of recurrence
113: time is connected to the power spectrum of the signal. Several research
114: groups \cite{Lowen:1f,Pacz:AvalacheDyna,Pacz:1f,MKV:Ion} have derived
115: $1/f$ power spectrum using the powerlaw distribution of recurrence-time.
116:
117: There is a recent work on spatial correlations in driven nonequilibrium
118: systems. Davidsen and Paczuski \cite{Pacs:EQSpatial} computed the
119: correlation between the epicenters of subsequent earthquakes, and
120: found the pdf for the distance $\Delta r$ between the subsequent
121: epicenters to be a powerlaw. Earlier, Krishnamurthy and Barma \cite{Supr1,Supr2}
122: found a powerlaw distribution for the long-range jumps in self-organized
123: interface depinning model.
124:
125: Corral \cite{Corr:PRL_Cluster,Corr:PRL_RG} and Bak et al. \cite{Bak:EQ}
126: combined the pdfs of magnitude and recurrence-time and derived a unified
127: scaling law. For events greater than a threshold $I_{c}$, then the
128: pdf for recurrence-time is
129:
130: \begin{equation}
131: P(\tau)=Rf(\tau R)\label{eq:earthquake1}\end{equation}
132: with \begin{equation}
133: f(\tau R)\approx CR^{-0.33}\left(\frac{{\tau}}{a}\right)^{-0.33}\exp(-\tau R/a)\label{eq:earthquake2}\end{equation}
134: and $R=R_{0}10^{-bM_{c}}$($b\approx1$) is the average number of
135: earthquakes (with $I>I_{c}$) per unit time, and $a$ is a constant.ll
136: these work however are model specific. In our paper we present a general
137: scaling laws which should be applicable to all the above systems and
138: other driven nonequilibrium systems. Here we also derive universal
139: relationships between the exponents. Note that $M_{c}=\log_{10}A_{c}$
140: is the Richter scale.
141:
142: Eqs. (\ref{eq:earthquake1}, \ref{eq:earthquake2}) form the unified
143: scaling law for the earthquakes. One of the intriguing questions is
144: whether similar scaling laws exist for other nonequilibrium systems.
145: Another important puzzle is the origin of the unified scaling law.
146: In the present paper we address the origin of pdfs in driven nonequilibrium
147: systems, and we derive universal scaling laws connecting the magnitudes,
148: recurrence-time, and spatial intervals of large events. These scaling
149: laws should be applicable to all the above systems and other driven
150: nonequilibrium systems. Here we also derive universal relationships
151: between the scaling exponents.
152:
153: In the above mentioned driven-nonequilibrium system the large events
154: do not have a typical energy scale. Hence \emph{under steady-state,
155: the distribution of magnitude of large events follows a powerlaw.}
156: That is,\[
157: P(I)\sim I^{-\alpha},\]
158: where $I$ is the strength of the events, and $\alpha>0$. Note however
159: that the events cannot have energy more than what is available in
160: the system, hence the distribution of the tail is decaying, mostly
161: exponential.
162:
163: For temporal correlation study, we take events larger than a threshold
164: value $I_{c}$. Driven nonequilibrium systems can be classified in
165: two classes, ones in which the events are uncorrelated, and the others
166: in which the events are correlated. In the uncorrelated case, the
167: pdf for recurrence of large events is Poissonian because the probability
168: of large event is small. If the large events occur with certain rate
169: $R$, then $1/R$ provides the time-scale for the recurrence of large
170: events; consequently the pdf will be $P(\tau)\sim R\exp(-R\tau)$.
171: The above feature is observed in BTW model \cite{Sanc:SOCorNot}.
172: In the correlated case, the nonequilibrium system is {}``\emph{critica}l''
173: under steady-state, and the system does not have any time-scale for
174: the recurrence time. Hence, the pdf of recurrence time is a powerlaw.
