1. Introduction
Over decades, stochastic growth processes, kinetic roughening phenomena, and fluctuating surfaces or interfaces have been attracting constant attention. The most prominent examples include the deposition of a substance on a surface and the growth of the corresponding phase boundary; propagation of flame, smoke, and solidification fronts; growth of vicinal surfaces and bacterial colonies; erosion of landscapes and seabed profiles; molecular beam epitaxy; and many others, see [
1,
2,
3,
4,
5,
6,
7,
8,
9,
10,
11,
12,
13] and references therein.
Another vast area of research is that of diffusion and random walks in a random environment, such as disordered, inhomogeneous, porous, or turbulent media, see, e.g., [
14,
15,
16,
17,
18,
19,
20,
21,
22,
23,
24].
In this paper, we study a simple model of a random walk on a rough fluctuating surface. We consider the Fokker–Planck equation for a particle in a uniform gravitational field. The surface is modeled by the generalized Edwards–Wilkinson linear stochastic equation for the height field [
1]. The generalized model involves two arbitrary exponents,
and
, related to the spectrum and the dispersion law of the height field, respectively. A detailed description of the model and its relation to various special cases is given in
Section 2.
Using the general Martin–Siggia–Rose–de Dominicis–Janssen theorem, the original stochastic problem is reformulated as a certain field-theoretic model. This allows one to apply the well-developed formalism of Feynman diagrammatic techniques, renormalization theory, and a renormalization group (RG). The model is shown to be multiplicatively renormalizable, so that the RG equation can be derived in a standard way. The corresponding renormalization constants and the RG functions (anomalous dimensions and
functions) are explicitly calculated in the leading one-loop order of the RG perturbation theory. These issues are discussed in
Section 3 and
Section 4.
The RG equations have two Gaussian (free) fixed points and two nontrivial ones. Those points are infrared (IR) attractive depending on the values of the parameters
and
, which implies the existence of scaling (self-similar) asymptotic regimes in the IR range (long times and large distances) for the various response and correlation functions of the model (
Section 4). The critical dimensions for those regimes are found exactly as functions of
and
. As an indicative application, the time dependence of the mean-square radius of a cloud of randomly walking particles is obtained (
Section 5). It is described by a power law with the exponent that depends on the fixed point, is known exactly as a function of
and
, and for nontrivial points, it differs from the ordinary random walk:
.
Some implications and possible generalizations are discussed in
Section 6.
2. Description of the Model
We consider the random walk of a point particle on a two-dimensional rough surface embedded into the -dimensional space. The particle is located on the surface on the height , where is the particle’s coordinate projection on the d-dimensional substrate, . Thus, d is an arbitrary (for generality) dimension of the substrate space.
While the coordinates with determine the particle’s location, the coordinate with is not treated as an independent Cartesian coordinate but is restricted to the surface, . This setup excludes the possibility for a particle to “jump off” or “escape” the surface; such interesting phenomena are not included in our simple model.
The basic stochastic equation of motion of a particle located at the point
in an external drift field
F has the form [
14,
15,
16,
17,
18,
19,
20,
21,
22,
23,
24]:
Here,
is a Gaussian noise with zero mean and a given pair correlation function,
will play a role of the diffusion coefficient, and
F is an external drift field (a force or an advecting velocity, depending on the specific context).
1The probability distribution function
satisfies the Fokker–Planck equation
Here and below, summation over repeated indices is implied.
From physics reasons, the drift
F in a gravitational field should obey the
symmetry and the invariance with the respect to the shift
const, so that it should be built of gradients of the field
h. Thus, in the simplest linear approximation, it is taken in the form
with the parameter
proportional to the particle’s mass
m and the gravitational acceleration
g. Possible higher-order corrections to the linear approximation should also be constructed from gradients of the field
h and obey the
symmetry, for example,
. They have higher canonical dimensions in comparison with (
3) (see
Section 3), are IR irrelevant (in the sense of Wilson), and should be dropped in the analysis of IR scaling.
