ANOMALOUS DIFFUSION AND POWER LAW STATISTICS

Rytis Kazakevičius1, Julius Ruseckas1

1 Institute of Theoretical Physics and Astronomy, Vilnius University, Lithuania

One of the fundamental aspects of complex systems are the transport properties. Transport properties in complex systems are usually characterized by anomalous scaling, that is by a non-linear time dependency in the growth of the variance, $\sigma^2(t) \sim t^{\mu}$, where t is the elapsed time. This condition is known as an anomalous diffusion. In contrast to the anomalous diffusion, in a typical diffusion process the variance of the particle position (or mean squared displacement) is a linear function of time. Anomalous diffusion is classified by its power law exponent $\mu$. If $\mu > 1$, the phenomenon is called super-diffusion. Super-diffusion has been experimentally observed in a study of tracer particles in a two-dimensional rotating flow [1]. If $\mu < 1$, the particle undergoes sub-diffusion. Sub-diffusion has been proposed as a measure of macromolecular crowding in the cytoplasm [2]. Experimental studies shows that anomalous scaling of variance and power law statistics (power law spectral density (PSD) and power law steady state probability densities (PDF)) can occur together[3]. Usually, these processes are modeled as separate and independent. So there is a need for model describing a system with bought anomalous diffusion and power law statistics.

Figure 1
Fig. 1. Dependence of the variance $\sigma^2(t)$ of the signal generated by Eq. (2) on time t. The stability index of Lévy noise $\alpha=1$ and power law exponent $\lambda=3$. Gray and black strait lines show the power law dependence of the variance on time, $\sigma^2 \sim t^{\mu}$. (a) Black curve corresponds to $\eta=2.8$, gray curve to $\eta=2.9$. The numerically determined values of the index $\mu$ are $\mu=1.35$ and $\mu=0.9$, respectively. (b) Black curve corresponds to $\eta=2.4$, gray curve to $\eta=2.1$. The numerically determined values of the index $\mu$ are $\mu=0.5$. Eq. (2) is solved with reflective boundary conditions $x_{min}=1$, $x_{max}=10^4$.

Stochastic differential equation (SDE) that generate signals with power law statistical properties have been proposed

$$dx = \left(\eta - \frac{1}{2}\lambda\right)x^{2\eta - 1}dt + x^{\eta}dW(t). \tag{1}$$

However, proposed SDE generated signals variance do not exhibit anomalous power law growth [4]. The simplest known model of super-diffusion is Lévy $\alpha$ stable process $L_t^{\alpha}$. We modified Eq. (1) by replacing Wiener $W_t$ process to Lévy stable process $L_t^{\alpha}$

$$dx = \gamma(\eta, \lambda, \alpha) x^{\alpha(\eta - 1) + 1} dt + x^{\eta} dL_t^{\alpha}. \tag{2}$$

Proposed SDE generate signals with power law PDF $(P(x) \sim x^{-\lambda})$ and spectral density S(f)

$$S(f) \sim \frac{1}{f^{\beta}}, \qquad \beta = 1 + \frac{\lambda - 3}{\alpha(\eta - 1)}. \tag{3}$$

In addition, numerical results show that due to presence of the multiplicative Lévy $\alpha$ stable noise process in Eq. (2) both sub-diffusion and super-diffusion can occur (see Fig. 1). This is in contrast to the SDEs with additive stable stochastic process, when only sub-diffusion is possible unless more complex non-Markovian models are used [5].

In conclusion, we have proposed nonlinear stochastic differential equations driven by Lévy $\alpha$ stable process that generate signals exhibiting power law statistical properties. In addition, such nonlinear SDEs can lead to Lévy flights with anomalous diffusion, both sub-diffusion and supper-diffusion.


[1] T. H. Solomon, E. R. Weeks, H. L. Swinney, Observation of anomalous diffusion and Lévy flights in a two-dimensional rotating flow, Phys. Rev. Lett. 71 3975 (1993).

[2] M. Weiss, M. Elsner, F. Kartberg et al., Anomalous subdiffusion is a measure for cytoplasmic crowding in living cells, Biophys. J. 87 3518 (2004).

[3] C. Scalliet, A. Gnoli, A. Puglisi et al., Cages and Anomalous Diffusion in Vibrated Dense Granular Media, Phys. Rev. Lett. 114, 198001 (2015).

[4] A. G. Cherstvy and R. Metzler, Nonergodicity, fluctuations, and criticality in heterogeneous diffusion processes, Phys. Rev. E 90, 012134 (2014).

[5] R. Metzler and J. Klafter, The random walk's guide to anomalous diffusion: a fractional dynamics approach, Phys. Rep. 339, 1-77 (2014).