175: We can combine the pdf of magnitude and recurrence time using the
176: following law:\begin{equation}
177: P(I_{c},\tau)=I_{c}^{-\alpha}\tau^{-\beta}f(-\tau/\tau_{c}),\label{eq:Ptau}\end{equation}
178: where $f(x)$ is an universal function. For small $x$, $f(x)=1$,
179: and for large $x$, $f(x)\rightarrow\exp(-x)$. Further application
180: of scaling yields $\tau_{c}\sim I_{c}^{\gamma}$. It is expected that
181: when $I_{c1}>I_{c2}$, $P(I_{c1},\tau)<P(I_{c2},\tau)$, but $\tau_{c1}>\tau_{c2}$
182: \cite{Corr:PRL_Cluster,Corr:PhysicaA}. This observation implies that
183: $\gamma>0$. Corral's law for earthquake \cite{Corr:PRL_Cluster,Corr:PRL_RG}
184: is a good example which illustrates the above argument.
185:
186: The exponents $\alpha,\beta$, and $\gamma$ are not all independent.
187: Using conservation of probability\begin{equation}
188: \int P(I_{c},\tau)d\tau=1,\label{eq:P-conserve}\end{equation}
189: and equating the powers of $I_{c}$ on both sides, we obtain
190:
191: \begin{equation}
192: \alpha=\gamma(1-\beta).\label{eq:reln1}\end{equation}
193: As an example, the earthquake exponents of Corral \cite{Corr:PRL_Cluster,Corr:PhysicaA}
194: are $\alpha\approx2/3$, $\gamma\approx1$, $\beta\approx0.3$, which
195: satisfy the above relationship to a good approximation.
196:
197: Unfortunately Eq. (\ref{eq:P-conserve}) does not converge for $P(I_{c},\tau)$
198: with $\beta\ge1$ {[}Eq. (\ref{eq:Ptau}){]}. However we can circumvent
199: this difficulty by computing the higher-order moments of $\tau$.
200: If the integer part of $\beta$ is $n$, then we compute\begin{equation}
201: \left\langle \tau^{n}\right\rangle =\int\tau^{n}P(I_{c},\tau)d\tau.\label{eq:tau_n}\end{equation}
202: Clearly the above integral converges. By postulating $\left\langle \tau^{n}\right\rangle \sim I_{c}^{\delta_{n}}$,
203: and equating the exponents of $I_{c}$ on both sides, we obtain
204:
205: \begin{equation}
206: \alpha=\gamma(n+1-\beta)-\delta_{n}.\label{eq:reln2}\end{equation}
207: Hence the scaling arguments provide us an expression for $P(I_{c},\tau)$
208: {[}Eq. (\ref{eq:Ptau}){]} and relationships between exponents.
209:
210: The form of pdf of spatial separations between consecutive large events
211: can be derived in the similar lines. The events could be either spatially
212: uncorrelated or correlated. For uncorrelated case, the pdf of spatial
213: separation is expected to be Poissonian: $P(\Delta r)\sim\exp(-\Delta r/(\Delta r)_{c})/(\Delta r)_{c}$,
214: where the spatial density of the large events provide a length-scale
215: $(\Delta r)_{c}$. Nonequilibrium systems with spatially correlated
216: events are \emph{critical} under steady-state, and they do not have
217: any length-scale for the separation between the consecutive large
218: events. For this class of systems, the pdf is \cite{Pacs:EQSpatial}\begin{equation}
219: P(I_{c},\Delta r)=I_{c}^{-\alpha}(\Delta r)^{-\zeta}\exp(-\Delta r/(\Delta r)_{c}),\label{eq:Pdr}\end{equation}
220: with $(\Delta r)_{c}\sim I_{c}^{-\mu}$. If we assume dynamical scaling
221: $\tau\sim(\Delta r)^{z}$, then $P(I_{c},\Delta r)=P(I_{c},\tau)d\tau/d(\Delta r)$
222: yields \begin{equation}
223: \zeta=z(\beta-1)+1.\label{eq:spacetime}\end{equation}
224: Currently we do not know all the above parameters for any driven nonequilibrium
225: systems. It will be interesting to test this conjecture.
226:
227: It should be noted that earlier Maslov et al. \cite{Pacz:1f}, Paczuski
228: et al. \cite{Pacz:AvalacheDyna}, and others had derived relationships
229: between scaling exponents for several driven nonequilibrium systems,
230: e. g., Bak-Sneppen model, Directed Percolation etc. \cite{Pacz:AvalacheDyna}.
231: These relationships differ significantly from our equations, which
232: are based on large events.