Careful interpretation of the gradient
for a rough surface and the very existence (in a rigorous mathematical sense) of corresponding continuous equations is a serious problem. Recently, important progress was achieved for the Kardar–Parisi–Zhang (KPZ) model, see [
25,
26,
27,
28,
29] and references therein.
From a more practical physical point of view, there is a very small microscopical scale
a below which the field
h becomes smooth and differentiable. In our and the KPZ’s treatment, this scale is tacitly set to zero, so that the field becomes rough. There is an apparent analogy with the well-known dissipative anomaly in turbulence, see, e.g., [
30]. Practically, in this study, we use the formal perturbation theory, where this problem does not arise, and the expression (
3) is applied without further comments.
The simplest model of surface roughening, proposed within the context of landscape erosion, is the one due to Edwards and Wilkinson [
1]. In the continuous formulation, it is described by the diffusion-type stochastic equation for the height field
:
where
is (a kind of) surface tension coefficient,
is the Laplace operator, and
f is a Gaussian random noise with zero mean and a given pair correlation function. The most popular choices are the white noise
with the positive amplitude
and the quenched noise; the simplified version of the latter is
In this paper, we consider a generalized equation
written here in the symbolic notation with
k being the wave number,
2 while the correlation function is taken in a power-like form:
Here,
and
y are arbitrary exponents and
d is the dimension of space. Clearly, the choice
,
corresponds to the model (
4) and (
5); as we will see, the model (
4) and (
6) can also be obtained from (
7) and (
8).
For a linear stochastic equation with a Gaussian additive random noise, the field
h is also a Gaussian field defined by its pair correlation function. For the model (
7) and (
8), the latter has the following form in the Fourier (
–
) representation
In the second relation, we introduced the new variables: the exponent
and the amplitudes
,
, defined by the relations
They are convenient, in particular, because the equal-time correlation function
involves the parameters
,
, while the dispersion law
is expressed only via
,
.
The choice
can be justified by the ideas of self-organized criticality (SOC), according to which the evolution of a sandpile surface is not an ordinary diffusion-type process but involves several discrete steps: expectation period, reaching a threshold, and avalanche, see, e.g., [
31,
32,
33].
Thus, according to [
32], self-organized critical dynamical systems give rise to the so-called
noise because the characteristic size of an avalanche is related to its lifetime via a power law
, where the exponent
is the rate at which the event propagates across the system, see also, e.g., Section 1.3.2 in the book [
31] and papers [
33,
34]. In the
–
representation, this corresponds to the dispersion law (
12) with the exponent
. It is also worth noting that the
noise appears also in models of random walks in random media, see, e.g., [
14,
15].
The model (
9) includes two special cases interesting on their own. In the limit
and
fixed, the function
becomes independent of the frequency
, and the field
becomes white in time. Indeed, one obtains in the (
t–
) representation
Here, the exponent plays a role of a Hölder’s exponent that measures “roughness” of the field h (“Batchelor limit” corresponds to a smooth field).
In the limit
and
fixed, the function
in (
11) remains finite, so that (
9) tends to
which corresponds to the time-independent (quenched or frozen) field
h. Surprisingly enough, for
, this reproduces the model (
4) and (
6) where one has
.
Substituting the gravitational force (
3) with the random height field from (
7) and (
8) into the Fokker–Planck equation (
2) turns the latter into a stochastic equation in its own right. It has the form
where the random field
can be interpreted as the density of walking particles, while the role of the (deterministic) probability distribution function
is now conveyed to the linear response function, see Equation (
44) in
Section 5.
This completes formulation of the problem.