233:
234: The scaling laws discussed above are quite general, and are expected
235: to hold for systems with correlated large events. Earthquake is one
236: such system for which Corral \cite{Corr:PRL_Cluster,Corr:PhysicaA}
237: , Bak et al. \cite{Bak:EQ}, and Davidsen and Paczuski \cite{Pacs:EQSpatial}
238: have computed some of the scaling exponents. Solar flares, Bak-Sneppen
239: model for punctuated evolution, and turbulence are other nonequilibrium
240: systems which too have similar behaviour. \emph{It should be noted
241: however that the exponents vary for different systems, but the form
242: of pdfs is the same for all them.} In the following discussion we
243: apply the above scaling ansatz to ion-channel currents in voltage-dependent
244: anion channels (VDAC), and test if it works.
245:
246: Our experimental setup on ion-channel current consists of two chambers
247: containing buffer solutions \emph{KCl}, $MgCl_{2}$, and HEPES. The
248: chambers are separated by a thin wall of a polystyrene cuvette. A
249: matched pair of \emph{Ag/AgCl} electrodes, connected to Axopatch-200
250: amplifier, are put into the buffer solution of two chambers. A bilayer
251: lipid membrane is formed on the pore. The protein molecules of VDAC
252: purified from rat brain mitochondria (De Pinto et al. \cite{Pinto:ion})
253: are put into the buffer solution of the outer chamber, after which
254: they are stirred magnetically. The VDAC molecules now reconstitute
255: the lipid membrane, and a channel is formed (Roos et al. \cite{Roos:ion}).
256: After one channel is formed, the buffer solution (containing more
257: protein molecules) of the outer chamber is replaced by fresh buffer
258: solution. Thus we form a single ion-channel.
259:
260: Now the voltage is applied across the membrane, and the resulting
261: current is recorded using the data acquisition software Clampex. This
262: method is same as that used by Bezrukov et al. \cite{Bezr:Ion} and
263: Banerjee et al \cite{MKV:Ion}. In our experiment we measure currents
264: with sampling frequency 1000 Hz
265: for various applied voltages in the range $-30mV$ to $30mV$. Typically
266: the current has open and closed states with a fractal structure \cite{Lieb:ion}.
267: We find the length of a single open state to be too small. Therefore,
268: we patch four long open-states and use the resulting timeseries for
269: our analysis (see Fig. \ref{Fig:timeseries}). Due to a large variation
270: of currents in various patches we perform the running average of the
271: timeseries over eleven neighboring points. The noise is obtained by
272: filtering the running average from the time-series. The statistics
273: is performed on the resulting noise time-series.
274:
275: We take the absolute value of the current fluctuations, and compute
276: the pdf of the magnitude of current fluctuation $P(I)$. The log-log
277: plot of $P(I)$ vs. $I$ is shown in Fig. \ref{Fig:PIvsI}. The slope
278: of the plot provides us the exponent $\alpha=3.86\pm0.08$.
279:
280: For the computation of temporal correlation, we filter the current
281: using a threshold value $I_{c}$. From the filtered current, we compute
282: time-interval between subsequent current peaks. From this set of time-interval
283: we compute the pdf of recurrence time $P(I_{c},\tau)$. We perform
284: this exercise for three values of $I_{c}:3\sigma,3.5\sigma,4\sigma$,
285: where $\sigma$ is the standard deviation of the current fluctuations.
286: The corresponding plots $P(I_{c},\tau)$ vs. $\tau$ are shown in
287: Fig. \ref{Fig:Ptvst}. All the three plots have approximately the
288: same exponent, $\beta=1.47\pm0.05$. Note that for higher $I_{c}$,
289: $P(I_{c},\tau)$ is lower, but the powerlaw extends up to higher $\tau$.
290: This feature is also seen in earthquakes.
291:
292: To obtain the universal function, we need to scale both $x$ and $y$
293: axes. If we scale only the $y$-axis to $P(I,\tau)I_{c}^{\alpha}\tau^{\beta}$
294: and plot of $P(I,\tau)I_{c}^{\alpha}\tau^{\beta}$ vs. $\tau$, then
295: we obtain a reasonable compactification along $y$ axis, but not along
296: $x$ axis. See Fig. \ref{Fig:norm1} for illustration. When we scale
297: $x$-axis by $\tau_{c}=I_{c}^{\gamma}$ and plot $P(I,\tau)I_{c}^{\alpha}\tau^{\beta}$vs.
298: $\tau I_{c}^{-\alpha}$ (see Fig. \ref{Fig:norm2}), the compactification
299: is quite good. The resulting plot is the universal function $f(x)$
300: of Eq. (\ref{eq:Ptau}). Note that we have taken $\gamma=\alpha$,
301: which is an approximation. The actual parameters $\gamma$ and $\delta_{1}$
302: {[}Eq. (\ref{eq:reln2}){]} are not accessible in the present experiment.
303: Since the current flow in the ion-channels is a localized, we do not
304: have the pdf for spatial separation between consecutive events.
305:
306: In summary, the pdf of magnitudes of events in a driven nonequilibrium
307: systems follow a powerlaw due to lack of any typical energy scale
308: for the events. The pdf of recurrence-time could be either an exponential
309: or a powerlaw depending on whether the events are temporally uncorrelated
310: or correlated. The pdf of recurrence time for the many correlated
311: nonequilibrium systems is a powerlaw due to lack of any timescale
312: in the system. When the pdfs of magnitudes and recurrence-time for
313: the large events are powerlaws, they could be combined into a single
314: universal scaling law. In the present paper we derive the universal
315: scaling law as well as relationships between the scaling exponents
316: for driven-nonequilibrium systems. We show that universal scaling
317: law is satisfied by ion-channel currents. Earlier, Corral \cite{Corr:PRL_Cluster,Corr:PRL_RG}
318: and Bak et al. \cite{Bak:EQ} had derived similar law for earthquakes.
319: We have also derived similar scaling law for the pdf of spatial separation
320: between subsequent large events. We should however keep in mind that
321: the exponents vary for different systems, but the form of pdfs is
322: the same for all them. \emph{}
323:
324: Temporal and spatial correlations between events are seen in solar
325: flares, turbulence, interface depinning models, music, punctuated
326: equilibria etc. A timeseries of major historical events would presumably
327: have temporal and spatial correlations. We should check how well scaling
328: laws describe the above correlations. Higher-order correlations between
329: the events should be another source for understanding nonequilibrium
330: systems.
331:
332: \begin{acknowledgments}
333: MKV acknowledges the kind hospitality of D. Ghosh, his group, and
334: his mother during his visits to Delhi University. MKV also thanks
335: Harshawardhan Wanare, A. Corral, Amit Dutta, Sutapa Mukherjee, Archisman
336: Ghosh, Amar Chandra, and Mustansir Barma for useful discussions.
337: \end{acknowledgments}
338: \bibliographystyle{apsrev}
339: %\bibliography{/home/mkv/res/bib/res}
340: \begin{thebibliography}{23}
341: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
342: \expandafter\ifx\csname bibnamefont\endcsname\relax
343: \def\bibnamefont#1{#1}\fi
344: \expandafter\ifx\csname bibfnamefont\endcsname\relax
345: \def\bibfnamefont#1{#1}\fi
346: \expandafter\ifx\csname citenamefont\endcsname\relax
347: \def\citenamefont#1{#1}\fi
348: \expandafter\ifx\csname url\endcsname\relax
349: \def\url#1{\texttt{#1}}\fi
350: \expandafter\ifx\csname urlprefix\endcsname\relax\def\urlprefix{URL }\fi
351: \providecommand{\bibinfo}[2]{#2}
352: \providecommand{\eprint}[2][]{\url{#2}}
353:
354: \bibitem[{\citenamefont{Bak et~al.}(1987)\citenamefont{Bak, Tang, and
355: Wiesenfeld}}]{Bak:SOC_PRL}
356: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Bak}},
357: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Tang}}, \bibnamefont{and}
358: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Wiesenfeld}},
359: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{59}},
360: \bibinfo{pages}{381} (\bibinfo{year}{1987}).
361:
362: \bibitem[{\citenamefont{Bak et~al.}(1988)\citenamefont{Bak, Tang, and
363: Wiesenfeld}}]{Bak:SOC_PRA}
364: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Bak}},
365: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Tang}}, \bibnamefont{and}
366: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Wiesenfeld}},
367: \bibinfo{journal}{Phys. Rev. A} \textbf{\bibinfo{volume}{38}},
368: \bibinfo{pages}{364} (\bibinfo{year}{1988}).