3. Field-Theoretic Formulation and Renormalization of the Model
According to the general theorem (see, e.g., Section 5.3 in the monograph [
35]), the full stochastic problem (
8), (
15) is equivalent to the field-theoretic model for the doubled set of fields
with the de Dominicis–Janssen action functional:
Here,
is the correlator (
8),
is the density field,
h is the height field, and
,
are the corresponding Martin–Siggia–Rose response fields; all the needed integrations over their arguments
and summations over repeated indices are implied. The field-theoretic formulation means that various correlation and response functions of the original stochastic problem are represented by functional averages with the weight
. The field
can easily be removed by Gaussian integration, then
would be replaced with
with
from (
9), but the expanded representation (
17) is more convenient for the renormalization purposes. The constant
can be removed by rescaling of the fields
and other parameters. Thus, in the following, with no loss of generality, we set
.
The model (
16) and (
17) corresponds to Feynman diagrammatic technique with bare propagators
,
,
(the latter does not enter into relevant diagrams) and the only vertex
.
It is well known that an analysis of ultraviolet (UV) divergences is based on an analysis of canonical dimensions, see, e.g., [
35] (Sections 1.15 and 1.16). In contrast to conventional static models, dynamic ones have two independent scales: a time scale
and a spatial scale
, see [
35] (Sections 1.17 and 5.14). Thus, the canonical dimension of any quantity
F (a field or a parameter) is determined by two numbers: the frequency dimension
and the momentum dimension
:
The dimensions are found from obvious normalization conditions
and from the requirement that all terms in the action functional be dimensionless with respect to both the canonical dimensions separately. The total canonical dimension is defined as
(the coefficient 2 follows from the relation
in the free theory). In the renormalization procedure,
plays the same role as the conventional (momentum) dimension does in static models, see Section 5.14 in [
35].
Canonical dimensions of all the fields and parameters of our model are given in
Table 1. It also involves renormalized parameters (without subscript “0”) and the reference mass
, an additional parameter of the renormalized theory; they all will appear later on.
Note that for the fields
,
all these dimensions can be unambiguously defined only for the product
. Formally, this follows from the invariance of the action functional (
16) under the dilatation
,
.
As can be seen from
Table 1, the model becomes logarithmic (both coupling constants
,
become dimensionless) for
(or equivalently for
) and arbitrary
d.
3 According to the general strategy of renormalization, the exponents
,
y, or
that “measure” the deviation from the logarithmicity should be treated as formal small parameters of the same order. The UV divergences manifest themselves as singularities at
, etc., in the correlation functions; in the one-loop approximation, they have the form of simple poles.
The total canonical dimension of a certain 1-irreducible Green’s function is given by
where
are the numbers of the fields
entering the Green’s function and
are their total canonical dimensions.
The formal index of divergence is the total dimension of the Green’s function in the logarithmic theory (), that is, . Superficial UV divergences, whose removal requires introducing counterterms, can be present in the Green’s function if is a non-negative integer.
When analyzing the divergences in the model (
16) and (
17), the following additional considerations should be taken into account, see, e.g., [
35] (Section 5.15) and [
36] (Section 1.4).
(i) For any dynamic model of this type, all the 1-irreducible functions without the response fields contain closed circuits of retarded propagators and vanish. Thus, it is sufficient to consider the functions with .
(ii) For all non-vanishing functions,
(otherwise no diagrams can be constructed). Formally, this is a consequence of the invariance of the action functional (
16) with respect to dilatation
,
.
(iii) Using integration by parts, one derivative in the vertex can be moved onto the field , i.e., . Thus, in any 1-irreducible diagram, each external field or “releases” the external momentum, and the real index of divergence decreases by the corresponding number of units, i.e., . Furthermore, these fields enter the counterterms only in the form of spatial gradients. This observation excludes the counterterms and , the latter allowed by the formal index for .
(iv) It is clear that the fields , do not affect the statistics of the field h. In the field-theoretic terms, this “passivity” means that any 1-irreducible Green’s function with , , and vanishes: no corresponding diagrams can be constructed.
Taking into account these considerations, one obtains:
(we recall that
, so that only
is indicated).