369:
370: \bibitem[{\citenamefont{Boffetta et~al.}(1999)\citenamefont{Boffetta, Carbone,
371: and G}}]{Boff:SolarflarePRL}
372: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Boffetta}},
373: \bibinfo{author}{\bibfnamefont{V.}~\bibnamefont{Carbone}}, \bibnamefont{and}
374: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{G}}, \bibinfo{journal}{Phys.
375: Rev. Lett.} \textbf{\bibinfo{volume}{83}}, \bibinfo{pages}{4662}
376: (\bibinfo{year}{1999}).
377:
378: \bibitem[{\citenamefont{S\'anchez et~al.}(2002)\citenamefont{S\'anchez, Newman,
379: and Carreeras}}]{Sanc:SOCorNot}
380: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{S\'anchez}},
381: \bibinfo{author}{\bibfnamefont{D.~E.} \bibnamefont{Newman}},
382: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{B.~A.}
383: \bibnamefont{Carreeras}}, \bibinfo{journal}{Phys. Rev. Lett.}
384: \textbf{\bibinfo{volume}{88}}, \bibinfo{pages}{068302}
385: (\bibinfo{year}{2002}).
386:
387: \bibitem[{\citenamefont{Lu and Hamilton}(1991)}]{Lu:Solarflare}
388: \bibinfo{author}{\bibfnamefont{E.~T.} \bibnamefont{Lu}} \bibnamefont{and}
389: \bibinfo{author}{\bibfnamefont{R.~J.} \bibnamefont{Hamilton}},
390: \bibinfo{journal}{Astrophys. J.} \textbf{\bibinfo{volume}{380}},
391: \bibinfo{pages}{L89} (\bibinfo{year}{1991}).
392:
393: \bibitem[{\citenamefont{Baiesi et~al.}(2006)\citenamefont{Baiesi, Paczuski, and
394: Stella}}]{Pacz:EQ_Intesity}
395: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Baiesi}},
396: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Paczuski}}, \bibnamefont{and}
397: \bibinfo{author}{\bibfnamefont{A.~L.} \bibnamefont{Stella}},
398: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{96}},
399: \bibinfo{pages}{051103} (\bibinfo{year}{2006}).
400:
401: \bibitem[{\citenamefont{Corral}(2004{\natexlab{a}})}]{Corr:PRL_Cluster}
402: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Corral}},
403: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{92}},
404: \bibinfo{pages}{108501} (\bibinfo{year}{2004}{\natexlab{a}}).
405:
406: \bibitem[{\citenamefont{Corral}(2005)}]{Corr:PRL_RG}
407: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Corral}},
408: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{95}},
409: \bibinfo{pages}{028501} (\bibinfo{year}{2005}).
410:
411: \bibitem[{\citenamefont{Bak et~al.}(2002)\citenamefont{Bak, Christensen, Danon,
412: and Scanlon}}]{Bak:EQ}
413: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Bak}},
414: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Christensen}},
415: \bibinfo{author}{\bibfnamefont{L.}~\bibnamefont{Danon}}, \bibnamefont{and}
416: \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Scanlon}},
417: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{88}},
418: \bibinfo{pages}{178501} (\bibinfo{year}{2002}).
419:
420: \bibitem[{\citenamefont{Gutenberg and Richter}(1936)}]{GuteRich}
421: \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Gutenberg}} \bibnamefont{and}
422: \bibinfo{author}{\bibfnamefont{C.~F.} \bibnamefont{Richter}},
423: \bibinfo{journal}{Science} \textbf{\bibinfo{volume}{83}},
424: \bibinfo{pages}{183} (\bibinfo{year}{1936}).
425:
426: \bibitem[{\citenamefont{Paczuski et~al.}(1996)\citenamefont{Paczuski, Maslov,
427: and Bak}}]{Pacz:AvalacheDyna}
428: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Paczuski}},
429: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Maslov}}, \bibnamefont{and}
430: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Bak}},
431: \bibinfo{journal}{Phys. Rev. E} \textbf{\bibinfo{volume}{53}},
432: \bibinfo{pages}{414} (\bibinfo{year}{1996}).
433:
434: \bibitem[{\citenamefont{Banerjee et~al.}(2006)\citenamefont{Banerjee, Verma,
435: Manna, and Ghosh}}]{MKV:Ion}
436: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Banerjee}},
437: \bibinfo{author}{\bibfnamefont{M.~K.} \bibnamefont{Verma}},
438: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Manna}}, \bibnamefont{and}
439: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Ghosh}},
440: \bibinfo{journal}{Europhys. Lett.} \textbf{\bibinfo{volume}{73}},
441: \bibinfo{pages}{457} (\bibinfo{year}{2006}).
442:
443: \bibitem[{\citenamefont{Sreenivasan et~al.}(2002)\citenamefont{Sreenivasan,
444: Bershadskii, and Niemela}}]{KRS:Wind}
445: \bibinfo{author}{\bibfnamefont{K.~R.} \bibnamefont{Sreenivasan}},
446: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Bershadskii}},
447: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{J.~J.}
448: \bibnamefont{Niemela}}, \bibinfo{journal}{Phys. Rev. E}
449: \textbf{\bibinfo{volume}{65}}, \bibinfo{pages}{56306} (\bibinfo{year}{2002}).
450:
451: \bibitem[{\citenamefont{Lowen and Teich}(1993)}]{Lowen:1f}
452: \bibinfo{author}{\bibfnamefont{S.~B.} \bibnamefont{Lowen}} \bibnamefont{and}
453: \bibinfo{author}{\bibfnamefont{M.~C.} \bibnamefont{Teich}},
454: \bibinfo{journal}{Phys. Rev. E} \textbf{\bibinfo{volume}{47}},
455: \bibinfo{pages}{992} (\bibinfo{year}{1993}).
456:
457: \bibitem[{\citenamefont{Maslow et~al.}(1994)\citenamefont{Maslow, Paczuski, and
458: Bak}}]{Pacz:1f}
459: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Maslow}},
460: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Paczuski}}, \bibnamefont{and}
461: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Bak}},
462: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{73}},
463: \bibinfo{pages}{2162} (\bibinfo{year}{1994}).
464:
465: \bibitem[{\citenamefont{Davidsen and Paczuski}(2005)}]{Pacs:EQSpatial}
466: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Davidsen}} \bibnamefont{and}
467: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Paczuski}},
468: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{94}},
469: \bibinfo{pages}{048501} (\bibinfo{year}{2005}).
470:
471: \bibitem[{\citenamefont{Krishnamurthy and Barma}(1996)}]{Supr1}
472: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Krishnamurthy}}
473: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Barma}},
474: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{76}},
475: \bibinfo{pages}{423} (\bibinfo{year}{1996}).
476:
477: \bibitem[{\citenamefont{Krishnamurthy and Barma}(1998)}]{Supr2}
478: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Krishnamurthy}}
479: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Barma}},
480: \bibinfo{journal}{Phys. Rev. E} \textbf{\bibinfo{volume}{57}},
481: \bibinfo{pages}{2949} (\bibinfo{year}{1998}).
482:
483: \bibitem[{\citenamefont{Corral}(2004{\natexlab{b}})}]{Corr:PhysicaA}
484: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Corral}},
485: \bibinfo{journal}{Physica A} \textbf{\bibinfo{volume}{340}},
486: \bibinfo{pages}{590} (\bibinfo{year}{2004}{\natexlab{b}}).
487:
488: \bibitem[{\citenamefont{De~Pinto et~al.}(1987)\citenamefont{De~Pinto, Prezioso,
489: and Palmieri}}]{Pinto:ion}
490: \bibinfo{author}{\bibfnamefont{V.}~\bibnamefont{De~Pinto}},
491: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Prezioso}}, \bibnamefont{and}
492: \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Palmieri}},
493: \bibinfo{journal}{Biochim. Biophys. Acta} \textbf{\bibinfo{volume}{905}},
494: \bibinfo{pages}{499} (\bibinfo{year}{1987}).
495:
496: \bibitem[{\citenamefont{Roos et~al.}(1982)\citenamefont{Roos, Benz, and
497: Brdiczka}}]{Roos:ion}
498: \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Roos}},
499: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Benz}}, \bibnamefont{and}
500: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Brdiczka}},
501: \bibinfo{journal}{Biochim. Biophys. Acta} \textbf{\bibinfo{volume}{686}},
502: \bibinfo{pages}{204} (\bibinfo{year}{1982}).