Then, the straightforward analysis shows that the superficial divergences in our model are present only in the 1-irreducible functions
and
, and the corresponding counterterms necessarily contract to the forms
(
,
) and
(
,
). Such terms are already present in the action (
16), which means that our model (
16) and (
17) is multiplicatively renormalizable with only two independent renormalization constants
and
.
The renormalized action has the form
which is naturally reproduced as renormalization of the field
h and the coefficient
; no renormalization of the product
is needed:
The functional (
17) is not renormalized,
, but it should be expressed in renormalized variables, taking into account Equations (
8) and (
9):
where the renormalization mass
is introduced so that renormalized couplings
g and
u are completely dimensionless. Then, it follows from the absence of renormalization of
that
Along with (
21), this finally gives the following relations:
We calculated the renormalization constants
and
in the leading one-loop approximation (the first order of the perturbative expansion in
g). It is sufficient to find them for
, because the anomalous dimensions in the minimal subtraction (MS) renormalization scheme are independent of the parameters such as
and
y, while the exponent
y alone provides UV regularization. Then, one obtains:
with the higher-order corrections in
g. Here,
,
is the surface area of the unit sphere in
d-dimensional space. It is convenient to absorb overall factors into the coupling constant
g, which gives
For
, the expressions (
25) and (
26) would be infinite sums, see, e.g., [
37].
4. RG Equations, RG Functions, and Fixed Points
Because our model is multiplicatively renormalizable, the corresponding RG equations are derived in a standard fashion. In particular, for a certain renormalized (full or connected) Green’s function
, the RG equation reads
Here, the ellipsis stands for other variables (times and coordinates or frequencies and momenta), , for any variable x, and the sum runs over all fields .
The coefficients in the RG differential operator (
27)—the anomalous dimensions
and the
functions—are defined as
where
is the differential operation
at fixed bare (unrenormalized) parameters, see, e.g., Sections 1.24 and 1.25 in the monograph [
35].
From (
21)–(
24) and definition (
28), it follows that
From (
30) and the one-loop result (
26), one obtains
with the higher-order corrections in
g.
The IR asymptotic behavior of the Green’s functions is determined by IR attractive fixed points of the corresponding RG equations. The coordinates of fixed points
,
are found from the requirement that all the
functions vanish simultaneously:
The type of a fixed point is determined by the matrix of derivatives at the given point : for an IR attractive point, all the eigenvalues should have positive real parts.
An analysis of the expressions (
32) reveals four fixed points:
(i) Gaussian (free) fixed point:
(ii) Nontrivial fixed point:
The point (i) is IR attractive for , while the point (ii) is IR attractive for .
Two more points are found in the following way. In order to explore the limiting case
with
fixed, we have to pass to new variables:
and
. For this case, we obtain
Finding the zeros of the functions, we find two additional fixed points:
(iii) Gaussian (free) fixed point:
(iv) Nontrivial fixed point:
The point (iii) is IR attractive if , and the point (iv) is IR attractive if or .
The general stability pattern of the fixed points in the
–
plane is shown in
Figure 1.
In the one-loop approximation, the regions of IR stability for all the points are given by sectors that cover the full plane without gaps or overlaps between them.
Some remarks are in order. Clearly, the Gaussian points correspond to cases in which the dynamics of the field are not affected by the statistics of the height field h (only in the leading order of the IR asymptotic behavior!). In these cases, we deal with an ordinary random walk.
The point (iv) corresponds to the limiting case (
13) when the field
h, in comparison with
, behaves as if it was
correlated in time.
However, we did not find a nontrivial point that would correspond to the frozen limit (
14). This follows from the fact that the function
in (
32) becomes trivial for
:
. A similar triviality was observed earlier in models of diffusion in time-independent potential vector fields where it was shown to be exact in all orders of perturbation theory [
18,
19]. Because those models have a close formal resemblance with the limit (
14) of our model and its special case (
4) and (
6), we believe that in the latter cases
is also trivial exactly.
5. Critical Dimensions and Scaling Behavior
The existence of IR attractive fixed points of the RG equations implies the existence of the scaling behavior of the correlation functions in the IR range.
In dynamical models, the critical dimension of any quantity
F (a field or a parameter) is given by the expression (see, e.g., Sections 5.16 and 6.7 in [
35] and Section 2.1 in [
36])
(with the standard normalization convention that
). Here and below,
denotes the value of the anomalous dimension
at a fixed point.
For the Gaussian points (i) and (iii), one has
For the fixed point (ii), one obtains the exact results from the relation (
29) and definition (
30):
As already mentioned, the point (iv) corresponds to the limit (
13), where the propagator
becomes
-correlated in time. As a result, closed circuits of retarded propagators
appear in almost all diagrams relevant for a renormalization procedure and they therefore vanish. The only exception is the one-loop diagram contributing to
. Thus, one has
identically, while
is given exactly by the one-loop expression, cf. the discussion of Kraichnan’s rapid-change model of passive scalar advection [
38]. Then, one readily derives the exact expressions for the critical dimensions:
As an illustrative application, consider the mean-square distance of a random walker on a rough surface. For such a particle that started moving at
from the origin
, it is given by
where
is a later time and
is the corresponding current position. Substituting the scaling representation for the linear response function
gives
Taking into account the exact relation
, valid for all fixed points (i)-(iv), one arrives at the spreading law
with the exact expressions
for the points (i), (iii),
for (ii), and
for (iv).
6. Conclusions
We studied a model of a random walk of a particle on a rough fluctuating surface described by the Fokker–Planck equation for a particle in a constant gravitational field while the surface was modeled by the (generalized) Edwards–Wilkinson model. The full stochastic problem, (
2), (
3), (
7) and (
8), is mapped onto a multiplicatively renormalizable field-theoretic model (
16) and (
17).
The corresponding RG equations reveal two Gaussian (free) and two nontrivial fixed points, which means that the system exhibits various types of IR scaling behavior (long times and large distances). Although the practical calculation is confined within the leading one-loop approximation, the main critical dimensions are found exactly.
As an illustrative example, we considered the mean-square displacement of a walking particle (in another interpretation, the radius of particles’ cloud). It shows that the particle is not trapped in a finite area but travels all across the system with a spreading law similar to the ordinary random walk but, in general, with different exponents, see (
46) and the text below.
As one can see, even a comparatively simple model demonstrates interesting types of IR behavior. Thus, it is interesting to study more involved situations. There are several directions for possible generalizations.
Linear stochastic equations such as (
4) and (
7) (corresponding to Gaussian statistics for the height field) can be replaced by nonlinear models, such as the KPZ [
2] or Pavlik’s [
5,
8] ones.
Although our expressions (
41) and (
42) for the critical dimensions are exact, they are derived within perturbation theory based on the assumption that the expansion parameters
and
are small. Then, it is supposed that the one-loop pattern of fixed points is qualitatively correct. However, in some cases, a crossover in the scaling behavior occurs for finite values of parameters analogous to
and
[
37,
39]. In the field-theoretic approach, this effect can be related to the appearance of composite operators with negative dimensions [
37]. This issue requires a special investigation.
On some occasions, the motion of a particle is not an ordinary random walk (
1) but is described, e.g., by Lévy flights, see, e.g., [
21]. This possibility is supported by the ideas of self-organized criticality that the underlying surface evolves via avalanches [
31,
32,
33,
34], while the particle can slide upon the surface. If so, it is natural to replace the Laplace operator in the Fokker–Planck equation (
2) with a fractional derivative,
, with a certain new exponent
.
It is especially interesting to include anisotropy (as a consequence of an overall tilt of the surface). This can be done by describing the field
h by the Pastor-Satorras–Rothman model for an eroding landscape [
9,
10] or the Hwa–Kardar model of a running sandpile [
40,
41].
This work remains for the future and is partly in progress.