503:
504: \bibitem[{\citenamefont{Bezrukov and Winterhalter}(2000)}]{Bezr:Ion}
505: \bibinfo{author}{\bibfnamefont{S.~M.} \bibnamefont{Bezrukov}} \bibnamefont{and}
506: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Winterhalter}},
507: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{85}},
508: \bibinfo{pages}{202} (\bibinfo{year}{2000}).
509:
510: \bibitem[{\citenamefont{Liebovitch et~al.}(2001)\citenamefont{Liebovitch,
511: Scheurle, Rusek, and Zochowski}}]{Lieb:ion}
512: \bibinfo{author}{\bibfnamefont{L.~S.} \bibnamefont{Liebovitch}},
513: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Scheurle}},
514: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Rusek}}, \bibnamefont{and}
515: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Zochowski}},
516: \bibinfo{journal}{Methods} \textbf{\bibinfo{volume}{24}},
517: \bibinfo{pages}{359} (\bibinfo{year}{2001}).
518:
519: \end{thebibliography}
520:
521:
522:
523: \begin{center}\textbf{\Large Figure Caption}\end{center}{\Large \par}
524:
525: Figure 1: Timeseries of ion-channel currents in open state of Voltage-dependent
526: Anion Channels (VDAC). Four parts have been patched here for extending
527: the data range.
528:
529: Figure 2: Plot of the probability distribution of magnitudes of current
530: fluctuation in ion-channel. The best fit is a powerlaw with slope
531: $\alpha=3.86\pm0.08$.
532:
533: Figure 3: Plot of the probability distribution of recurrence time
534: $P(I_{c},\tau)$ for three different threshold currents $I_{c}=3\sigma,3.5\sigma,4\sigma$,
535: where $\sigma$ is the standard deviations. The best fit for all the
536: three plots is a powerlaw with the exponent $\beta=1.47\pm0.05$.
537:
538: Figure 4: Plot of $P(I_{c},\tau)\tau^{\beta}$ vs. $\tau$.
539:
540: Figure 5: Plot of $P(I_{c},\tau)\tau^{\beta}I_{c}^{\alpha}$ vs. $\tau I_{c}^{-\alpha}$.
541: The resulting plot is the universal function $f(x)$.
542:
543: \pagebreak
544:
545: \begin{center}\textbf{\Large Figures}\end{center}{\Large \par}
546:
547: %
548: \begin{figure}
549: \includegraphics{figures/timeseries.ps}
550:
551:
552: \caption{\label{Fig:timeseries} Timeseries of ion-channel currents in open
553: state of Voltage-dependent Anion Channels (VDAC). Four parts have
554: been patched here for extending the data range.}
555: \end{figure}
556: \pagebreak
557:
558: %
559: \begin{figure}
560: \includegraphics{figures/PIvsI.ps}
561:
562:
563: \caption{\label{Fig:PIvsI} Plot of the probability distribution of magnitudes
564: of current fluctuation $P(I)$ in ion-channel. The best fit is a powerlaw
565: with slope $\alpha=3.86\pm0.08$.}
566: \end{figure}
567:
568:
569: \pagebreak
570:
571: %
572: \begin{figure}
573: \includegraphics[%
574: scale=0.5]{figures/Ptvst.ps}
575:
576:
577: \caption{\label{Fig:Ptvst} Plot of the probability distribution of recurrence
578: time $P(I_{c},\tau)$for three different threshold currents $I_{c}=3\sigma,3.5\sigma,4\sigma$,
579: where $\sigma$ is the standard deviations. The best fit for all the
580: three plots is a powerlaw with the exponent $\beta=1.47\pm0.05$. }
581: \end{figure}
582:
583:
584: %
585: \begin{figure}
586: \includegraphics[%
587: scale=0.8]{figures/norm1.ps}
588:
589:
590: \caption{\label{Fig:norm1} Plot of $P(I_{c},\tau)\tau^{\beta}$ vs. $\tau$.}
591: \end{figure}
592: \pagebreak
593:
594: %
595: \begin{figure}
596: \includegraphics[%
597: scale=0.8]{figures/norm2.ps}
598:
599:
600: \caption{\label{Fig:norm2} Plot of $P(I_{c},\tau)\tau^{\beta}I_{c}^{\alpha}$
601: vs. $\tau I_{c}^{-\alpha}$. The resulting plot is the universal function
602: $f(x)$.}
603: \end{figure}
604:
605: \end{document}
606: