Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
APPLICATIONS
G.A. Pavliotis
Department of Mathematics
Imperial College London
London SW7 2AZ, UK
June 9, 2011
2
Contents
Preface vii
1 Introduction 1
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Historical Overview . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.3 The One-Dimensional Random Walk . . . . . . . . . . . . . . . . 3
1.4 Stochastic Modeling of Deterministic Chaos . . . . . . . . . . . . 6
1.5 Why Randomness . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.6 Discussion and Bibliography . . . . . . . . . . . . . . . . . . . . 7
1.7 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
i
ii CONTENTS
4 Markov Processes 57
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.2 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.3 Definition of a Markov Process . . . . . . . . . . . . . . . . . . . 62
4.4 The Chapman-Kolmogorov Equation . . . . . . . . . . . . . . . . 64
4.5 The Generator of a Markov Processes . . . . . . . . . . . . . . . 67
4.5.1 The Adjoint Semigroup . . . . . . . . . . . . . . . . . . 69
4.6 Ergodic Markov processes . . . . . . . . . . . . . . . . . . . . . 70
4.6.1 Stationary Markov Processes . . . . . . . . . . . . . . . . 72
4.7 Discussion and Bibliography . . . . . . . . . . . . . . . . . . . . 73
4.8 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
5 Diffusion Processes 77
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
5.2 Definition of a Diffusion Process . . . . . . . . . . . . . . . . . . 77
5.3 The Backward and Forward Kolmogorov Equations . . . . . . . . 79
5.3.1 The Backward Kolmogorov Equation . . . . . . . . . . . 79
5.3.2 The Forward Kolmogorov Equation . . . . . . . . . . . . 82
5.4 Multidimensional Diffusion Processes . . . . . . . . . . . . . . . 84
5.5 Connection with Stochastic Differential Equations . . . . . . . . . 85
5.6 Examples of Diffusion Processes . . . . . . . . . . . . . . . . . . 86
5.7 Discussion and Bibliography . . . . . . . . . . . . . . . . . . . . 86
5.8 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
CONTENTS iii
9 The Mean First Passage time and Exit Time Problems 191
9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
9.2 Brownian Motion in a Bistable Potential . . . . . . . . . . . . . . 191
9.3 The Mean First Passage Time . . . . . . . . . . . . . . . . . . . . 194
9.3.1 The Boundary Value Problem for the MFPT . . . . . . . . 194
9.3.2 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . 197
9.4 Escape from a Potential Barrier . . . . . . . . . . . . . . . . . . . 199
9.4.1 Calculation of the Reaction Rate in the Overdamped Regime200
9.4.2 The Intermediate Regime: = O(1) . . . . . . . . . . . 201
9.4.3 Calculation of the Reaction Rate in the energy-diffusion-
limited regime . . . . . . . . . . . . . . . . . . . . . . . 202
9.5 Discussion and Bibliography . . . . . . . . . . . . . . . . . . . . 202
9.6 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
The purpose of these notes is to present various results and techniques from the
theory of stochastic processes and are useful in the study of stochastic problems
in physics, chemistry and other areas. These notes have been used for several
years for a course on applied stochastic processes offered to fourth year and to
MSc students in applied mathematics at the department of mathematics, Imperial
College London.
G.A. Pavliotis
London, December 2010
vii
viii PREFACE
Chapter 1
Introduction
1.1 Introduction
In this chapter we introduce some of the concepts and techniques that we will study
in this book. In Section 1.2 we present a brief historical overview on the develop-
ment of the theory of stochastic processes in the twentieth century. In Section 1.3
we introduce the one-dimensional random walk an we use this example in order to
introduce several concepts such Brownian motion, the Markov property. In Sec-
tion 1.4 we discuss about the stochastic modeling of deterministic chaos. Some
comments on the role of probabilistic modeling in the physical sciences are of-
fered in Section 1.5. Discussion and bibliographical comments are presented in
Section 1.6. Exercises are included in Section 1.7.
1
2 CHAPTER 1. INTRODUCTION
the particle (molecule, pollen grain...). Furthermore, he introduced the idea that it
makes more sense to talk about the probability of finding the particle at position x
at time t, rather than about individual trajectories.
In his work many of the main aspects of the modern theory of stochastic pro-
cesses can be found:
d2 x dx
m = 6a + ,
dt2 dt
where is a random force. It can be shown that there is complete agreement be-
tween Einsteins theory and Langevins theory. The theory of Brownian motion
was developed independently by Smoluchowski, who also performed several ex-
periments.
The approaches of Langevin and Einstein represent the two main approaches
in the theory of stochastic processes:
1
= (F (x)) + : (A(x)),
t 2
The theory of stochastic processes was developed during the 20th century by sev-
eral mathematicians and physicists including Smoluchowksi, Planck, Kramers,
Chandrasekhar, Wiener, Kolmogorov, Ito, Doob.
0 5 10 15 20 25 30 35 40 45 50
20
10
10
20
30
40
50
0 100 200 300 400 500 600 700 800 900 1000
2
mean of 1000 paths
5 individual paths
1.5
0.5
U(t)
0
0.5
1.5
0 0.2 0.4 0.6 0.8 1
t
Every path of the random walk is different: it depends on the outcome of a se-
quence of independent random experiments. We can compute statistics by gen-
erating a large number of paths and computing averages. For example, E(Sn ) =
0, E(Sn2 ) = n. The paths of the random walk (without the linear interpolation) are
not continuous: the random walk has a jump of size 1 at each time step. This is an
example of a discrete time, discrete space stochastic processes. The random walk
is a time-homogeneous Markov process. If we take a large number of steps, the
random walk starts looking like a continuous time process with continuous paths.
We can quantify this observation by introducing an appropriate rescaled pro-
cess and by taking an appropriate limit. Consider the sequence of continuous time
stochastic processes
1
Ztn := Snt .
n
In the limit as n , the sequence {Ztn } converges (in some appropriate sense,
that will be made precise in later chapters) to a Brownian motion with diffusion
2
coefficient D = x 1
2t = 2 . Brownian motion W (t) is a continuous time stochastic
processes with continuous paths that starts at 0 (W (0) = 0) and has indepen-
dent, normally. distributed Gaussian increments. We can simulate the Brownian
6 CHAPTER 1. INTRODUCTION
2
= D 2, (y, 0) = (y x).
t y
To describe systems for which we are not confident about the validity of our
mathematical model.
1.7 Exercises
1. Read the papers by Einstein, Ornstein-Uhlenbeck, Doob etc.
2. Write a computer program for generating the random walk in one and two di-
mensions. Study numerically the Brownian limit and compute the statistics of
the random walk.
8 CHAPTER 1. INTRODUCTION
Chapter 2
2.1 Introduction
In this chapter we put together some basic definitions and results from probability
theory that will be used later on. In Section 2.2 we give some basic definitions
from the theory of probability. In Section 2.3 we present some properties of ran-
dom variables. In Section 2.4 we introduce the concept of conditional expectation
and in Section 2.5 we define the characteristic function, one of the most useful
tools in the study of (sums of) random variables. Some explicit calculations for
the multivariate Gaussian distribution are presented in Section 2.6. Different types
of convergence and the basic limit theorems of the theory of probability are dis-
cussed in Section 2.7. Discussion and bibliographical comments are presented in
Section 2.8. Exercises are included in Section 2.9.
Definition 2.2.1. The set of all possible outcomes of an experiment is called the
sample space and is denoted by .
9
10 CHAPTER 2. ELEMENTS OF PROBABILITY THEORY
i. F;
ii. if A F then Ac F;
iii. If A, B F then A B F.
A1 , . . . An F ni=1 Ai F, ni=1 Ai F.
i. F;
ii. if A F then Ac F;
iii. If A1 , A2 , F then
i=1 Ai F.
i. P() = 0, P() = 1;
Definition 2.2.8. The triple , F, P comprising a set , a -algebra F of sub-
sets of and a probability measure P on (, F) is a called a probability space.
Example 2.2.10. Take = [0, 1], F = B([0, 1]), P = Leb([0, 1]). Then
(, F, P) is a probability space.
12 CHAPTER 2. ELEMENTS OF PROBABILITY THEORY
Bi Bj = , i 6= j and iI Bi = .
Proposition 2.2.13. Law of total probability. For any event A and any partition
{Bi : i I} we have
X
P(A) = P(A|Bi )P(Bi ).
iI
The proof of this result is left as an exercise. In many cases the calculation of
the probability of an event is simplified by choosing an appropriate partition of
and using the law of total probability.
Let (, F, P) be a probability space and fix B F. Then P(|B) defines a
probability measure on F. Indeed, we have that
P(|B) = 0, P(|B) = 1
{ : X() A} =: {X A} (2.1)
Let U be a topological space. We will use the notation B(U ) to denote the Borel
algebra of U : the smallest algebra containing all open sets of U . Every ran-
dom variable from a probability space (, F, ) to a measurable space (E, B(E))
induces a probability measure on E:
Consider the case where E = R equipped with the Borel algebra. In this
case a random variable is defined to be a function X : R such that
{ : X() 6 x} F x R.
with > 0.
with a < b.
Definition 2.3.9. Two random variables X and Y are independent if the events
{ | X() 6 x} and { | Y () 6 y} are independent for all x, y R.
2F
The mixed derivative of the distribution function fX,Y (x, y) := xy (x, y), if it
exists, is called the joint PDF of the random vector {X, Y }:
Z x Z y
FX,Y (x, y) = fX,Y (x, y) dxdy.
and
fX,Y (x, y) = fX (x)fY (y).
We can extend the above definition to random vectors of arbitrary finite dimen-
sions. Let X be a random variable from (, F, ) to (Rd , B(Rd )). The (joint)
distribution function FX Rd [0, 1] is defined as
E(X) = b (2.7)
Since the mean and variance specify completely a Gaussian random variable on
R, the Gaussian is commonly denoted by N (m, ). The standard normal random
variable is N (0, 1). Similarly, since the mean and covariance matrix completely
specify a Gaussian random variable on Rd , the Gaussian is commonly denoted by
N (m, ).
Some analytical calculations for Gaussian random variables will be presented
in Section 2.6.
We can define E[f (X)|G] and the conditional probability P[X F |G] = E[IF (X)|G],
where IF is the indicator function of F , in a similar manner.
We list some of the most important properties of conditional expectation.
(c) (Order) If X1 , X2 are integrable and X1 6 X2 a.s., then E(X1 |G) 6 E(X2 |G)
a.s.
(f) (Convergence) Let {Xn }n=1 be a sequence of random variables such that, for
all n, |Xn | 6 Z where Z is integrable. If Xn X a.s., then E(Xn |G)
E(X|G) a.s. and in L1 .
Proof. See Exercise 10.
For a continuous random variable for which the distribution function F has a den-
sity, dF () = p()d, (2.9) gives
Z
(t) = eit p() d.
R
From the properties of the Fourier transform we conclude that the characteristic
function determines uniquely the distribution function of the random variable, in
the sense that there is a one-to-one correspondance between F () and (t). Fur-
thermore, in the exercises at the end of the chapter the reader is asked to prove the
following two results.
Lemma 2.5.1. Let {X1 , X2 , . . . Xn } be independent random variables with char-
P
acteristic functions j (t), j = 1, . . . n and let Y = nj=1 Xj with characteristic
function Y (t). Then
Y (t) = nj=1 j (t).
Lemma 2.5.2. Let X be a random variable with characteristic function (t) and
assume that it has finite moments. Then
1 (k)
E(X k ) = (0).
ik
20 CHAPTER 2. ELEMENTS OF PROBABILITY THEORY
Theorem 2.6.1. Let b Rd and Rdd a symmetric and positive definite ma-
trix. Let X be the multivariate Gaussian random variable with probability density
function
1 1 1
,b (x) = exp h (x b), x bi .
Z 2
Then
EX = b
and
E((X EX) (X EX)) = .
Proof. i. From the spectral theorem for symmetric positive definite matrices
we have that there exists a diagonal matrix with positive entries and an
orthogonal matrix B such that
1 = B T 1 B.
h1 z, zi = hB T 1 Bz, zi
= h1 Bz, Bzi = h1 y, yi
d
X
= 1 2
i yi .
i=1
2.6. GAUSSIAN RANDOM VARIABLES 21
,b (x) dx = ,b (B T y + b) dy
d
Y
1 1
= p exp i yi2 dyi .
(2)d/2 det() i=1 2
Consequently
Z
EX = x,b (x) dx
d
ZR
= (B T y + b),b (B T y + b) dy
RZd
= b ,b (B T y + b) dy = b.
Rd
more, z = B T y. We calculate
Z
E((Xi bi )(Xj bj )) = zi zj ,b (z + b) dz
Rd
Z X X
1 1 X 1 2
= p Bki yk
y Bmi ym exp
(2)d/2 det() Rd k m
2
Z !
1 X 1 X 1 2
= p Bki Bmj yk ym exp y d
(2)d/2 det() k,m Rd 2
X
= Bki Bmj k km
k,m
= ij .
iii. Let y be a multivariate Gaussian random variable with mean 0 and covari-
ance I. Let also C = B . We have that = CC T = C T C. We have
that
X = CY + b.
To see this, we first note that X is Gaussian since it is given through a linear
transformation of a Gaussian random variable. Furthermore,
Now we have:
1
= eihb,ti e 2 hCt,Cti
1 T Cti
= eihb,ti e 2 ht,C
1
= eihb,ti e 2 ht,ti .
Consequently,
1
(t) = eihb,ti 2 ht,ti .
2.7. TYPES OF CONVERGENCE AND LIMIT THEOREMS 23
(c) Zn converges to Z in Lp if
p
lim E Zn Z = 0.
n+
Let {Xn }n=1 be iid random variables with EXn = V . Then, the strong law
of large numbers states that average of the sum of the iid converges to V with
probability one:
N
1 X
P lim Xn = V = 1.
N + N
n=1
The strong law of large numbers provides us with information about the behav-
ior of a sum of random variables (or, a large number or repetitions of the same
experiment) on average. We can also study fluctuations around the average be-
havior. Indeed, let E(Xn V )2 = 2 . Define the centered iid random variables
P
Yn = Xn V . Then, the sequence of random variables 1N N n=1 Yn converges
in distribution to a N (0, 1) random variable:
N
! Z
1 X a
1 1 2
lim P Yn 6 a = e 2 x dx.
n+ N n=1 2
2.9 Exercises
1. Show that the intersection of a family of -algebras is a -algebra.
3. Calculate the mean, variance and characteristic function of the following prob-
ability density functions.
with > 0.
(b) The uniform distribution with density
1
f (x) = ba a < x < b,
0 x
/ (a, b),
with a < b.
26 CHAPTER 2. ELEMENTS OF PROBABILITY THEORY
5. Let X and Y be Gaussian random variables. Show that they are uncorrelated if
and only if they are independent.
7. Let X be a discrete random variable taking vales on the set of nonnegative inte-
P
gers with probability mass function pk = P(X = k) with pk > 0, + k=0 pk =
1. The generating function is defined as
+
X
X
g(s) = E(s ) = pk s k .
k=0
2.9. EXERCISES 27
e k
pk = P(X = k) = , k = 0, 1, 2, . . . and > 0.
k!
(c) Prove that the generating function of a sum of independent nonnegative
integer valued random variables is the product of their generating func-
tions.
8. Write a computer program for studying the law of large numbers and the central
limit theorem. Investigate numerically the rate of convergence of these two
theorems.
9. Study the properties of Gaussian measures on separable Hilbert spaces from [?,
Ch. 2].
3.1 Introduction
In this chapter we present some basic results form the theory of stochastic pro-
cesses and we investigate the properties of some of the standard stochastic pro-
cesses in continuous time. In Section 3.2 we give the definition of a stochastic pro-
cess. In Section 3.3 we present some properties of stationary stochastic processes.
In Section 3.4 we introduce Brownian motion and study some of its properties.
Various examples of stochastic processes in continuous time are presented in Sec-
tion 3.5. The Karhunen-Loeve expansion, one of the most useful tools for repre-
senting stochastic processes and random fields, is presented in Section 3.6. Further
discussion and bibliographical comments are presented in Section 3.7. Section 3.8
contains exercises.
29
30 CHAPTER 3. BASICS OF THE THEORY OF STOCHASTIC PROCESSES
The set T can be either discrete, for example the set of positive integers Z+ , or
continuous, T = [0, +). The state space E will usually be Rd equipped with the
algebra of Borel sets.
A stochastic process X may be viewed as a function of both t T and .
We will sometimes write X(t), X(t, ) or Xt () instead of Xt . For a fixed sample
point , the function Xt () : T 7 E is called a sample path (realization,
trajectory) of the process X.
F (x) = P(X(ti ) 6 xi , i = 1, . . . , k)
with x = (x1 , . . . , xk ).
Definition 3.2.3. We will say that two processes Xt and Yt are equivalent if they
have same finite dimensional distributions.
From the above definition we conclude that the Finite dimensional distributions
of a Gaussian continuous time stochastic process are Gaussian with PFG
n/2 1/2 1 1
k ,Kk (x) = (2) (detKk ) exp hKk (x k ), x k i ,
2
where x = (x1 , x2 , . . . xk ).
1
In fact, what we need is the stochastic process to be separable. See the discussion in Section 3.7
3.3. STATIONARY PROCESSES 31
almost surely. In fact, Birkhoffs ergodic theorem states that, for any function f
such that Ef (Y0 ) < +, we have that
N 1
1 X
lim f (Xj ) = Ef (Y0 ), (3.1)
N + N
j=0
32 CHAPTER 3. BASICS OF THE THEORY OF STOCHASTIC PROCESSES
Ergodic strictly stationary processes satisfy (3.1) Hence, we can calculate the
statistics of a sequence stochastic process Xn using a single sample path, provided
that it is long enough (N 1).
Example 3.3.3. Let Z be a random variable and define the stochastic process
Xn = Z, n = 0, 1, 2, . . . . Then Xn is a strictly stationary process (see Exercise 2).
We can calculate the long time average of this stochastic process:
N 1 N 1
1 X 1 X
Xj = Z = Z,
N N
j=0 j=0
which is independent of N and does not converge to the mean of the stochastic pro-
cesses EXn = EZ (assuming that it is finite), or any other deterministic number.
This is an example of a non-ergodic processes.
Remark 3.3.5. Let Xt be a strictly stationary stochastic process with finite second
moment (i.e. Xt L2 ). The definition of strict stationarity implies that EXt = , a
constant, and E((Xt )(Xs )) = C(ts). Hence, a strictly stationary process
with finite second moment is also stationary in the wide sense. The converse is not
true.
Example 3.3.6.
Let Y0 , Y1 , . . . be a sequence of independent, identically distributed random vari-
ables and consider the stochastic process Xn = Yn . From Example 3.3.2 we
know that this is a strictly stationary process, irrespective of whether Y0 is such
that EY02 < +. Assume now that EY0 = 0 and EY02 = 2 < +. Then
Xn is a second order stationary process with mean zero and correlation function
R(k) = 2 k0 . Notice that in this case we have no correlation between the values
of the stochastic process at different times n and k.
Example 3.3.7. Let Z be a single random variable and consider the stochastic
process Xn = Z, n = 0, 1, 2, . . . . From Example 3.3.3 we know that this is a
strictly stationary process irrespective of whether E|Z|2 < + or not. Assume
now that EZ = 0, EZ 2 = 2 . Then Xn becomes a second order stationary
process with R(k) = 2 . Notice that in this case the values of our stochastic
process at different times are strongly correlated.
We will see in Section 3.3.3 that for second order stationary processes, ergod-
icity is related to fast decay of correlations. In the first of the examples above,
there was no correlation between our stochastic processes at different times and
the stochastic process is ergodic. On the contrary, in our second example there is
very strong correlation between the stochastic process at different times and this
process is not ergodic.
Remark 3.3.8. The first two moments of a Gaussian process are sufficient for a
34 CHAPTER 3. BASICS OF THE THEORY OF STOCHASTIC PROCESSES
lim E|Xt+h Xt |2 = 0.
h0
Lemma 3.3.9. Assume that the covariance function C(t) of a second order sta-
tionary process is continuous at t = 0. Then it is continuous for all t R. Fur-
thermore, the continuity of C(t) is equivalent to the continuity of the process Xt in
the L2 -sense.
Notice that form (3.4) we immediately conclude that C(0) > C(h), h R.
The Fourier transform of the covariance function of a second order stationary
process always exists. This enables us to study second order stationary processes
using tools from Fourier analysis. To make the link between second order station-
ary processes and Fourier analysis we will use Bochners theorem, which applies
to all nonnegative functions.
for all n N, t1 , . . . tn R, c1 , . . . cn C.
3.3. STATIONARY PROCESSES 35
= E|Xtc |2 > 0.
In the following we will assume that the spectral measure is absolutely contin-
uous with respect to the Lebesgue measure on R with density f (x), i.e. (dx) =
f (x)dx. The Fourier transform f (x) of the covariance function is called the spec-
tral density of the process:
Z
1
f (x) = eitx C(t) dt.
2
From (3.6) it follows that that the autocorrelation function of a mean zero, second
order stationary process is given by the inverse Fourier transform of the spectral
density: Z
C(t) = eitx f (x) dx. (3.7)
There are various cases where the experimentally measured quantity is the spec-
tral density (or power spectrum) of a stationary stochastic process. Conversely,
36 CHAPTER 3. BASICS OF THE THEORY OF STOCHASTIC PROCESSES
The slower the decay of the correlation function, the larger the correlation time
is. Notice that when the correlations do not decay sufficiently fast so that C(t) is
integrable, then the correlation time will be infinite.
Example 3.3.14. Consider a mean zero, second order stationary process with cor-
relation function
R(t) = R(0)e|t| (3.8)
D
where > 0. We will write R(0) = where D > 0. The spectral density of this
process is:
Z +
1 D
f (x) = eixt e|t| dt
2
Z0 Z +
1 D
= eixt et dt + eixt et dt
2
0
1 D 1 1
= +
2 ix + ix +
D 1
= .
x2 + 2
This function is called the Cauchy or the Lorentz distribution. The correlation
time is (we have that R(0) = D/)
Z
cor = et dt = 1 .
0
The Ornstein Uhlenbeck process is used as a model for the velocity of a Brown-
ian particle. It is of interest to calculate the statistics of the position of the Brownian
particle, i.e. of the integral
Z t
X(t) = Y (s) ds, (3.9)
0
where the symmetry of the function R(u) was used in the last step.
which, from our assumption on R(t), is a finite quantity. 2 The above calculation
suggests that, for T 1, we have that
Z t 2
E X(t) dt 2DT.
0
This implies that, at sufficiently long times, the mean square displacement of the
integral of the ergodic second order stationary process Xt scales linearly in time,
with proportionality coefficient 2D.
2
Notice however that we do not know whether it is nonzero. This requires a separate argument.
3.4. BROWNIAN MOTION 39
Assume that Xt is the velocity of a (Brownian) particle. In this case, the inte-
gral of Xt
Z t
Zt = Xs ds,
0
represents the particle position. From our calculation above we conclude that
EZt2 = 2Dt.
where Z Z
D= R(t) dt = E(Xt X0 ) dt (3.14)
0 0
is the diffusion coefficient. Thus, one expects that at sufficiently long times and
under appropriate assumptions on the correlation function, the time integral of a
stationary process will approximate a Brownian motion with diffusion coefficient
D. The diffusion coefficient is an example of a transport coefficient and (3.14) is
an example of the Green-Kubo formula: a transport coefficient can be calculated
in terms of the time integral of an appropriate autocorrelation function. In the
case of the diffusion coefficient we need to calculate the integral of the velocity
autocorrelation function.
i. W (0) = 0.
ii. W (t) has independent increments.
iii. For every t > s > 0 W (t) W (s) has a Gaussian distribution with
mean 0 and variance t s. That is, the density of the random variable
W (t) W (s) is
1
2 x2
g(x; t, s) = 2(t s) exp ; (3.15)
2(t s)
2
mean of 1000 paths
5 individual paths
1.5
0.5
U(t)
0
0.5
1.5
0 0.2 0.4 0.6 0.8 1
t
Remark 3.4.5. Equivalently, we could have defined the one dimensional standard
Brownian motion as a stochastic process on a probability space , F, P with
continuous paths for almost all , and Gaussian finite dimensional distri-
butions with zero mean and covariance E(Wti Wtj ) = min(ti , tj ). One can then
show that Definition 3.4.1 follows from the above definition.
where [] denotes the integer part of a number. Then Wtn converges weakly, as
n + to a one dimensional standard Brownian motion.
Brownian motion is a Gaussian process. For the ddimensional Brownian mo-
tion, and for I the d d dimensional identity, we have (see (2.7) and (2.8))
EW (t) = 0 t > 0
and
E (W (t) W (s)) (W (t) W (s)) = (t s)I. (3.17)
Moreover,
E W (t) W (s) = min(t, s)I. (3.18)
From the formula for the Gaussian density g(x, t s), eqn. (3.15), we immedi-
ately conclude that W (t) W (s) and W (t + u) W (s + u) have the same pdf.
Consequently, Brownian motion has stationary increments. Notice, however, that
Brownian motion itself is not a stationary process. Since W (t) = W (t) W (0),
the pdf of W (t) is
1 x2 /2t
g(x, t) = e .
2t
We can easily calculate all moments of the Brownian motion:
Z +
1 2
E(xn (t)) = xn ex /2t dx
2t
n
1.3 . . . (n 1)tn/2 , n even,
=
0, n odd.
ii. (Shifting). For each c > 0 Wc+t Wc , t > 0 is a Brownian motion which is
independent of Wu , u [0, c].
iii. (Time reversal). Define Xt = W1t W1 , t [0, 1]. Then (Xt , t [0, 1]) =
(Wt , t [0, 1]) in law.
We emphasize that the equivalence in the above theorem holds in law and not
in a pathwise sense.
We can also add a drift and change the diffusion coefficient of the Brownian
motion: we will define a Brownian motion with drift and variance 2 as the
process
Xt = t + Wt .
The mean and variance of Xt are
dXt = dt + dWt .
Lemma 3.4.8. Let W (t) be a standard Brownian motion and consider the process
V (t) = et W (e2t ).
Then V (t) is a Gaussian stationary process with mean 0 and correlation function
For the proof of this result we first need to show that time changed Gaussian
processes are also Gaussian.
Lemma 3.4.9. Let X(t) be a Gaussian stochastic process and let Y (t) = X(f (t))
where f (t) is a strictly increasing function. Then Y (t) is also a Gaussian process.
Proof. We need to show that, for all positive integers N and all sequences of times
{t1 , t2 , . . . tN } the random vector
which is Gaussian for all N and all choices of times s1 , s2 , . . . sN . Hence Y (t) is
also Gaussian.
Proof of Lemma 3.4.8. The fact that V (t) is mean zero follows immediately
from the fact that W (t) is mean zero. To show that the correlation function of V (t)
is given by (3.19), we calculate
E(V (t)V (s)) = ets E(W (e2t )W (e2s )) = ets min(e2t , e2s )
= e|ts| .
The Gaussianity of the process V (t) follows from Lemma 3.4.9 (notice that the
transformation that gives V (t) in terms of W (t) is invertible and we can write
W (s) = s1/2 V ( 12 ln(s))).
Conversely, we can write the Brownian motion as a time change of the Brownian
bridge:
t
Wt = (t + 1)B , t > 0.
1+t
1 2H
E(WtH WsH ) = s + t2H |t s|2H . (3.24)
2
H
(Wt , t > 0) = (H WtH , t > 0), > 0, (3.25)
where Z
fn = f (x)en (x) dx.
The convergence is in L2 ():
N
X
lim
f (x) fn en (x)
= 0.
N
n=1 L2 ()
It turns out that we can obtain a similar expansion for an L2 mean zero process
which is continuous in the L2 sense:
For simplicity we will take T = [0, 1]. Let R(t, s) = E(Xt Xs ) be the autocorrela-
tion function. Notice that from (3.26) it follows that R(t, s) is continuous in both t
and s (exercise 21).
Let us assume an expansion of the form
X
Xt () = n ()en (t), t [0, 1] (3.27)
n=1
3.6. THE KARHUNEN-LOEVE EXPANSION 47
where {en } 2
n=1 is an orthonormal basis in L (0, 1). The random variables n are
calculated as
Z 1 Z 1X
Xt ek (t) dt = n en (t)ek (t) dt
0 0 n=1
X
= n nk = k ,
n=1
E(n m ) = n nm ,
where {n }
n=1 are positive numbers that will be determined later.
Assuming that an expansion of the form (3.27) exists, we can calculate
X
!
X
R(t, s) = E(Xt Xs ) = E k ek (t) e (s)
k=1 =1
X
X
= E (k ) ek (t)e (s)
k=1 =1
X
= k ek (t)ek (s).
k=1
Hence, in order to prove the expansion (3.27) we need to study the eigenvalue
problem for the integral operator R : L2 [0, 1] 7 L2 [0, 1]. It easy to check that
this operator is self-adjoint ((Rf, h) = (f, Rh) for all f, h L2 (0, 1)) and non-
negative (Rf, f > 0 for all f L2 (0, 1)). Hence, all its eigenvalues are real
and nonnegative. Furthermore, it is a compact operator (if {n } n=1 is a bounded
2
sequence in L (0, 1), then {Rn }n=1 has a convergent subsequence). The spec-
tral theorem for compact, self-adjoint operators implies that R has a countable
sequence of eigenvalues tending to 0. Furthermore, for every f L2 (0, 1) we can
write
X
f = f0 + fn en (t),
n=1
where Z 1
n = Xt en (t) dt, En = 0, E(n m ) = nm . (3.31)
0
Proof. The fact that En = 0 follows from the fact that Xt is mean zero. The
orthogonality of the random variables {n }
n=1 follows from the orthogonality of
3.6. THE KARHUNEN-LOEVE EXPANSION 49
the eigenfunctions of R:
Z 1Z 1
E(n m ) = E Xt Xs en (t)em (s) dtds
0 0
Z 1Z 1
= R(t, s)en (t)em (s) dsdt
0 0
Z 1
= n en (s)em (s) ds
0
= n nm .
P
Consider now the partial sum SN = N n=1 n en (t).
XN
= R(t, t) k |ek (t)|2 0,
k=1
by Mercers theorem.
Remark 3.6.2. Let Xt be a Gaussian second order process with continuous co-
variance R(t, s). Then the random variables {k } k=1 are Gaussian, since they
are defined through the time integral of a Gaussian processes. Furthermore, since
they are Gaussian and orthogonal, they are also independent. Hence, for Gaussian
processes the Karhunen-Loeve expansion becomes:
+ p
X
Xt = k k ek (t), (3.32)
k=1
where {k }
k=1 are independent N (0, 1) random variables.
Let us assume that n > 0 (it is easy to check that 0 is not an eigenvalue). Upon
setting t = 0 we obtain n (0) = 0. The eigenvalue problem can be rewritten in
the form
Z t Z 1
sn (s) ds + t n (s) ds = n n (t).
0 t
n (t) = n n (t),
The inner product can be obtained through polarization. This norm enables us to
measure the regularity of the function f (t).3 Let Xt be a mean zero second order
(i.e. with finite second moment) process with continuous autocorrelation function.
Define the space H := L2 ((, P ), H (0, 1)) with (semi)norm
X
kXt k2 = EkXt k2H = |k |1 . (3.34)
k
Notice that the regularity of the stochastic process Xt depends on the decay of the
R1
eigenvalues of the integral operator R := 0 R(t, s) ds.
As an example, consider the L2 -regularity of Brownian motion. From Exam-
ple 3.6.3 we know that k k2 . Consequently, from (3.34) we get that, in order
for Wt to be an element of the space H , we need that
X
|k |2(1) < +,
k
from which we obtain that < 1/2. This is consistent with the Holder continuity
of Brownian motion from Theorem 3.4.6. 4
3.8 Exercises
1. Let Y0 , Y1 , . . . be a sequence of independent, identically distributed random
variables and consider the stochastic process Xn = Yn .
3
Think of R as being the inverse of the Laplacian with periodic boundary conditions. In this case
H coincides with the standard fractional Sobolev space.
4
Notice, however, that Wieners theorem refers to a.s. Holder continuity, whereas the calculation
presented in this section is about L2 -continuity.
52 CHAPTER 3. BASICS OF THE THEORY OF STOCHASTIC PROCESSES
Xn = a1 n + a2 n1 + . . . am nm+1 .
(a) Calculate the mean, variance and the covariance function of Xn . Show
that it is a weakly stationary process.
(b) Set ak = 1/ m for k = 1, . . . m. Calculate the covariance function and
study the cases m = 1 and m +.
5. Let W (t) be a standard one dimensional Brownian motion. Calculate the fol-
lowing expectations.
where {j , j }N
j=1 are positive real numbers.
(a) Calculate the spectral density and the correlaction time of this process.
(b) Show that the assumptions of Theorem 3.3.17 are satisfied and use the
argument presented in Section 3.3.3 (i.e. the Green-Kubo formula) to cal-
Rt
culate the diffusion coefficient of the process Zt = 0 Xs ds.
(c) Under what assumptions on the coefficients {j , j }N
j=1 can you study
the above questions in the limit N +?
10. Let W (t) be the standard one-dimensional Brownian motion and let , s1 , s2 >
0. Calculate
11. Let Wt be a one dimensional Brownian motion and let , > 0 and define
St = et+Wt .
14. Use Lemma 3.4.8 to calculate the distribution function of the stationary Ornstein-
Uhlenbeck process.
15. Calculate the mean and the correlation function of the integral of a standard
Brownian motion Z t
Yt = Ws ds.
0
18. The autocorrelation function of the velocity Y (t) a Brownian particle moving
in a harmonic potential V (x) = 12 02 x2 is
1
R(t) = e|t| cos(|t|) sin(|t|) ,
p
where is the friction coefficient and = 02 2 .
3.8. EXERCISES 55
19. Show the scaling property (3.25) of the fractional Brownian motion.
20. Use Theorem (3.4.4) to show that there does not exist a continuous modification
of the Poisson process.
21. Show that the correlation function of a process Xt satisfying (3.26) is continu-
ous in both t and s.
22. Let Xt be a stochastic process satisfying (3.26) and R(t, s) its correlation func-
tion. Show that the integral operator R : L2 [0, 1] 7 L2 [0, 1]
Z 1
Rf := R(t, s)f (s) ds (3.35)
0
is self-adjoint and nonnegative. Show that all of its eigenvalues are real and
nonnegative. Show that eigenfunctions corresponding to different eigenvalues
are orthogonal.
Let R : L2 [0, 1] 7 L2 [0, 1] be the operator defined in (3.35) with R(t, s) being
continuous both in t and s. Show that it is a Hilbert-Schmidt operator.
24. Let Xt a mean zero second order stationary process defined in the interval [0, T ]
with continuous covariance R(t) and let {n }+n=1 be the eigenvalues of the
covariance operator. Show that
X
n = T R(0).
n=1
25. Calculate the Karhunen-Loeve expansion for a second order stochastic process
with correlation function R(t, s) = ts.
56 CHAPTER 3. BASICS OF THE THEORY OF STOCHASTIC PROCESSES
26. Calculate the Karhunen-Loeve expansion of the Brownian bridge on [0, 1].
27. Let Xt , t [0, T ] be a second order process with continuous covariance and
Karhunen-Loeve expansion
X
Xt = k ek (t).
k=1
Markov Processes
4.1 Introduction
In this chapter we will study some of the basic properties of Markov stochastic
processes. In Section 4.2 we present various examples of Markov processes, in
discrete and continuous time. In Section 4.3 we give the precise definition of a
Markov process. In Section 4.4 we derive the Chapman-Kolmogorov equation,
the fundamental equation in the theory of Markov processes. In Section 4.5 we
introduce the concept of the generator of a Markov process. In Section 4.6 we study
ergodic Markov processes. Discussion and bibliographical remarks are presented
in Section 4.7 and exercises can be found in Section 4.8.
4.2 Examples
Roughly speaking, a Markov process is a stochastic process that retains no mem-
ory of where it has been in the past: only the current state of a Markov process
can influence where it will go next. A bit more precisely: a Markov process is
a stochastic process for which, given the present, past and future are statistically
independent.
Perhaps the simplest example of a Markov process is that of a random walk
in one dimension. We defined the one dimensional random walk as the sum of
independent, mean zero and variance 1 random variables i , i = 1, . . . :
N
X
XN = n , X0 = 0.
n=1
57
58 CHAPTER 4. MARKOV PROCESSES
1 In
words, the probability that the random walk will be at in+m at time n + m
depends only on its current value (at time n) and not on how it got there.
The random walk is an example of a discrete time Markov chain:
for all h > 0. A continuous-time, discrete state space Markov process is called a
continuous-time Markov chain.
Example 4.2.3. The Brownian motion is a Markov process with conditional prob-
ability density
1 |x y|2
p(y, t|x, s) := p(Wt = y|Ws = x) = p exp . (4.4)
2(t s) 2(t s)
1
In fact, it is sufficient to take m = 1 in (4.1). See Exercise 1.
4.2. EXAMPLES 59
2(e2t e2s )
Z y |et xes |2
1
= p e 2(e2t (1e2(ts) ) d
2e2t (1 e2(ts) )
Z y |x|2
1 2(ts) )
= p e 2(1e d.
2(1 e2(ts) )
Consequently, the transition probability density for the OU process is given by the
formula
p(y, t|x, s) = P(Vt 6 y|Vs = x)
y
!
1 |y xe(ts) |2
= p exp .
2(1 e2(ts) ) 2(1 e2(ts) )
We will refer to the matrix P = {pij } as the transition matrix. It is each to check
that the transition matrix is a stochastic matrix, i.e. it has nonnegative entries and
P
j pij = 1. Similarly, we can define the n-step transition matrix Pn = {pij (n)}
as
pij (n) = P(Xm+n = j|Xm = i).
We can study the evolution of a Markov chain through the Chapman-Kolmogorov
equation: X
pij (m + n) = pik (m)pkj (n). (4.7)
k
(n)
Indeed, let i := P(Xn = i). The (possibly infinite dimensional) vector n
determines the state of the Markov chain at time n. A simple consequence of the
Chapman-Kolmogorov equation is that we can write an evolution equation for the
vector (n)
(n) = (0) P n , (4.8)
where P n denotes the nth power of the matrix P . Hence in order to calculate the
state of the Markov chain at time n all we need is the initial distribution 0 and the
transition matrix P . Componentwise, the above equation can be written as
(n)
X (0)
j = i ij (n).
i
In particular,
pij (t) = P(Xt = j|X0 = i).
The Chapman-Kolmogorov equation for a continuous time Markov chain is
dpij X
= pik (t)gkj , (4.9)
dt
k
4.2. EXAMPLES 61
where the matrix G is called the generator of the Markov chain. Equation (4.9)
can also be written in matrix notation:
dP
= Pt G.
dt
The generator of the Markov chain is defined as
1
G = lim (Ph I).
h0 h
Let now it = P(Xt = i). The vector t is the distribution of the Markov chain at
time t. We can study its evolution using the equation
t = 0 Pt .
Thus, as in the case if discrete time Markov chains, the evolution of a continuous
time Markov chain is completely determined by the initial distribution and and
transition matrix.
Consider now the case a continuous time Markov process with continuous state
space and with continuous paths. As we have seen in Example 4.2.3 the Brownian
motion is an example of such a process. It is a standard result in the theory of par-
tial differential equations that the conditional probability density of the Brownian
motion (4.4) is the fundamental solution of the diffusion equation:
p 1 2p
= , lim p(y, t|x, s) = (y x). (4.10)
t 2 y 2 ts
Similarly, the conditional distribution of the OU process satisfies the initial value
problem
p (yp) 1 2 p
= + , lim p(y, t|x, s) = (y x). (4.11)
t y 2 y 2 ts
The Brownian motion and the OU process are examples of a diffusion process.
A diffusion process is a continuous time Markov process with continuous paths.
We will see in Chapter 5, that the conditional probability density p(y, t|x, s) of a
diffusion process satisfies the forward Kolmogorov or Fokker-Planck equation
p 1 2
= (a(y, t)p) + (b(y, t)p), lim p(y, t|x, s) = (y x). (4.12)
t y 2 y 2 ts
Remark 4.3.6. The filtration FtX is generated by events of the form {|Xs1
B1 , Xs2 B2 , . . . Xsn Bn , } with 0 6 s1 < s2 < < sn 6 s and Bi
B(E). The definition of a Markov process is thus equivalent to the hierarchy of
equations
The particle position depends on the past of the OU process and, consequently,
is not a Markov process. However, the joint position-velocity process {Xt , Yt } is.
Its transition probability density p(x, y, t|x0 , y0 ) satisfies the forward Kolmogorov
equation
p p 1 2p
= p + (yp) + .
t x y 2 y 2
for all t, s T with t > s and all B(E). Assume that Xs = x. Since
P Xt |FsX = P [Xt |Xs ] we can write
Now we use the Markov property, together with equations (4.16) and (4.17) and
the fact that s < u FsX FuX to calculate:
I () denotes the indicator function of the set . We have also set E = R. The
CK equation is an integral equation and is the fundamental equation in the theory
of Markov processes. Under additional assumptions we will derive from it the
Fokker-Planck PDE, which is the fundamental equation in the theory of diffusion
processes, and will be the main object of study in this course.
Definition 4.4.1. A Markov process is homogeneous if
Let Xt be a homogeneous Markov process and assume that the initial distri-
bution of Xt is given by the probability measure () = P (X0 ) (for deter-
ministic initial conditionsX0 = x we have that () = I (x) ). The transition
function P (x, t, ) and the initial distribution determine the finite dimensional
distributions of X by
Theorem 4.4.2. ([12, Sec. 4.1]) Let P (t, x, ) satisfy (4.18) and assume that
(E, ) is a complete separable metric space. Then there exists a Markov process
X in E whose finite-dimensional distributions are uniquely determined by (4.19).
66 CHAPTER 4. MARKOV PROCESSES
Thus, the initial distribution and the transition function are sufficient to character-
ize a homogeneous Markov process. Notice that they do not provide us with any
information about the actual paths of the Markov process. The transition proba-
bility P (, t|x, s) is a probability measure. Assume that it has a density for all
t > s: Z
P (, t|x, s) = p(y, t|x, s) dy.
Clearly, for t = s we have P (, s|x, s) = I (x). The Chapman-Kolmogorov
equation becomes:
Z Z Z
p(y, t|x, s) dy = p(y, t|z, u)p(z, u|x, s) dzdy,
R
P0 = I, Pt+s = Pt Ps t, s > 0.
Furthermore:
Z
(Pt+s f )(x) = f (y)P (t + s, x, dy)
Z Z
= f (y)P (s, z, dy)P (t, x, dz)
Z Z
= f (y)P (s, z, dy) P (t, x, dz)
Z
= (Ps f )(z)P (t, x, dz)
= (Pt Ps f )(x).
Consequently:
Pt+s = Pt Ps .
for all f (x) Cb (E) (continuous bounded functions on E). Assume for simplicity
that Pt : Cb (E) Cb (E). Then the one-parameter family of operators Pt forms
a semigroup of operators on Cb (E). We define by D(L) the set of all f Cb (E)
such that the strong limit
Pt f f
Lf = lim ,
t0 t
exists.
Definition 4.5.2. The operator L : Cb (E) Cb (E) defined above is called the
generator of the Markov process {Xt ; t > 0}.
The semigroup property and the definition of the generator of a semigroup
imply that, formally at least, we can write:
Pt = exp(Lt).
Consider the function u(x, t) := (Pt f )(x). We calculate its time derivative:
u d d Lt
= (Pt f ) = e f
t dt dt
= L eLt f = LPt f = Lu.
Thus, given the generator of a Markov process L, we can calculate all the statistics
of our process by solving the backward Kolmogorov equation. In the case where
the Markov process is the solution of a stochastic differential equation, then the
generator is a second order elliptic operator and the backward Kolmogorov equa-
tion becomes an initial value problem for a parabolic PDE.
The space Cb (E) is natural in a probabilistic context, but other Banach spaces
often arise in applications; in particular when there is a measure on E, the spaces
Lp (E; ) sometimes arise. We will quite often use the space L2 (E; ), where
will is the invariant measure of our Markov process. The generator is fre-
quently taken as the starting point for the definition of a homogeneous Markov
process. Conversely, let Pt be a contraction semigroup (Let X be a Banach space
and T : X X a bounded operator. Then T is a contraction provided that
kT f kX 6 kf kX f X), with D(Pt ) Cb (E), closed. Then, under mild
technical hypotheses, there is an Evalued homogeneous Markov process {Xt }
associated with Pt defined through
Notice that the transition probability density t,x of the one dimensional
Brownian motion is the fundamental solution (Greens function) of the heat (diffu-
sion) PDE
u 1 2u
= .
t 2 x2
Pt = exp(L t),
Let t := Pt . This is the law of the Markov process and is the initial dis-
tribution. An argument similar to the one used in the derivation of the backward
70 CHAPTER 4. MARKOV PROCESSES
= L , (y, 0) = 0 (y). (4.23)
t
This is the forward Kolmogorov or Fokker-Planck equation. When the initial
conditions are deterministic, X0 = x, the initial condition becomes 0 = (y x).
Given the initial distribution and the generator of the Markov process Xt , we can
calculate the transition probability density by solving the Forward Kolmogorov
equation. We can then calculate all statistical quantities of this process through the
formula Z
E(f (Xt )|X0 = x) = f (y)(t, y; x) dy.
We will derive rigorously the backward and forward Kolmogorov equations for
Markov processes that are defined as solutions of stochastic differential equations
later on.
We can study the evolution of a Markov process in two different ways: Either
through the evolution of observables (Heisenberg/Koopman)
(Pt f )
= L(Pt f ),
t
or through the evolution of states (Schrodinger/Frobenious-Perron)
(Pt )
= L (Pt ).
t
We can also study Markov processes at the level of trajectories. We will do this
after we define the concept of a stochastic differential equation.
Pt g = g, g Cb (E) t > 0
If this measure is unique, then the Markov process is ergodic. Using this, we can
obtain an equation for the invariant measure in terms of the adjoint of the generator
L , which is the generator of the semigroup Pt . Indeed, from the definition of the
generator of a semigroup and the definition of an invariant measure, we conclude
that a measure is invariant if and only if
L = 0
L = 0.
The invariant measure (distribution) governs the long-time dynamics of the Markov
process.
72 CHAPTER 4. MARKOV PROCESSES
There are no solutions to Equation (4.24), subject to the constraints (4.25). 2 Thus,
the one dimensional Brownian motion is not an ergodic process.
Example 4.6.3. Consider a one-dimensional Brownian motion on [0, 1], with pe-
riodic boundary conditions. The generator of this Markov process L is the differ-
d2
ential operator L = 12 dx 2 , equipped with periodic boundary conditions on [0, 1].
This operator is self-adjoint. The null space of both L and L comprises constant
functions on [0, 1]. Both the backward Kolmogorov and the Fokker-Planck equation
reduce to the heat equation
1 2
=
t 2 x2
with periodic boundary conditions in [0, 1]. Fourier analysis shows that the solu-
tion converges to a constant at an exponential rate. See Exercise 6.
Example 4.6.4. The one dimensional Ornstein-Uhlenbeck (OU) process is a
Markov process with generator
d d2
L = x + D 2.
dx dx
2
The general solution to Equation (4.25) is (x) = Ax + B for arbitrary constants
R A and B. This
function is not normalizable, i.e. there do not exist constants A and B so that R rho(x) dx = 1.
4.7. DISCUSSION AND BIBLIOGRAPHY 73
4.8 Exercises
1. Let {Xn } be a stochastic process with state space S = Z. Show that it is a
Markov process if and only if for all n
2. Show that (4.4) is the solution of initial value problem (4.10) as well as of the
final value problem
p 1 2p
= , lim p(y, t|x, s) = (y x).
s 2 x2 st
3. Use (4.5) to show that the forward and backward Kolmogorov equations for the
OU process are
p 1 2p
= (yp) +
t y 2 y 2
and
p p 1 2 p
= x + .
s x 2 x2
4. Let W (t) be a standard one dimensional Brownian motion, let Y (t) = W (t)
with > 0 and consider the process
Z t
X(t) = Y (s) ds.
0
Show that the joint process {X(t), Y (t)} is Markovian and write down the
generator of the process.
Show that the joint process {X(t), Y (t)} is Markovian and write down the
generator of the process.
operator is self-adjoint. Show that the null space of both L and L comprises
constant functions on [0, 1]. Conclude that this process is ergodic. Solve the
corresponding Fokker-Planck equation for arbitrary initial conditions 0 (x) .
Show that the solution converges to a constant at an exponential rate. .
R(t)
E[Xt+s |Xs ] = X(s), s, t > 0.
R(0)
(c) Use the previous result to show that the only stationary Gaussian Markov
process with continuous autocorrelation function is the stationary OU pro-
cess.
Diffusion Processes
5.1 Introduction
77
78 CHAPTER 5. DIFFUSION PROCESSES
ii. (Definition of drift coefficient). There exists a function a(x, s) such that for
every x and every > 0
Z
(y x)P (dy, t|x, s) = a(x, s)(t s) + o(t s). (5.2)
|yx|6
iii. (Definition of diffusion coefficient). There exists a function b(x, s) such that
for every x and every > 0
Z
(y x)2 P (dy, t|x, s) = b(x, s)(t s) + o(t s). (5.3)
|yx|6
then we can extend the integration over the whole Rd and use expectations in the
definition of the drift and the diffusion coefficient. Indeed, ,let k = 0, 1, 2 and
notice that
Z
|y x|k P (dy, t|x, s)
|yx|>
Z
= |y x|2+ |y x|k(2+) P (dy, t|x, s)
|yx|>
Z
1
6 2+k |y x|2+ P (dy, t|x, s)
|yx|>
Z
1
6 2+k |y x|2+ P (dy, t|x, s).
R d
5.3. THE BACKWARD AND FORWARD KOLMOGOROV EQUATIONS 79
This implies that assumption (5.4) is sufficient for the sample paths to be continuous
(k = 0) and for the replacement of the truncated integrals in (8.73) and (5.3) by
integrals over R (k = 1 and k = 2, respectively). The definitions of the drift and
diffusion coefficients become:
Xt Xs
lim E Xs = x = a(x, s) (5.5)
ts ts
and
|Xt Xs |2
lim E Xs = x = b(x, s) (5.6)
ts ts
Assume furthermore that the functions a(x, s), b(x, s) are continuous in both x
and s. Then u(x, s) C 2,1 (R R+ ) and it solves the final value problem
u u 1 2u
= a(x, s) + b(x, s) 2 , lim u(s, x) = f (x). (5.7)
s x 2 x st
80 CHAPTER 5. DIFFUSION PROCESSES
Proof. First we notice that, the continuity assumption (5.1), together with the fact
that the function f (x) is bounded imply that
Z
u(x, s) = f (y) P (dy, t|x, s)
ZR Z
= f (y)P (dy, t|x, s) + f (y)P (dy, t|x, s)
|yx|6 |yx|>
Z Z
6 f (y)P (dy, t|x, s) + kf kL P (dy, t|x, s)
|yx|6 |yx|>
Z
= f (y)P (dy, t|x, s) + o(t s).
|yx|6
We add and subtract the final condition f (x) and use the previous calculation to
obtain:
Z Z
u(x, s) = f (y)P (dy, t|x, s) = f (x) + (f (y) f (x))P (dy, t|x, s)
R Z R Z
= f (x) + (f (y) f (x))P (dy, t|x, s) + (f (y) f (x))P (dy, t|x, s)
|yx|6 |yx|>
Z
= f (x) + (f (y) f (x))P (dy, t|x, s) + o(t s).
|yx|6
Now the final condition follows from the fact that f (x) Cb (R) and the arbitrari-
ness of .
Now we show that u(s, x) solves the backward Kolmogorov equation. We use
the Chapman-Kolmogorov equation (4.15) to obtain
Z
u(x, ) = f (z)P (dz, t|x, ) (5.8)
R
Z Z
= f (z)P (dz, t|y, )P (dy, |x, )
ZR R
Assume now that the transition function has a density p(y, t|x, s). In this case
the formula for u(x, s) becomes
Z
u(x, s) = f (y)p(y, t|x, s) dy.
R
where
1 2
As,x := a(x, s) + b(x, s) 2 .
x 2 x
Since (5.11) is valid for arbitrary functions f (y), we obtain a partial differential
equations for the transition probability density:
Theorem 5.3.2. (Kolmogorov) Assume that conditions (5.1), (8.73), (5.3) are sat-
isfied and that p(y, t|, ), a(y, t), b(y, t) C 2,1 (R R+ ). Then the transition
probability density satisfies the equation
p 1 2
= (a(t, y)p) + (b(t, y)p) , lim p(t, y|x, s) = (x y). (5.13)
t y 2 y 2 ts
Proof. Fix a function f (y) C02 (R). An argument similar to the one used in the
proof of the backward Kolmogorov equation gives
Z
1 1
lim f (y)p(y, s + h|x, s) ds f (x) = a(x, s)fx (x) + b(x, s)fxx (x),
h0 h 2
(5.14)
where subscripts denote differentiation with respect to x. On the other hand
Z Z
f (y) p(y, t|x, s) dy = f (y)p(y, t|x, s) dy
t t
Z
1
= lim (p(y, t + h|x, s) p(y, t|x, s)) f (y) dy
h0 h
Z Z
1
= lim p(y, t + h|x, s)f (y) dy p(z, t|s, x)f (z) dz
h0 h
Z Z Z
1
= lim p(y, t + s|z, t)p(z, t|x, s)f (y) dydz p(z, t|s, x)f (z
h0 h
Z Z
1
= lim p(z, t|x, s) p(y, t + h|z, t)f (y) dy f (z) dz
h0 h
Z
1
= p(z, t|x, s) a(z, t)fz (z) + b(z)fzz (z) dz
2
Z
1 2
= (a(z)p(z, t|x, s)) + (b(z)p(z, t|x, s) f (z) dz.
z 2 z 2
In the above calculation used the Chapman-Kolmogorov equation. We have also
performed two integrations by parts and used the fact that, since the test function f
has compact support, the boundary terms vanish.
5.3. THE BACKWARD AND FORWARD KOLMOGOROV EQUATIONS 83
Since the above equation is valid for every test function f (y), the forward
Kolmogorov equation follows.
Assume now that initial distribution of Xt is 0 (x) and set s = 0 (the initial
time) in (5.13). Define
Z
p(y, t) := p(y, t|x, 0)0 (x) dx. (5.15)
We multiply the forward Kolmogorov equation (5.13) by 0 (x) and integrate with
respect to x to obtain the equation
p(y, t) 1 2
= (a(y, t)p(y, t)) + (b(y, t)p(t, y)) , (5.16)
t y 2 y 2
together with the initial condition
The solution of equation (5.16), provides us with the probability that the diffusion
process Xt , which initially was distributed according to the probability density
0 (x), is equal to y at time t. Alternatively, we can think of the solution to (5.13)
as the Greens function for the PDE (5.16). Using (5.16) we can calculate the
expectation of an arbitrary function of the diffusion process Xt :
Z Z
E(f (Xt )) = f (y)p(y, t|x, 0)p(x, 0) dxdy
Z
= f (y)p(y, t) dy,
where p(y, t) is the solution of (5.16). Quite often we need to calculate joint prob-
ability densities. For, example the probability that Xt1 = x1 and Xt2 = x2 . From
the properties of conditional expectation we have that
Using the joint probability density we can calculate the statistics of a function of
the diffusion process Xt at times t and s:
Z Z
E(f (Xt , Xs )) = f (y, x)p(y, t|x, s)p(x, s) dxdy. (5.18)
84 CHAPTER 5. DIFFUSION PROCESSES
In particular, Z Z
E(Xt X0 ) = yxp(y, t|x, 0)p(x, 0) dxdy.
and Z
1
lim (y x) (y x)P (dy, t|x, s) = b(x, s).
ts t s |yx|<
The drift coefficient a(x, s) is a d-dimensional vector field and the diffusion coef-
ficient b(x, s) is a d d symmetric matrix (second order tensor). The generator of
a d dimensional diffusion process is
1
L = a(x, s) + b(x, s) :
2
Xd d
1 X 2
= aj (x, s) + bij (x, s) 2 .
xj 2 xj
j=1 i,j=1
Exercise 5.4.1. Derive rigorously the forward and backward Kolmogorov equa-
tions in arbitrary dimensions.
Assuming that the first and second moments of the multidimensional diffusion
process exist, we can write the formulas for the drift vector and diffusion matrix as
Xt Xs
lim E Xs = x = a(x, s) (5.19)
ts ts
and
(Xt Xs ) (Xt Xs )
lim E Xs = x = b(x, s) (5.20)
ts ts
Notice that from the above definition it follows that the diffusion matrix is sym-
metric and nonnegative definite.
5.5. CONNECTION WITH STOCHASTIC DIFFERENTIAL EQUATIONS 85
dXt = dWt , X0 = x.
Xt = x + Wt .
d D d2
L = x + .
dx 2 dx2
Z t (ts)
Xt = et X0 + D e dWs .
0
The argument used in the derivation of the forward and backward Kolmogorov
equations goes back to Kolmogorovs original work. More material on diffusion
processes can be found in [26], [32].
5.8. EXERCISES 87
5.8 Exercises
1. Prove equation (5.14).
6.1 Introduction
In the previous chapter we derived the backward and forward (Fokker-Planck) Kol-
mogorov equations and we showed that all statistical properties of a diffusion pro-
cess can be calculated from the solution of the Fokker-Planck equation. 1 In this
long chapter we study various properties of this equation such as existence and
uniqueness of solutions, long time asymptotics, boundary conditions and spectral
properties of the Fokker-Planck operator. We also study in some detail various ex-
amples of diffusion processes and of the associated Fokker-Palnck equation. We
will restrict attention to time-homogeneous diffusion processes, for which the drift
and diffusion coefficients do not depend on time.
In Section 6.2 we study various basic properties of the Fokker-Planck equa-
tion, including existence and uniqueness of solutions, writing the equation as a
conservation law and boundary conditions. In Section 6.3 we present some exam-
ples of diffusion processes and use the corresponding Fokker-Planck equation in
order to calculate various quantities of interest such as moments. In Section 6.4 we
study the multidimensional Onrstein-Uhlenbeck process and we study the spectral
properties of the corresponding Fokker-Planck operator. In Section 6.5 we study
stochastic processes whose drift is given by the gradient of a scalar function, gra-
dient flows. In Section 6.7 we solve the Fokker-Planck equation for a gradient
SDE using eigenfunction expansions and we show how the eigenvalue problem
for the Fokker-Planck operator can be reduced to the eigenfunction expansion for
1
In this chapter we will call the equation Fokker-Planck, which is more customary in the physics
literature. rather forward Kolmogorov, which is more customary in the mathematics literature.
89
90 CHAPTER 6. THE FOKKER-PLANCK EQUATION
a Schrodinger operator. In Section 8.2 we study the Langevin equation and the
associated Fokker-Planck equation. In Section 8.3 we calculate the eigenvalues
and eigenfunctions of the Fokker-Planck operator for the Langevin equation in a
harmonic potential. Discussion and bibliographical remarks are included in Sec-
tion 6.8. Exercises can be found in Section 6.9.
By definition (see equation (5.20)), the diffusion matrix is always symmetric and
nonnegative. We will assume that it is actually uniformly positive definite, i.e. we
will impose the uniform ellipticity condition:
d
X
bij (x)i j > kk2 , Rd , (6.3)
i,j=1
6.2. BASIC PROPERTIES OF THE FP EQUATION 91
Definition 6.2.1. We will call a solution to the Cauchy problem for the Fokker
Planck equation (6.2) a classical solution if:
i. u C 2,1 (Rd , R+ ).
2
ku(t, x)kL (0,T ) 6 cekxk
Theorem 6.2.2. Assume that conditions (6.3) and (6.4) are satisfied, and assume
2
that |f | 6 cekxk . Then there exists a unique classical solution to the Cauchy
problem for the FokkerPlanck equation. Furthermore, there exist positive con-
stants K, so that
1
|p|, |pt |, kpk, kD 2 pk 6 Kt(n+2)/2 exp kxk2 . (6.5)
2t
Notice that from estimates (6.5) it follows that all moments of a uniformly
elliptic diffusion process exist. In particular, we can multiply the Fokker-Planck
equation by monomials xn and then to integrate over Rd and to integrate by parts.
No boundary terms will appear, in view of the estimate (6.5).
Remark 6.2.3. The solution of the Fokker-Planck equation is nonnegative for all
times, provided that the initial distribution is nonnegative. This is follows from the
maximum principle for parabolic PDEs.
92 CHAPTER 6. THE FOKKER-PLANCK EQUATION
d
1X
Ji := ai (x)p bij (x)p . (6.6)
2 xj
j=1
Consequently:
kp(, t)kL1 (Rd ) = kp(, 0)kL1 (Rd ) = 1. (6.7)
Hence, the total probability is conserved, as expected. Equation (6.7) simply means
that
E(Xt Rd ) = 1, t > 0.
When studying a diffusion process that can take values on the whole of Rd , then
we study the pure initial value (Cauchy) problem for the Fokker-Planck equation,
equation (6.1). The boundary condition was that the solution decays sufficiently
fast at infinity. For ergodic diffusion processes this is equivalent to requiring that
the solution of the backward Kolmogorov equation is an element of L2 () where
is the invariant measure of the process. There are many applications where it is
important to study stochastic process in bounded domains. In this case it is neces-
sary to specify the value of the stochastic process (or equivalently of the solution
to the Fokker-Planck equation) on the boundary.
6.2. BASIC PROPERTIES OF THE FP EQUATION 93
p(x, t) = 0, on .
n J(x, t) = 0, on .
iii. The transition probability density is a periodic function in the case of peri-
odic boundary conditions.
Notice that, using the terminology customary to PDEs theory, absorbing boundary
conditions correspond to Dirichlet boundary conditions and reflecting boundary
conditions correspond to Neumann. Of course, on consider more complicated,
mixed boundary conditions.
2
Of course, the random walk is not a diffusion process. However, as we have already seen the
Brownian motion can be defined as the limit of an appropriately rescaled random walk. A similar
construction exists for more general diffusion processes.
94 CHAPTER 6. THE FOKKER-PLANCK EQUATION
Consider now a diffusion process in one dimension on the interval [0, L]. The
boundary conditions are
p(0, t) = J(L, t) = 0.
Set a(y, t) 0, b(y, t) 2D > 0. This diffusion process is the Brownian motion
with diffusion coefficient D. Let us calculate the transition probability density of
this process assuming that the Brownian particle is at y at time s. The Fokker-
Planck equation for the transition probability density p(x, t|y, s) is:
p 2p
= D 2, p(x, s|y, s) = (x y). (6.8)
t x
The solution to this equation is the Greens function (fundamental solution) of the
heat equation:
1 (x y)2
p(x, t|y, s) = p exp . (6.9)
4D(t s) 4D(t s)
Notice that using the Fokker-Planck equation for the Brownian motion we can
immediately show that the mean squared displacement grows linearly in time. As-
6.3. EXAMPLES OF DIFFUSION PROCESSES 95
where we performed two integrations by parts and we used the fact that, in view
of (6.9), no boundary terms remain. From this calculation we conclude that
EWt2 = 2Dt.
Assume now that the initial condition W0 of the Brownian particle is a random
variable with distribution 0 (x). To calculate the probability density function (dis-
tribution function) of the Brownian particle we need to solve the Fokker-Planck
equation with initial condition 0 (x). In other words, we need to take the aver-
age of the probability density function p(x, t|y, 0) over all initial realizations of
the Brownian particle. The solution of the Fokker-Planck equation, the distribution
function, is Z
p(x, t) = p(x, t|y, 0)0 (y) dy. (6.10)
Notice that only the transition probability density depends on x and y only through
their difference. Thus, we can write p(x, t|y, 0) = p(x y, t). From (6.10) we see
that the distribution function is given by the convolution between the transition
probability density and the initial condition, as we know from the theory of partial
differential equations.
Z
p(x, t) = p(x y, t)0 (y) dy =: p 0 .
We can also consider Brownian motion in a bounded domain, with either absorb-
ing, reflecting or periodic boundary conditions. Set D = 1 and consider the
Fokker-Planck equation (6.8) on [0, 1] with absorbing boundary conditions:
p 1 2p
= , p(0, t) = p(1, t) = 0. (6.11)
t 2 x2
96 CHAPTER 6. THE FOKKER-PLANCK EQUATION
Notice that the boundary conditions are automatically satisfied. The initial condi-
tion is
p(x, 0) = (x x0 ),
We substitute the expansion (6.12) into (6.11) and use the orthogonality properties
of the Fourier basis to obtain the equations
n2 2
pn = pn n = 1, 2, . . .
2
The solution of this equation is
n2 2
pn (t) = pn (0)e 2
t
.
Consequently, the transition probability density for the Brownian motion on [0, 1]
with absorbing boundary conditions is
X n2 2
p(x, t|x0 , 0) = 2 e 2
t
sin nx0 sin(nx).
n=1
Notice that
lim p(x, t|x0 , 0) = 0.
t
This is not surprising, since all Brownian particles will eventually get absorbed at
the boundary.
Consider now Brownian motion on the interval [0, 1] with reflecting boundary con-
ditions and set D = 1 for simplicity. In order to calculate the transition probability
6.3. EXAMPLES OF DIFFUSION PROCESSES 97
density we have to solve the Fokker-Planck equation which is the heat equation on
[0, 1] with Neumann boundary conditions:
p 1 2p
= , x p(0, t) = x p(1, t) = 0, p(x, 0) = (x x0 ).
t 2 x2
The boundary conditions are satisfied by functions of the form cos(nx). We look
for a solution in the form of a cosine Fourier series
X
1
p(x, t) = a0 + an (t) cos(nx).
2 n=1
We substitute the expansion into the PDE and use the orthonormality of the Fourier
basis to obtain the equations for the Fourier coefficients:
n2 2
a n = an
2
from which we deduce that
n2 2
an (t) = an (0)e 2
t
.
Consequently
X n2 2
p(x, t|x0 , 0) = 1 + 2 cos(nx0 ) cos(nx)e 2
t
.
n=1
2 ps
= 0, x ps (0) = x ps (1) = 0.
x2
The unique normalized solution to this boundary value problem is ps (x) = 1.
Indeed, we multiply the equation by ps , integrate by parts and use the boundary
conditions to obtain Z 1
dps 2
dx dx = 0,
0
98 CHAPTER 6. THE FOKKER-PLANCK EQUATION
from which it follows that ps (x) = 1. Alternatively, by taking the limit of p(x, t|x0 , 0)
as t we obtain the invariant distribution:
p (xp) 2p
= +D 2. (6.13)
t x x
This is the Fokker-Planck equation for the Ornstein-Uhlenbeck process. The cor-
responding stochastic differential equation is
dXt = Xt + 2DdWt .
So, in addition to Brownian motion there is a linear force pulling the particle to-
wards the origin. We know that Brownian motion is not a stationary process, since
the variance grows linearly in time. By adding a linear damping term, it is rea-
sonable to expect that the resulting process can be stationary. As we have already
seen, this is indeed the case.
The transition probability density pOU (x, t|y, s) for an OU particle that is lo-
cated at y at time s is
r !
(x e(ts) y)2
pOU (y, t|x, s) = exp . (6.14)
2D(1 e2(ts) ) 2D(1 e2(ts) )
We obtained this formula in Example (4.2.4) (for = D = 1) by using the fact that
the OU process can be defined through the a time change of the Brownian motion.
6.3. EXAMPLES OF DIFFUSION PROCESSES 99
We can also derive it by solving equation (6.13). To obtain (6.14), we first take
the Fourier transform of the transition probability density with respect to x, solve
the resulting first order PDE using the method of characteristics and then take the
inverse Fourier transform3
Notice that from formula (6.14) it immediately follows that in the limit as the
friction coefficient goes to 0, the transition probability of the OU processes con-
verges to the transition probability of Brownian motion. Furthermore, by taking
the long time limit in (6.14) we obtain (we have set s = 0)
r
x2
lim pOU (x, t|y, 0) = exp ,
t+ 2D 2D
irrespective of the initial position y of the OU particle. This is to be expected, since
as we have already seen the Ornstein-Uhlenbeck process is an ergodic Markov
process, with a Gaussian invariant distribution
r
x2
ps (x) = exp . (6.15)
2D 2D
Using now (6.14) and (6.15) we obtain the stationary joint probability density
where p(x y, t) := p(x, t|y, 0). When the OU process is distributed initially
according to its invariant distribution, 0 (x) = ps (x) given by (6.15), then the
Ornstein-Uhlenbeck process becomes stationary. The distribution function is given
by ps (x) at all times and the joint probability density is given by (6.16).
Knowledge of the distribution function enables us to calculate all moments of
the OU process using the formula
Z
n
E((Xt ) ) = xn p(x, t) dx,
We will calculate the moments by using the Fokker-Planck equation, rather than
the explicit formula for the transition probability density. Let Mn (t) denote the nth
moment of the OU process,
Z
Mn := xn p(x, t) dx, n = 0, 1, 2, . . . ,
R
multiply the FP equation for the OU process by x, integrate over R and perform
and integration by parts to obtain:
d
M1 = M1 .
dt
Consequently, the first moment converges exponentially fast to 0:
M1 (t) = et M1 (0).
Let now n > 2. We multiply the FP equation for the OU process by xn and
integrate by parts (once on the first term on the RHS and twice on the second) to
obtain: Z Z Z
d
y p = n y p + Dn(n 1) y n2 p.
n n
dt
Or, equivalently:
d
Mn = nMn + Dn(n 1)Mn2 , n > 2.
dt
This is a first order linear inhomogeneous differential equation. We can solve it
using the variation of constants formula:
Z t
nt
Mn (t) = e Mn (0) + Dn(n 1) en(ts) Mn2 (s) ds. (6.20)
0
We can use this formula, together with the formulas for the first two moments in
order to calculate all higher order moments in an iterative way. For example, for
n = 2 we have
Z t
M2 (t) = e2t M2 (0) + 2D e2(ts) M0 (s) ds
0
2t D
= e M2 (0) + e2t (e2t 1)
D D
= + e2t M2 (0) .
Consequently, the second moment converges exponentially fast to its stationary
D
value 2 . The stationary moments of the OU process are:
r Z
n y 2
hy iOU := y n e 2D dx
2D R
n
D n/2
= 1.3 . . . (n 1) , n even,
0, n odd.
102 CHAPTER 6. THE FOKKER-PLANCK EQUATION
exponentially fast4 . Since we have already shown that the distribution function of
the OU process converges to the Gaussian distribution in the limit as t +, it
is not surprising that the moments also converge to the moments of the invariant
Gaussian measure. What is not so obvious is that the convergence is exponentially
fast. In the next section we will prove that the Ornstein-Uhlenbeck process does,
indeed, converge to equilibrium exponentially fast. Of course, if the initial condi-
tions of the OU process are stationary, then the moments of the OU process become
independent of time and given by their equilibrium values
dXt = Xt dt + Xt dWt .
This equation is one of the basic models in mathematical finance. The coefficient
is called the volatility. The generator of this process is
x2 2
L = x + .
x 2 x2
Notice that this operator is not uniformly elliptic. The Fokker-Planck equation of
the geometric Brownian motion is:
p 2 2 x2
= (x) + 2 p .
t x x 2
We can easily obtain an equation for the nth moment of the geometric Brownian
motion:
d 2
Mn = n + n(n 1) Mn , n > 2.
dt 2
4
Of course, we need to assume that the initial distribution has finite moments of all orders in order
to justify the above calculations.
6.4. THE ORNSTEIN-UHLENBECK PROCESS AND HERMITE POLYNOMIALS103
2
Mn (t) = e(+(n1) 2
)nt
Mn (0), n>2
and
M1 (t) = et M1 (0).
Notice that the nth moment might diverge as t , depending on the values of
and . Consider for example the second moment and assume that < 0. We have
2 )t
Mn (t) = e(2+ M2 (0),
where denotes the inverse temperature. We have already seen that the OU pro-
cess is an ergodic Markov process whose unique invariant measure is absolutely
continuous with respect to the Lebesgue measure on Rd with Gaussian density
C (Rd )
1 |p|2
2
(p) = e .
(2 1 )d/2
104 CHAPTER 6. THE FOKKER-PLANCK EQUATION
The natural function space for studying the generator of the OU process is the L2 -
space weighted by the invariant measure of the process. This is a separable Hilbert
space with norm Z
kf k2 := f 2 dp.
Rd
and corresponding inner product
Z
(f, h) = f h dp.
R
= 1 (f, h) .
(Lf, f ) = 1 kf k2 6 0.
6.4. THE ORNSTEIN-UHLENBECK PROCESS AND HERMITE POLYNOMIALS105
(Lf, f ) = 1 kf k2
6 kf k2
Theorem 6.4.2. Consider the eigenvalue problem for the generator of the OU
process in one dimension
Lfn = n fn . (6.27)
Then the eigenvalues of L are the nonnegative integers:
n = n, n = 0, 1, 2, . . . .
2
H(p; ) = ep 2 , p R.
Then
X n
H(p; ) = Hn (p), p R, (6.30)
n!
n=0
The first few Hermite polynomials and the corresponding rescaled/normalized eigen-
functions of the generator of the OU process are:
H0 (p) = 1, f0 (p) = 1,
p
H1 (p) = p, f1 (p) = p,
1
H2 (p) = p2 1, f2 (p) = p2 ,
2 2
3/2
3
H3 (p) = p3 3p, f3 (p) = 3
p p
6 6
1
H4 (p) = p4 3p2 + 3, f4 (p) = 2 p4 3p2 + 3
24
1 5/2 5
H5 (p) = p5 10p3 + 15p, f5 (p) = p 10 3/2 p3 + 15 1/2 p .
120
6.4. THE ORNSTEIN-UHLENBECK PROCESS AND HERMITE POLYNOMIALS107
The proof of Theorem 6.4.2 follows essentially from the properties of the Hermite
polynomials. First, notice that by combining (6.28) and (6.30) we obtain
p +
X n
H( p, ) = fn (p)
n=0 n!
p p +
X n
H( p, ) = p fn (p),
n=1
n!
a+ = p + p.
L = a+ a .
[a+ , a ] = 1
1
S+ = p a+
(n + 1)
and
1
S = a .
n
Then
S + fn = fn+1 and S fn = fn1 . (6.34)
In particular,
1
fn = (a+ )n 1 (6.35)
n!
and
1
1 = (a )n fn . (6.36)
n!
6.4. THE ORNSTEIN-UHLENBECK PROCESS AND HERMITE POLYNOMIALS109
Now,
a+ a = (p + p)p = p pp = L.
Similarly,
a a+ = p2 + pp + 1.
and
[a+ , a ] = 1
Forumlas (6.34) follow from (6.31) and (6.33). Finally, formulas (6.35) and (6.36)
are a consequence of (6.31) and (6.33), together with a simple induction argument.
Notice that upon using (6.35) and (6.36) and the fact that a+ is the adjoint of
a we can easily check the orthonormality of the eigenfunctions:
Z Z
1
fn fm = fn (a )m 1
m! Z
1
= (a )m fn
Z m!
= fnm = nm .
From the eigenfunctions and eigenvalues of L we can easily obtain the eigenvalues
and eigenfunctions of L , the Fokker-Planck operator.
L = p2 +p (p)
are
n = n, n = 0, 1, 2, . . . and fn = fn .
110 CHAPTER 6. THE FOKKER-PLANCK EQUATION
Proof. We have
L (fn ) = fn L + Lfn
= nfn .
An immediate corollary of the above calculation is that we can the nth eigen-
function of the Fokker-Planck operator is given by
1 + n
fn = (p) (a ) 1.
n!
In this section we study a more general class (in fact, as we will see later the
most general class) of reversible Markov processes, namely stochastic perturba-
tions of ODEs with a gradient structure.
Let V (x) = 21 x2 . The generator of the OU process can be written as:
L = x V x + 1 x2 .
L = V (x) + 1 (6.40)
p
= (V p) + 1 p. (6.42)
t
It is not possible to calculate the time dependent solution of this equation for an
arbitrary potential. We can, however, always calculate the stationary solution, if it
exists.
for all R+ .
Proposition 6.5.3. Let V (x) be a smooth confining potential. Then the Markov
process with generator (6.40) is ergodic. The unique invariant distribution is the
Gibbs distribution
1
p(x) = eV (x) (6.44)
Z
where the normalization factor Z is the partition function
Z
Z= eV (x) dx.
Rd
The fact that the Gibbs distribution is an invariant distribution follows by direct
substitution. Uniqueness follows from a PDEs argument (see discussion below). It
is more convenient to normalize the solution of the Fokker-Planck equation with
respect to the invariant distribution.
Theorem 6.5.4. Let p(x, t) be the solution of the Fokker-Planck equation (6.42),
assume that (6.43) holds and let (x) be the Gibbs distribution (10.11). Define
h(x, t) through
p(x, t) = h(x, t)(x).
h
= V h + 1 h, h(x, 0) = p(x, 0)1 (x). (6.45)
t
112 CHAPTER 6. THE FOKKER-PLANCK EQUATION
Proof. The initial condition follows from the definition of h. We calculate the
gradient and Laplacian of p:
p = h hV
and
p = h 2V h + hV + h|V |2 2 .
We substitute these formulas into the FP equation to obtain
h
= V h + 1 h ,
t
from which the claim follows.
where (x) is the Gibbs distribution. This is a Hilbert space with inner product
Z
(f, h) = f h(x) dx.
Rd
Theorem 6.5.5. Assume that V (x) is a smooth potential and assume that condi-
tion (6.43) holds. Then the operator
L = V (x) + 1
= 1 f h dx,
Rd
(Lf, f ) = 1 kf k2 ,
0 = 1 kf k2 ,
Remark 6.5.6. The expression (Lf, f ) is called the Dirichlet form of the op-
erator L. In the case of a gradient flow, it takes the form
(Lf, f ) = 1 kf k2 . (6.46)
Using the properties of the generator L we can show that the solution of the
Fokker-Planck equation converges to the Gibbs distribution exponentially fast. For
this we need to use the fact that, under appropriate assumptions on the potential V ,
the Gibbs measure (dx) = Z 1 eV (x) satisfies Poincar`es inequality:
Theorem 6.5.7. Assume that the potential V satisfies the convexity condition
D 2 V > I.
Then the corresponding Gibbs measure satisfies the Poincare inequality with con-
stant : Z
f = 0 kf k > kf k . (6.47)
Rd
Theorem 6.5.8. Assume that p(x, 0) L2 (eV ). Then the solution p(x, t) of the
Fokker-Planck equation (6.42) converges to the Gibbs distribution exponentially
fast:
k1 2 k2L1 6 CH(1 |2 )
Theorem 6.5.10. Let p denote the solution of the FokkerPlanck equation (6.42)
where the potential is smooth and uniformly convex. Assume that the the initial
conditions satisfy
H(p(, 0)|V ) < .
Then p converges to the Gibbs distribution exponentially fast in relative entropy:
1 t
H(p(, t)|V ) 6 e H(p(, 0)|V ).
L = b(x) + 1
and invariant measure . Then the following three statements are equivalent.
b(x) = V (x).
kt kL1 6 et k0 kL1 .
1
The exponent is related to the spectral gap of the generator L = (x) (x)
+ . This technique for sampling from a given distribution is an example of the
Markov Chain Monte Carlo (MCMC) methodology.
Proposition 6.6.1. Let V (x) be a confining potential, (x) a smooth vector field
and consider the diffusion process
p
dXt = (V (Xt ) + (x)) dt + 2 1 dWt . (6.50)
116 CHAPTER 6. THE FOKKER-PLANCK EQUATION
Then the invariant measure of the process Xt is the Gibbs measure (dx) =
1 V (x)
Ze dx if and only if (x) is divergence-free with respect to the density of
this measure:
(x)eV (x)) = 0. (6.51)
where is the Poincare constant of the potential V (i.e. for the Gibbs measure
Z 1 eV (x) dx). The above imply that we can study the spectral problem for L:
Lfn = n fn , n = 0, 1, . . .
X
= n fn , n = (, fn ) (6.53)
n=0
We multiply this equation by fm , integrate wrt the Gibbs measure and use the
orthonormality of the eigenfunctions to obtain the sequence of equations
h n = n hn , n = 0, 1,
The solution is
h0 (t) = 0 , hn (t) = en t n , n = 1, 2, . . .
Notice that
Z Z
1 = p(x, 0) dx = p(x, t) dx
d Rd
ZR
= h(x, t)Z 1 eV dx = (h, 1) = (, 1)
Rd
= 0 .
This expansion, together with the fact that all eigenvalues are positive (n > 1),
shows that the solution of the backward Kolmogorov equation converges to 1 ex-
ponentially fast. The solution of the FokkerPlanck equation is
!
X
1 V (x) n t
p(x, t) = Z e 1+ e n fn .
n=1
Define now (x, t) = eV /2D p(x, t). Then solves the PDE
|V |2 V
= D U (x), U (x) := . (6.57)
t 4D 2
Let H := D + U . Then L and H have the same eigenvalues. The nth eigen-
function n of L and the nth eigenfunction n of H are associated through the
transformation
V (x)
n (x) = n (x) exp .
2D
Remarks 6.7.2. i. From equation (6.56) shows that the FP operator can be
written in the form
L = D eV /D eV /D .
ii. The operator that appears on the right hand side of eqn. (6.57) has the form
of a Schrodinger operator:
H = D + U (x).
iii. The spectral problem for the FP operator can be transformed into the spec-
tral problem for a Schrodinger operator. We can thus use all the available
results from quantum mechanics to study the FP equation and the associated
SDE.
From this we conclude that eV /2D Hn = n n eV /2D from which the equiva-
lence between the two eigenvalue problems follows.
U U
H = DA A, A=+ , A = + .
2D 2D
ii. These are creation and annihilation operators. They can also be written in
the form
A = eU/2D eU/2D , A = eU/2D eU/2D
iii. The forward the backward Kolmogorov operators have the same eigenvalues.
Their eigenfunctions are related through
B F
n = n exp (V /D) ,
where B F
n and n denote the eigenfunctions of the backward and forward
operators, respectively.
p
= p in Rd (0, T )
t
p(x, 0) = 0.
Each of these solutions besides the trivial solution p = 0 grows very rapidly as
x +. More details can be found in [34, Ch. 7].
120 CHAPTER 6. THE FOKKER-PLANCK EQUATION
6.9 Exercises
1. Solve equation (6.13) by taking the Fourier transform, using the method of char-
acteristics for first order PDEs and taking the inverse Fourier transform.
2. Use the formula for the stationary joint probability density of the Ornstein-
Uhlenbeck process, eqn. (6.17) to obtain the stationary autocorrelation function
of the OU process.
3. Use (6.20) to obtain formulas for the moments of the OU process. Prove, using
these formulas, that the moments of the OU process converge to their equilib-
rium values exponentially fast.
(a) Write down the generator and the forward and backward Kolmogorov
equations for Xt .
(b) Assume that X0 is a random variable with probability density 0 (x) that
has finite moments. Use the forward Kolmogorov equation to derive a
system of differential equations for the moments of Xt .
(c) Find the first three moments M0 , M1 , M2 in terms of the moments of the
initial distribution 0 (x).
(d) Under what conditions on the coefficients ai , bi > 0, i = 0, 1, 2 is M2
finite for all times?
(a) Write down the generator and the forward and backward Kolmogorov
equations for Xt .
(b) Assume that a(x) is divergence-free ( a(x) = 0). Show that Xt is
ergodic and find the invariant distribution.
(c) Show that the probability density p(x, t) (the solution of the forward Kol-
mogorov equation) converges to the invariant distribution exponentially
fast in L2 ([0, 1]d ). (Hint: Use Poincares inequality on [0, 1]d ).
(a) Write down the generator the forward and backward Kolmogorov equa-
tions for Xt .
(b) Show that this process is ergodic and find its invariant distribution.
122 CHAPTER 6. THE FOKKER-PLANCK EQUATION
9. Let x(t) = {x(t), y(t)} be the two-dimensional diffusion process on [0, 2]2
with periodic boundary conditions with drift vector a(x, y) = (sin(y), sin(x))
and diffusion matrix b(x, y) with b11 = b22 = 1, b12 = b21 = 0.
(a) Write down the generator of the process {x(t), y(t)} and the forward and
backward Kolmogorov equations.
(b) Show that the constant function
s (x, y) = C
is the unique stationary distribution of the process {x(t), y(t)} and calcu-
late the normalization constant.
(c) Let E denote the expectation with respect to the invariant distribution
s (x, y). Calculate
E cos(x) + cos(y) and E(sin(x) sin(y)).
10. Let a, D be positive constants and let X(t) be the diffusion process on [0, 1]
with periodic boundary conditions and with drift and diffusion coefficients a(x) =
a and b(x) = 2D, respectively. Assume that the process starts at x0 , X(0) =
x0 .
(a) Write down the generator of the process X(t) and the forward and back-
ward Kolmogorov equations.
(b) Solve the initial/boundary value problem for the forward Kolmogorov
equation to calculate the transition probability density p(x, t|x0 , 0).
(c) Show that the process is ergodic and calculate the invariant distribution
ps (x).
(d) Calculate the stationary autocorrelation function
Z 1Z 1
E(X(t)X(0)) = xx0 p(x, t|x0 , 0)ps (x0 ) dxdx0 .
0 0
Chapter 7
7.1 Introduction
In this part of the course we will study stochastic differential equation (SDEs):
ODEs driven by Gaussian white noise.
Let W (t) denote a standard mdimensional Brownian motion, h : Z Rd
a smooth vector-valued function and : Z Rdm a smooth matrix valued
function (in this course we will take Z = Td , Rd or Rl Tdl . Consider the SDE
dz dW
= h(z) + (z) , z(0) = z0 . (7.1)
dt dt
Z t Z t
z(t) = z0 + h(z(s))ds + (z(s))dW (s). (7.2)
0 0
In order to make sense of this equation we need to define the stochastic integral
against W (s).
123
124 CHAPTER 7. STOCHASTIC DIFFERENTIAL EQUATIONS
where W (t) is a standard one dimensional Brownian motion. This is not straight-
forward because W (t) does not have bounded variation. In order to define the
stochastic integral we assume that f (t) is a random process, adapted to the filtra-
tion Ft generated by the process W (t), and such that
Z T
E f (s)2 ds < .
0
The Ito stochastic integral I(t) is defined as the L2 limit of the Riemann sum
approximation of (7.3):
K1
X
I(t) := lim f (tk1 ) (W (tk ) W (tk1 )) , (7.4)
K
k=1
and
E[I(t)|Fs ] = I(s) t > s,
K1
X 1
Istrat (t) := lim f (tk1 ) + f (tk ) (W (tk ) W (tk1 )) , (7.6)
K 2
k=1
The limit in (7.6) gives rise to an integral which differs from the Ito integral. The
situation is more complex than that arising in the standard theory of Riemann in-
tegration for functions of bounded variation: in that case the points in [tk1 , tk ]
where the integrand is evaluated do not effect the definition of the integral, via a
limiting process. In the case of integration against Brownian motion, which does
not have bounded variation, the limits differ. When f and W are correlated through
an SDE, then a formula exists to convert between them.
126 CHAPTER 7. STOCHASTIC DIFFERENTIAL EQUATIONS
It is well known that existence and uniqueness of solutions for ODEs (i.e. when
0 in (7.1)) holds for globally Lipschitz vector fields h(x). A very similar
theorem holds when 6= 0. As for ODEs the conditions can be weakened, when a
priori bounds on the solution can be found.
Theorem 7.3.2. Assume that both h() and () are globally Lipschitz on Z and
that z0 is a random variable independent of the Brownian motion W (t) with
E|z0 |2 < .
Then the SDE (7.1) has a unique solution z(t) C(R+ ; Z) with
Z T
2
E |z(t)| dt < T < .
0
dz dW
= h(z) + (z) , z(0) = z0 . (7.7)
dt dt
By this we mean that z C(R+ , Z) satisfies the integral equation
Z t Z t
z(t) = z(0) + h(z(s))ds + (z(s)) dW (s). (7.8)
0 0
7.3. STOCHASTIC DIFFERENTIAL EQUATIONS 127
By using definitions (7.4) and (7.6) it can be shown that z satisfying the Stratonovich
SDE (7.7) also satisfies the Ito SDE
dz 1 1 dW
= h(z) + (z)(z)T (z) (z)T + (z) , (7.9a)
dt 2 2 dt
z(0) = z0 , (7.9b)
provided that (z) is differentiable. White noise is, in most applications, an ideal-
ization of a stationary random process with short correlation time. In this context
the Stratonovich interpretation of an SDE is particularly important because it often
arises as the limit obtained by using smooth approximations to white noise. On
the other hand the martingale machinery which comes with the Ito integral makes
it more important as a mathematical object. It is very useful that we can convert
from the Ito to the Stratonovich interpretation of the stochastic integral. There are
other interpretations of the stochastic integral, e.g. the Klimontovich stochastic
integral.
The Definition of Brownian motion implies the scaling property
W (ct) = cW (t),
where the above should be interpreted as holding in law. From this it follows that,
if s = ct, then
dW 1 dW
= ,
ds c dt
again in law. Hence, if we scale time to s = ct in (7.1), then we get the equation
dz 1 1 dW
= h(z) + (z) , z(0) = z0 .
ds c c ds
Z t
t
X(t) = e x+ 2 e(ts) dW (s).
0
We can use Itos formula to obtain equations for the moments of the OU process.
The generator is:
L = xx + x2 .
Consequently:
Z t
n n
X(t) = x + nX(t)n + n(n 1)X(t)n2 dt
0
Z t
+n 2 X(t)n1 dW.
0
By taking the expectation in the above equation we obtain the equation for the mo-
ments of the OU process that we derived earlier using the Fokker-Planck equation:
Z t
Mn (t) = xn + (nMn (s) + n(n 1)Mn2 (s)) ds.
0
where we use the Ito interpretation of the stochastic differential. The generator of
this process is
2 x2 2
L = xx + .
2 x
The solution to this equation is
2
X(t) = X(0) exp ( )t + W (t) . (7.11)
2
FORMULA AND THE FOKKER-PLANCK EQUATION129
7.4. THE GENERATOR, ITOS
To derive this formula, we apply Itos formula to the function f (x) = log(x):
d log(X(t)) = L log(X(t)) dt + xx log(X(t)) dW (t)
1 2 x2 1
= x + 2 dt + dW (t)
x 2 x
2
= dt + dW (t).
2
Consequently:
X(t) 2
log = t + W (t)
X(0) 2
from which (7.11) follows. Notice that the Stratonovich interpretation of this equa-
tion leads to the solution
Note that if W were a smooth time-dependent function this formula would not be
correct: there is an additional term in LV , proportional to , which arises from the
lack of smoothness of Brownian motion. The precise interpretation of the expres-
sion for the rate of change of V is in integrated form:
Lemma 7.4.1. (Itos Formula) Assume that the conditions of Theorem 7.3.2 hold.
Let x(t) solve (7.1) and let V C 2 (Z, Rn ). Then the process V (z(t)) satisfies
Z t Z t
V (z(t)) = V (z(0)) + LV (z(s))ds + hV (z(s)), (z(s)) dW (s)i .
0 0
where the expectation is with respect to all Brownian driving paths. By averaging
in the Ito formula, which removes the stochastic integral, and using the Markov
property, it is possible to obtain the Backward Kolmogorov equation.
Theorem 7.4.2. Assume that is chosen sufficiently smooth so that the backward
Kolmogorov equation
v
= Lv for (z, t) Z (0, ),
t
v = for (z, t) Z {0} , (7.15)
For a Stratonovich SDE the rules of standard calculus apply: Consider the
Stratonovich SDE (7.29) and let V (x) C 2 (R). Then
dV
dV (X(t)) = (X(t)) (f (X(t)) dt + (X(t)) dW (t)) .
dx
1
= (f ) + ( ())). (7.16)
t 2
Now we can derive rigorously the Fokker-Planck equation.
7.5. LINEAR SDES 131
Theorem 7.4.3. Consider equation (7.2) with z(0) a random variable with density
0 (z). Assume that the law of z(t) has a density (z, t) C 2,1 (Z (0, )). Then
satisfies the Fokker-Planck equation
= L for (z, t) Z (0, ), (7.17a)
t
= 0 for z Z {0}. (7.17b)
Proof. Let E denote averaging with respect to the product measure induced by the
measure with density 0 on z(0) and the independent driving Wiener measure
on the SDE itself. Averaging over random z(0) distributed with density 0 (z), we
find
Z
E ((z(t))) = v(z, t)0 (z) dz
Z
Z
= (eLt )(z)0 (z) dz
ZZ
= (eL t 0 )(z)(z) dz.
Z
We use a density argument so that the identity can be extended to all L2 (Z).
Hence, from the above equation we deduce that
(z, t) = eL t 0 (z).
Differentiation of the above equation gives (7.17a). Setting t = 0 gives the initial
condition (7.17b).
or, componentwise,
d
X
dXi (t) = Aij Xj (t) + 2D dWi (t), i = 1, . . . d.
j=1
p
= (Axp) + Dp
t
or
d d
p X X 2p
= (Aij xj p) + D .
t xi
i,j
x2j j=1
Let us now solve the Fokker-Planck equation with initial conditions p(x, t|x0 , 0) =
(x x0 ). We take the Fourier transform of the Fokker-Planck equation to obtain
p
= Ak k p D|k|2 p (7.18)
t
with Z
d
p(x, t|x0 , 0) = (2) eikx p(k, t|x0 , t) dk.
Rd
The initial condition is
p(k, 0|x0 , 0) = eikx0 (7.19)
We know that the transition probability density of a linear SDE is Gaussian. Since
the Fourier transform of a Gaussian function is also Gaussian, we look for a solu-
tion to (7.18) which is of the form
1
p(k, t|x0 , 0) = exp(ik M (t) kT (t)k).
2
We substitute this into (7.18) and use the symmetry of A to obtain the equations
dM d
= AM and = 2A + 2DI,
dt dt
with initial conditions (which follow from (10.13)) M (0) = x0 and (0) = 0
where 0 denotes the zero d d matrix. We can solve these equations using the
spectral resolution of A = B T B. The solutions are
and
(t) = DA1 DA1 e2At .
L = V (x) + D
P
with V (x) = 21 xT Ax = 21 di,j=1 Aij xi xj . This is a confining potential and from
the theory presented in Section 6.5 we know that the process Xt is ergodic. The
invariant distribution is
1 1 T
ps (x) = e 2 x Ax (7.21)
Z
R 1 T dp
with Z = Rd e 2 x Ax dx = (2) 2 det(A1 ). Using the above calculations, we
can calculate the stationary autocorrelation matrix is given by the formula
Z Z
E(X0T Xt ) = xT0 xp(x, t|x0 , 0)ps (x0 ) dxdx0 .
We substitute the formulas for the transitions probability density and the station-
ary distribution, equations (7.21) and (7.20) into the above equations and do the
Gaussian integration to obtain
Z t
At
Xt = e X0 + 2D eA(ts) dW (s).
0
eAt = B T et B.
134 CHAPTER 7. STOCHASTIC DIFFERENTIAL EQUATIONS
dx 1
= h(x) + f (x)y, (7.22a)
dt
r
dy y 2D dV
= 2 + , (7.22b)
dt 2 dt
with V being a standard one-dimensional Brownian motion. We say that the pro-
cess x(t) is driven by colored noise: the noise that appears in (7.22a) has non-zero
correlation time. The correlation function of the colored noise (t) := y(t)/ is
(we take y(0) = 0)
1 D 2 |ts|
R(t) = E ((t)(s)) = e .
2
The power spectrum of the colored noise (t) is:
1 D2 1
f (x) = 2
x + (2 )2
2
D 1 D
=
4 x2 + 2 2
and, consequently,
y(t) y(s) 2D
lim E = (t s),
0 2
If we neglect the O() term on the right hand side then we arrive, again, at the
heuristic (7.23). Both of these arguments lead us to conjecture the limiting Ito
SDE: r
dX 2D dV
= h(X) + f (X) . (7.25)
dt dt
In fact, as applied, the heuristic gives the incorrect limit. Whenever white noise is
approximated by a smooth process, the limiting equation should be interpreted in
the Stratonovich sense, giving
r
dX 2D dV
= h(X) + f (X) . (7.26)
dt dt
This is usually called the Wong-Zakai theorem. A similar result is true in arbitrary
finite and even infinite dimensions. We will show this using singular perturbation
theory.
Theorem 7.6.1. Assume that the initial conditions for y(t) are stationary and that
the function f is smooth. Then the solution of eqn (7.22a) converges, in the limit
as 0 to the solution of the Stratonovich SDE (7.26).
v. In higher dimensions an additional drift term might appear due to the non-
commutativity of the row vectors of the diffusion matrix. This is related to
the Levy area correction in the theory of rough paths.
136 CHAPTER 7. STOCHASTIC DIFFERENTIAL EQUATIONS
1 1
L = 2
yy + Dy2 + f (x)yx + h(x)x
1 1
=: L0 + L1 + L2 .
2
The fast process is an stationary Markov process with invariant density
r
y2
(y) = e 2D . (7.27)
2D
The backward Kolmogorov equation is
u 1 1
= L 0 + L 1 + L 2 u . (7.28)
t 2
We look for a solution to this equation in the form of a power series expansion in
:
u (x, y, t) = u0 + u1 + 2 u2 + . . .
We substitute this into (7.28) and equate terms of the same power in to obtain the
following hierarchy of equations:
L0 u0 = 0,
L0 u1 = L1 u0 ,
u0
L0 u2 = L1 u1 + L2 u0 .
t
The ergodicity of the fast process implies that the null space of the generator L0
consists only of constant in y. Hence:
u0 = u(x, t).
L0 u1 = f (x)yx u.
This equation is solvable since the right hand side is orthogonal to the null space of
the adjoint of L0 (this is the Fredholm alterantive). We solve it using separation
of variables:
1
u1 (x, y, t) = f (x)x uy + 1 (x, t).
7.6. DERIVATION OF THE STRATONOVICH SDE 137
In order for the third equation to have a solution we need to require that the right
hand side is orthogonal to the null space of L0 :
Z
u0
L1 u1 + L2 u0 (y) dy = 0.
R t
We calculate: Z
u0 u
(y) dy = .
R t t
Furthermore: Z
L2 u0 (y) dy = h(x)x u.
R
Finally
Z Z
1
L1 u1 (y) dy = f (x)yx f (x)x uy + 1 (x, t) (y) dy
R R
1
= f (x)x (f (x)x u) hy 2 i + f (x)x 1 (x, t)hyi
D
= f (x)x (f (x)x u)
2
D D
= f (x)x f (x)x u + 2 f (x)2 x2 u.
2
Putting everything together we obtain the limiting backward Kolmogorov equation
u D D
= h(x) + 2 f (x)x f (x) x u + 2 f (x)2 x2 u,
t
(x) is a probability density for all values of c R. The presence of additive noise
in some sense trivializes the dynamics. The dependence of various averaged
quantities on c resembles the physical situation of a second order phase transition.
7.9. NOISE INDUCED TRANSITIONS 139
Thus, we have been able to transform (7.33) into an SDE with additive noise:
h i
dYt = (c ) e2Yt dt + dWt . (7.34)
> 1 c > .
Not all multiplicative random perturbations lead to ergodic behavior. The depen-
dence of the invariant distribution on c is similar to the physical situation of first
order phase transitions.
140 CHAPTER 7. STOCHASTIC DIFFERENTIAL EQUATIONS
where (t) is colored noise with correlation time 2 . In the limit where both small
time scales go to 0 we can get either Ito or Stratonovich or neither. See [40, 56].
Noise induced transitions are studied extensively in [32]. The material in Sec-
tion 7.9 is based on [47]. See also [46].
7.11 Exercises
1. Calculate all moments of the geometric Brownian motion for the Ito and Stratonovich
interpretations of the stochastic integral.
dx
= x(c + 2x2 x4 ).
dt
8.1 Introduction
L = p q q V p + (pp + Dp ). (8.1)
The L2 (dpdq)-adjoint is
L = p q q V p + (p (p) + Dp ) .
141
142 CHAPTER 8. THE LANGEVIN EQUATION
Theorem 8.2.1. Let V (x) be a smooth confining potential. Then the Markov pro-
cess with generator (8.45) is ergodic. The unique invariant distribution is the
Maxwell-Boltzmann distribution
1 H(p,q)
(p, q) = e (8.3)
Z
where
1
H(p, q) = kpk2 + V (q)
2
1
is the Hamiltonian, = (kB T ) is the inverse temperature and the normaliza-
tion factor Z is the partition function
Z
Z= eH(p,q) dpdq.
R2d
The proof of this result is very complicated, since the generator L is degenerate
and non-selfadjoint. See for example and the references therein.
Let (q, p, t) be the solution of the Kramers equation and let (q, p) be the
Maxwell-Boltzmann distribution. We can write
Let Xi := p i
. The L2 -adjoint of Xi is
Xi = pi + .
pi
We have that
d
X
1
S= Xi Xi .
i=1
Consequently, the generator of the Markov process {q(t), p(t)} can be written in
Hormanders sum of squares form:
d
X
1
L = A + Xi Xi . (8.5)
i=1
which spans Tp,q R2d for all p, q Rd . This shows that the generator L is a
hypoelliptic operator.
Let now Yi = p i
with L2 -adjoint Yi = q i q
V
i
. We have that
V
Xi Yi Yi Xi = pi .
qi qi pi
Consequently, the generator can be written in the form
d
X
1
L= (Xi Yi Yi Xi + Xi Xi ) . (8.6)
i=1
The Fokker-Planck equation has a similar structure to the Boltzmann equation (the
basic equation in the kinetic theory of gases), with the difference that the collision
operator for the FP equation is linear. Convergence of solutions of the Boltzmann
equation to the Maxwell-Boltzmann distribution has also been proved. See ??.
We can study the backward and forward Kolmogorov equations for (9.13) by
expanding the solution with respect to the Hermite basis. We consider the problem
in 1d. We set D = 1. The generator of the process is:
L = pq V (q)p + pp + p2 .
=: L1 + L0 ,
where
L0 := pp + p2 and L1 := pq V (q)p .
The backward Kolmogorov equation is
h
= Lh. (8.7)
t
The solution should be an element of the weighted L2 -space
Z
2 2 1 H(p,q)
L = f | |f | Z e dpdq < .
R2
We notice that the invariant measure of our Markov process is a product measure:
1 2
eH(p,q) = e 2 |p| eV (q) .
1 2
The space L2 (e 2 |p| dp) is spanned by the Hermite polynomials. Consequently,
we can expand the solution of (8.7) into the basis of Hermite basis:
X
h(p, q, t) = hn (q, t)fn (p), (8.8)
n=0
where fn (p) = 1/ n!Hn (p). Our plan is to substitute (8.8) into (8.7) and obtain a
sequence of equations for the coefficients hn (q, t). We have:
X
X
L0 h = L0 hn fn = nhn fn
n=0 n=0
8.2. THE FOKKER-PLANCK EQUATION IN PHASE SPACE (KLEIN-KRAMERS EQUATION)145
Furthermore
L1 h = q V p h + pq h.
We calculate each term on the right hand side of the above equation separately. For
this we will need the formulas
p fn = nfn1 and pfn = nfn1 + n + 1fn+1 .
X
X
pq h = pq hn fn = pp h0 + q hn pfn
n=0 n=1
X
= q h0 f1 + q hn nfn1 + n + 1fn+1
n=1
X
= ( n + 1q hn+1 + nq hn1 )fn
n=0
with h1 0. Furthermore
X
X
q V p h = q V hn p fn = q V hn nfn1
n=0 n=0
X
= q V hn+1 n + 1fn .
n=0
Consequently:
Lh = L1 + L1 h
X
= nhn + n + 1q hn+1
n=0
+ nq hn1 + n + 1q V hn+1 fn
This is set of equations is usually called the Brinkman hierarchy (1956). We can
use this approach to develop a numerical method for solving the Klein-Kramers
146 CHAPTER 8. THE LANGEVIN EQUATION
or p
q = p, p = 02 q p + .
2 1 W (8.11)
This is a linear equation that can be solved explicitly. Rather than doing this, we
will calculate the eigenvalues and eigenfunctions of the generator, which takes the
form
L = pq 02 qp + (pp + 1 p2 ). (8.12)
The Fokker-Planck operator is
L = pq 02 qp + (pp + 1 p2 ). (8.13)
The process {q(t), p(t)} is an ergodic Markov process with Gaussian invariant
measure 2
0 2 p2 2
0
(q, p) dqdp = e q . (8.14)
2
8.3. THE LANGEVIN EQUATION IN A HARMONIC POTENTIAL 147
and
b = 01 1/2 q , b+ = 01 1/2 q + 0 1/2 p. (8.16)
We have that
a+ a = 1 p2 + pp
and
b+ b = 1 q2 + qq
Consequently, the operator
Lb = a+ a b+ b (8.17)
[a+ , a ] = 1, (8.18a)
[b+ , b ] = 1, (8.18b)
[a , b ] = 0. (8.18c)
See Exercise 3. Using now the operators a and b we can write the generator L
in the form
L = a+ a 0 (b+ a a+ b ), (8.19)
which is a particular case of (8.6). In order to calculate the eigenvalues and eigen-
functions of (8.19) we need to make an appropriate change of variables in order
to bring the operator L into the decoupled form (8.17). Clearly, this is a linear
transformation and can be written in the form
Y = AX
for some appropriate constants C and D. Since our goal is, essentially, to map L
to the two-dimensional OU process, we require that that the operators c and d
satisfy the canonical commutation relations
[c+ , c ] = 1, (8.21a)
[d+ , d ] = 1, (8.21b)
[c , d ] = 0. (8.21c)
The operators c and d should be given as linear combinations of the old op-
erators a and b . From the structure of the generator L (8.19), the decoupled
form (8.20) and the commutation relations (8.21) and (8.18) we conclude that c
and d should be of the form
c+ = 11 a+ + 12 b+ , (8.22a)
c = 21 a + 22 b , (8.22b)
d+ = 11 a+ + 12 b+ , (8.22c)
d = 21 a + 22 b . (8.22d)
Notice that the c and d are not the adjoints of c+ and d+ . If we substitute now
these equations into (8.20) and equate it with (8.19) and into the commutation re-
lations (8.21) we obtain a system of equations for the coefficients {ij }, {ij }. In
order to write down the formulas for these coefficients it is convenient to introduce
the eigenvalues of the deterministic problem
q = q 02 q.
q(t) = C1 e1 t + C2 e2 t
with q
1,2 = , = 2 402 . (8.23)
2
The eigenvalues satisfy the relations
1 + 2 = , 1 2 = , 1 2 = 02 . (8.24)
8.3. THE LANGEVIN EQUATION IN A HARMONIC POTENTIAL 149
Proposition 8.3.1. Let L be the generator (8.19) and let c , dpm be the operators
1 p p
c+ = 1 a+ + 2 b+ , (8.25a)
1 p p
c = 1 a 2 b , (8.25b)
1 p p
d+ = 2 a+ + 1 b+ , (8.25c)
1 p p
d = 2 a + 1 b . (8.25d)
Then c , d satisfy the canonical commutation relations (8.21) as well as
L = 1 c+ c 2 d+ d . (8.27)
1
[c+ , c ] = 1 [a+ , a ] 2 [b+ , b ]
1
= (1 + 2 ) = 1.
Similarly,
1
[d+ , d ] = 2 [a+ , a ] + 1 [b+ , b ]
1
= (2 1 ) = 1.
Clearly, we have that
[c+ , d+ ] = [c , d ] = 0.
Furthermore,
1 p p
[c+ , d ] = 1 2 [a+ , a ] + 1 2 [b+ , b ]
1 p p
= ( 1 2 + 1 2 ) = 0.
150 CHAPTER 8. THE LANGEVIN EQUATION
Finally:
[L, c+ ] = 1 c+ c c+ + 1 c+ c+ c
= 1 c+ (1 + c+ c ) + 1 c+ c+ c
= 1 c+ (1 + c+ c ) + 1 c+ c+ c
= 1 c+ ,
L = 1 c+ c 2 d+ d
22 21 + + 1 2 1p
= a a + 0b b + (1 2 )a+ b + 1 2 (1 + 2 )b+ a
= a+ a 0 (b+ a a+ b ),
Using now (8.27) we can readily obtain the eigenvalues and eigenfunctions of
L. From our experience with the two-dimensional OU processes (or, the Schrodinger
operator for the two-dimensional quantum harmonic oscillator), we expect that the
eigenfunctions should be tensor products of Hermite polynomials. Indeed, we have
the following, which is the main result of this section.
Theorem 8.3.2. The eigenvalues and eigenfunctions of the generator of the Markov
process {q, p} (8.11) are
1 1
nm = 1 n + 2 m = (n + m) + (n m), n, m = 0, 1, . . . (8.28)
2 2
and
1
nm (q, p) = (c+ )n (d+ )m 1, n, m = 0, 1, . . . (8.29)
n!m!
Proof. We have
and similarly [L, (d+ )2 ] = 21 (c+ )2 . A simple induction argument now shows
that (see Exercise 8.3.3)
L(c+ )n (d+ )n 1
= (c+ )n L(d+ )m 1 n1 (c+ )n (d+ m)1
= (c+ )n (d+ )m L1 m2 (c+ )n (d+ m)1 n1 (c+ )n (d+ m)1
= n1 (c+ )n (d+ m)1 m2 (c+ )n (d+ m)1
00 = 1.
1 p + 2 0 q
10 = .
2 p + 1 0 q
01 =
2 1 2 + 1 p2 2 + p1 0 q + 0 q2 p + 2 0 2 q 2 1
11 = .
1 + p2 1 + 2 2 p 1 0 q 2 + 0 2 q 2 2
20 = .
2
2 + p2 2 + 2 2 p 1 0 q 1 + 0 2 q 2 1
02 = .
2
152 CHAPTER 8. THE LANGEVIN EQUATION
nn = n.
On the other hand, in the overdamped regime, > 20 all eigenvalues are real:
q
1 1
nm = (n + m) + 2 402 (n m), > 20 .
2 2
In fact, in the overdamped limit + (which we will study in Chapter ??), the
eigenvalues of the generator L converge to the eigenvalues of the generator of the
OU process:
2
nm = n + 0 (n m) + O( 3 ).
This is consistent with the fact that in this limit the solution of the Langevin equa-
tion converges to the solution of the OU SDE. See Chapter ?? for details.
8.3. THE LANGEVIN EQUATION IN A HARMONIC POTENTIAL 153
1
)
nm
Im (
3
0 0.5 1 1.5 2 2.5 3
Re (nm)
Lb := A + S (8.32)
= 0 (b+ a b a+ ) + a+ a (8.33)
+ +
= 1 (c ) (c ) 2 (d ) (d ), (8.34)
where
1 p p
(c+ ) = 1 a + 2 b , (8.35a)
1 p p
(c ) = 1 a+ 2 b+ , (8.35b)
1 p p
(d+ ) = 2 a + 1 b , (8.35c)
1 p p
(d ) = 2 a+ + 1 b+ . (8.35d)
b nm = nm nm ,
L
1
nm = ((c ) )n ((d ) )m 1. (8.36)
n!m!
Proof. We will use formulas (8.31). Notice that using the third and fourth of these
equations together with the fact that c 1 = d 1 = 0 we can conclude that (for
n > )
(c ) (c+ )n 1 = n(n 1) . . . (n + 1)(c+ )n . (8.38)
8.4. ASYMPTOTIC LIMITS FOR THE LANGEVIN EQUATION 155
We have
Z Z Z Z
1
nm k dpdq = ((c+ ))n ((d+ ))m 1((c ) ) ((d ) )k 1 dpdq
n!m!!k!
Z Z
n(n 1) . . . (n + 1)m(m 1) . . . (m k + 1)
= ((c+ ))n ((d+ ))mk 1 d
n!m!!k!
= n mk ,
L (f ) = Lf
b
we immediately conclude that the the Fokker-Planck operator has the same eigen-
b The eigenfunctions are
values as those of L and L.
1
nm = nm = ((c ) )n ((d ) )m 1. (8.39)
n!m!
ii. Small noise asymptotics/large times (rare events): the theory of large devia-
tions, escape from a potential well, exit time problems.
iii. Small and large friction asymptotics for the Fokker-Planck equation: The
FreidlinWentzell (underdamped) and Smoluchowski (overdamped) limits.
iv. Large time asymptotics for the Langevin equation in a periodic potential:
homogenization and averaging.
156 CHAPTER 8. THE LANGEVIN EQUATION
We will study various asymptotic limits for the Langevin equation (we have set
m = 1)
p
q = V (q) q + 2 1 W . (8.40)
There are two parameters in the problem, the friction coefficient and the in-
verse temperature . We want to study the qualitative behavior of solutions to this
equation (and to the corresponding Fokker-Planck equation). There are various
asymptotic limits at which we can eliminate some of the variables of the equa-
tion and obtain a simpler equation for fewer variables. In the large temperature
limit, 1, the dynamics of (9.13) is dominated by diffusion: the Langevin
equation (9.13) can be approximated by free Brownian motion:
p
q = .
2 1 W
where can be either the reaction rate or the diffusion coefficient. The small
temperature asymptotics will be studied later for the case of a bistable potential
(reaction rate) and for the case of a periodic potential (diffusion coefficient).
Assuming that the temperature is fixed, the only parameter that is left is the
friction coefficient . The large and small friction asymptotics can be expressed in
terms of a slow/fast system of SDEs. In many applications (especially in biology)
the friction coefficient is large: 1. In this case the momentum is the fast
variable which we can eliminate to obtain an equation for the position. This is the
overdamped or Smoluchowski limit. In various problems in physics the friction
coefficient is small: 1. In this case the position is the fast variable whereas the
energy is the slow variable. We can eliminate the position and obtain an equation
for the energy. This is the underdampled or Freidlin-Wentzell limit. In both
cases we have to look at sufficiently long time scales.
8.4. ASYMPTOTIC LIMITS FOR THE LANGEVIN EQUATION 157
q (t) = (t/ ).
= 1, = , 1:
p
q = 2 V (q ) q + .
2 2 1 W (8.44)
Under this scaling the interesting limit is the underdamped limit, 1. We will
see later that in the limit as 0 the energy of the solution to (8.44) converges to
a stochastic process on a graph.
This systems of SDEs defined a Markov process in phase space. Its generator is
1 1
L = 2
p p + 1 + p q q V p
1 1
=: L0 + L1 .
2
This is a singularly perturbed differential operator. We will derive the Smolu-
chowski equation (8.46) using a pathwise technique, as well as by analyzing the
corresponding Kolmogorov equations.
We apply Itos formula to p:
1 p 1
dp(t) = L p(t) dt + 2 p p(t) dW
1 1 1 p 1
= 2 p(t) dt q V (q(t)) dt + 2 dW.
Consequently:
Z Z t p
1 t
p(s) ds = q V (q(s)) ds + 2 1 W (t) + O().
0 0
E|p(t)|2 6 C.
This estimate is true, under appropriate assumptions on the potential V (q) and on
the initial conditions. In fact, we can prove a pathwise approximation result:
!1/p
p
E sup |q (t) q(t)| 6 C2 ,
t[0,T ]
The pathwise derivation of the Smoluchowski equation implies that the solu-
tion of the Fokker-Planck equation corresponding to the Langevin equation (8.45)
converges (in some appropriate sense to be explained below) to the solution of the
Fokker-Planck equation corresponding to the Smoluchowski equation (8.46). It is
important in various applications to calculate corrections to the limiting Fokker-
Planck equation. We can accomplish this by analyzing the Fokker-Planck equation
for (8.45) using singular perturbation theory. We will consider the problem in one
dimension. This mainly to simplify the notation. The multidimensional problem
can be treated in a very similar way.
The FokkerPlanck equation associated to equations (8.47) and (8.48) is
= L
t
1 1
= (pq + q V (q)p ) + 2 p (p) + 1 p2
1 1
=: L + L . (8.49)
2 0 1
The invariant distribution of the Markov process {q, p}, if it exists, is
Z
1 H(p,q)
(p, q) = e , Z= eH(p,q) dpdq,
Z R 2
Proposition 8.4.1. The function f (p, q, t) defined in (8.50) satisfies the equation
f 1 1 2
1
= pq + p (pq q V (q)p ) f
t 2
1 1
=: 2
L0 L1 f. (8.51)
Remark 8.4.2. This is almost the backward Kolmogorov equation with the dif-
ference that we have L1 instead of L1 . This is related to the fact that L0 is a
symmetric operator in L2 (R2 ; Z 1 eH(p,q) ), whereas L1 is antisymmetric.
Proof. We note that L0 0 = 0 and L1 0 = 0. We use this to calculate:
L0 = L0 (f 0 ) = p (f 0 ) + 1 p2 (f 0 )
= 0 pp f + 0 1 p2 f + f L0 0 + 2 1 p f p 0
= pp f + 1 p2 f 0 = 0 L0 f.
160 CHAPTER 8. THE LANGEVIN EQUATION
Similarly,
L1 = L1 (f 0 ) = (pq + q V p ) (f 0 )
= 0 (pq f + q V p f ) = 0 L1 f.
We will assume that the initial conditions for (8.51) depend only on q:
Another way for stating this assumption is the following: Let H = L2 (R2d ; (p, q))
and define the projection operator P : H 7 L2 (Rd ; (q)) with (q) = Z1q eV (q) , Zq =
R V (q) dq:
Rd e Z
1 |p|2
P := e 2 dp, (8.53)
Zp Rd
R 2
with Zp := Rd e|p| /2 dp. Then, assumption (10.13) can be written as
P fic = fic .
N
X
f (p, q, t) = n fn (p, q, t). (8.54)
n=0
We substitute this expansion into eqn. (8.51) to obtain the following system of
equations.
L0 f0 = 0, (8.55a)
L0 f1 = L1 f0 , (8.55b)
f0
L0 f2 = L1 f1 (8.55c)
t
fn2
L0 fn = L1 fn1 , n = 3, 4 . . . N. (8.55d)
t
8.4. ASYMPTOTIC LIMITS FOR THE LANGEVIN EQUATION 161
L1 f0 = pq f.
f1 = pq f + 1 (q, t).
1 d
k1 k2 6 Ck1 k2 .
2 dt
We use Gronwalls inequality now to conclude that
1 0.
Putting everything together we obtain the first two terms in the -expansion of the
FokkerPlanck equation (8.51):
(p, q, t) = Z 1 eH(p,q) f + (pq f ) + O(2 ) ,
where f is the solution of (8.56). Notice that we can rewrite the leading order term
to the expansion in the form
1 2 /2
(p, q, t) = (2 1 ) 2 ep V (q, t) + O(),
L0 k = k k .
8.4. ASYMPTOTIC LIMITS FOR THE LANGEVIN EQUATION 163
b n1 = 0,
Lf
s
k+1 b p
1
Lfn,k+1 + k 1 q fn,k1 = kfn+1,k , k = 1, 2 . . . , n 1,
p
n 1 q fn,n1 = nfn+1,n ,
p
(n + 1) 1 q fn,n = (n + 1)fn+1,n+1 .
Using this method we can obtain the first three terms in the expansion:
p 1
(x, y, t) = 0 (p, q) f + ( 1 q f 1 ) + 2 q2 f 2 + f20
2
r !!
3 p p
+3 3 f 3 + 1 L b 2 f 1 q f20 1
3! q q
+O(4 ),
This is the equation for an O(1/) Hamiltonian system perturbed by O(1) noise.
We expect that, to leading order, the energy is conserved, since it is conserved for
the Hamiltonian system. We apply Itos formula to the Hamiltonian of the system
to obtain
p
H = 1 p2 + 2 1 p2 W
The Hamiltonian is the slow variable, whereas the momentum (or position) is the
fast variable. Assuming that we can average over the Hamiltonian dynamics, we
obtain the limiting SDE for the Hamiltonian:
p
H = 1 hp2 i + 2 1 hp2 iW . (8.60)
The limiting SDE lives on the graph associated with the Hamiltonian system. The
domain of definition of the limiting Markov process is defined through appropriate
boundary conditions (the gluing conditions) at the interior vertices of the graph.
We identify all points belonging to the same connected component of the a
level curve {x : H(x) = H}, x = (q, p). Each point on the edges of the graph
correspond to a trajectory. Interior vertices correspond to separatrices. Let Ii , i =
1, . . . d be the edges of the graph. Then (i, H) defines a global coordinate system
on the graph.
We will study the small asymptotics by analyzing the corresponding back-
ward Kolmogorov equation using singular perturbation theory. The generator of
the process {q , p } is
L = 1 (pq q V p ) pp + 1 p2
= 1 L0 + L1 .
Let u = E(f (p (p, q; t), q (p, q; t))). It satisfies the backward Kolmogorov equa-
tion associated to the process {q , p }:
u 1
= L 0 + L 1 u . (8.61)
t
u = u0 + u1 + 2 u2 + . . .
We substitute this ansatz into (8.61) and equate equal powers in to obtain the
following sequence of equations:
L0 u0 = 0, (8.62a)
u0
L0 u1 = L1 u1 + , (8.62b)
t
u1
L0 u2 = L1 u1 + . (8.62c)
t
8.4. ASYMPTOTIC LIMITS FOR THE LANGEVIN EQUATION 165
.........
Notice that the operator L0 is the backward Liouville operator of the Hamiltonian
system with Hamiltonian
1
H = p2 + V (q).
2
We assume that there are no integrals of motion other than the Hamiltonian. This
means that the null space of L0 consists of functions of the Hamiltonian:
N (L0 ) = functions ofH . (8.63)
Let us now analyze equations (8.62). We start with (8.62a); eqn. (8.63) implies that
u0 depends on q, p through the Hamiltonian function H:
Now we proceed with (8.62b). For this we need to find the solvability condition
for equations of the form
L0 u = f (8.65)
This implies that the solvability condition for equation (8.83) is that
Z
f (p, q)F (H(p, q)) dpdq = 0, F Cb (R). (8.66)
R2
Thus, we have obtained the limiting backward Kolmogorov equation for the energy,
which is the slow variable. From this equation we can read off the limiting SDE
for the Hamiltonian:
H = b(H) + (H)W (8.69)
where
b(H) = 1 hp1 i1 hpi, (H) = 1 hp1 i1 hpi.
Notice that the noise that appears in the limiting equation (8.69) is multiplica-
tive, contrary to the additive noise in the Langevin equation.
As it well known from classical mechanics, the action and frequency are de-
fined as Z
I(E) = p(q, E) dq
and
dI 1
(E) = 2 ,
dE
respectively. Using the action and the frequency we can write the limiting Fokker
Planck equation for the distribution function of the energy in a very compact form.
Theorem 8.4.3. The limiting FokkerPlanck equation for the energy distribution
function (E, t) is
(E)
= I(E) + 1 . (8.70)
t E E 2
Proof. We notice that
Z Z
dI p
= dq = p1 dq
dE E
and consequently
(E)
hp1 i1 = .
2
Hence, the limiting FokkerPlanck equation can be written as
I(E)(E) 2 I
= 1 + 1
t E 2 E 2 2
1 I 1 dI 1
= + + + I
E E 2 E dE 2 E E 2
I
= + 1 I
E 2 E E 2
1 (E)
= I(E) + ,
E E 2
168 CHAPTER 8. THE LANGEVIN EQUATION
Remarks 8.4.4. i. We emphasize that the above formal procedure does not
provide us with the boundary conditions for the limiting FokkerPlanck equa-
tion. We will discuss about this issue in the next section.
We will use this equation later on to calculate the rate of escape from a
potential barrier in the energy-diffusion-limited regime.
Goal: Calculate the effective drift and the effective diffusion tensor
hx(t)i
Uef f = lim (8.73)
t t
and
hx(t) hx(t)i) (x(t) hx(t)i)i
Def f = lim . (8.74)
t 2t
where is the friction coefficient, kB the Boltzmann constant and T denotes the
temperature. (t) stands for the standard ddimensional white noise process, i.e.
V (x + ei ) = V (x), i = 1, . . . , d,
x(t)
= y(t), (8.76a)
p
= V (x(t)) y(t)
y(t) + 2kB T (t), (8.76b)
The process {x(t), y(t)} is Markovian with generator
L = y x V (x) y + (y y + Dy ) .
In writing the above we have set D = KB T . This process is ergodic. The unique
invariant measure is absolutely continuous with respect to the Lebesgue measure
and its density is the MaxwellBoltzmann distribution
1 1
(y, x) = n e D H(x,y) , (8.77)
(2D) Z
2
R
where Z = Td eV (x)/D dx and H(x, y) is the Hamiltonian of the system
1 2
H(x, y) = y + V (x).
2
The long time behavior of solutions to (8.75) is governed by an effective Brownian
motion. Indeed, the following central limit theorem holds [65, 55, ?]
where (dx dy) = (x, y)dxdy and the vector valued function is the solution of
the Poisson equation
L = y. (8.79)
We are interested in analyzing the dependence of Def f on . We will mostly
focus on the one dimensional case. We start by rescaling the Langevin equa-
tion (9.13) p
x = F (x) x + 2 1 W , (8.80)
where we have set F (x) = V (x). We will assume that the potential is periodic
with period 2 in every direction. Since we expect that at sufficiently long length
and time scales the particle performs a purely diffusive motion, we perform a dif-
fusive rescaling to the equations of motion (9.13): t t/2 , x x . Using the
(c t) = 1 W
fact that W (t) in law we obtain:
c
1 x p
,
2 x
= F x + 2 1 W
Introducing p = x and q = x/ we write this equation as a first order system:
1
x = p,
p = 1
F (q) 1
p ,
+ 12 1 W (8.81)
2 2
1
q = 2
p,
where E denotes the expectation with respect to the Brownian motion W (t) in
the Langevin equation and f is a smooth function.2 The evolution of the func-
tion u (p, q, x, t) is governed by the backward Kolmogorov equation associated to
2
In other words, we have that
Z
u (p, q, x, t) = f (x, v, t; p, q)(x, v, t; p, q)(p, q) dpdqdxdv,
8.5. BROWNIAN MOTION IN PERIODIC POTENTIALS 171
u 1 1
= p x u + 2 q V (q) p + p q + p p + 1 p u .
t
1 1
:= L 0 + L 1 u , (8.82)
2
where:
L0 = q V (q) p + p q + p p + 1 p ,
L1 = p x
The invariant distribution of the fast process q(t), p(t) in Td Rd is the Maxwell-
Boltzmann distribution
Z
1 H(q,p)
(q, p) = Z e , Z= eH(q,p) dqdp,
Td Rd
L0 (q, p) = 0,
where L0 denotes the Fokker-Planck operator which is the L2 -adjoint of the gen-
erator of the process L0 :
L0 f = q V (q) p f p q f + p (pf ) + 1 p f .
where (x, v, t; p, q) is the solution of the Fokker-Planck equation and (p, q) is the initial distribu-
tion.
3
it is more customary in the physics literature to use the forward Kolmogorov equation, i.e. the
Fokker-Planck equation. However, for the calculation presented below, it is more convenient to use
the backward as opposed to the forward Kolmogorov equation. The two formulations are equivalent.
See [57, Ch. 6] for details.
172 CHAPTER 8. THE LANGEVIN EQUATION
u = u0 + u1 + 2 u2 + . . . (8.86)
L0 u0 = 0, (8.87a)
L0 u1 = L1 u0 , (8.87b)
u0
L0 u2 = L1 u1 + . (8.87c)
t
From the first equation in (8.87) we deduce that u0 = u0 (x, t), since the null
space of L0 consists of functions which are constants in p and q. Now the second
equation in (8.87) becomes:
L0 u1 = p x u0 .
Since hpi = 0, the right hand side of the above equation is mean-zero with respect
to the Maxwell-Boltzmann distribution. Hence, the above equation is well-posed.
We solve it using separation of variables:
u1 = (p, q) x u0
with
L0 = p. (8.88)
This Poisson equation is posed on Td Rd . The solution is periodic in q and
satisfies condition (8.85). Now we proceed with the third equation in (8.87). We
apply the solvability condition to obtain:
Z
u0
= L1 u1 (p, q) dpdq
t Td Rd
X d Z 2
u0
= pi j (p, q) dpdq .
Td Rd xi xj
i,j=1
8.5. BROWNIAN MOTION IN PERIODIC POTENTIALS 173
This is the Backward Kolmogorov equation which governs the dynamics on large
scales. We write it in the form
d
X
u0 2 u0
= Dij (8.89)
t xi xj
i,j=1
The calculation of the effective diffusion tensor requires the solution of the bound-
ary value problem (8.88) and the calculation of the integral in (8.90). The limiting
backward Kolmogorov equation is well posed since the diffusion tensor is non-
negative. Indeed, let be a unit vector in Rd . We calculate (we use the notation
= and h, i for the Euclidean inner product)
Z Z
h, Di = (p )( ) dpdq = L0 dpdq
Z
= 1 p 2 dpdq > 0, (8.91)
where V (x) is periodic with period 2 and F is a constant force field. The formulas
for the effective drift and the effective diffusion tensor are
Z Z
V = p(q, p) dqdp, D = (p V ) (p, q) dpdq, (8.93)
Rd Td Rd Td
where
L = p V, (8.94a)
Z
L = 0, (p, q) dpdq = 1. (8.94b)
Rd Td
174 CHAPTER 8. THE LANGEVIN EQUATION
with
L = p q + (q V + F ) p + p p + 1 p . (8.95)
We have used to denote the tensor product between two vectors; L denotes the
L2 -adjoint of the operator L, i.e. the Fokker-Planck operator. Equations (8.94)
are equipped with periodic boundary conditions in q. The solution of the Poisson
equation (8.94) is also taken to be square integrable with respect to the invariant
density (q, p): Z
|(q, p)|2 (p, q) dpdq < +.
Rd Td
The diffusion tensor is nonnegative definite. A calculation similar to the one used
to derive (8.91) shows the positive definiteness of the diffusion tensor:
Z
h, Di = 1 p 2 (p, q) dpdq > 0, (8.96)
for every vector in Rd . The study of diffusion in a tilted periodic potential, in the
underdamped regime and in high dimensions, based on the above formulas for V
and D, will be the subject of a separate publication.
= x V x +
x
where (t) stands for Gaussian white noise in one dimension with correlation func-
tion
h(t)(s)i = 2kB T (t s).
We assume that the (x, v) process is stationary, i.e. that the initial conditions are
distributed according to the Maxwell-Boltzmann distribution
(q, p) = Z 1 eH(p,q) .
8.5. BROWNIAN MOTION IN PERIODIC POTENTIALS 175
The function v(t) satisfies the backward Kolmogorov equation which governs the
evolution of observables [59, Ch. 6]
v
= Lv, v(0; p, q) = p. (8.99)
t
We can write, formally, the solution of (8.99) as
v = eLt p. (8.100)
We combine now equations (8.98) and (8.100) to obtain the following formula for
the velocity autocorrelation function
Z Z
hv(t; q, p)v(0; q, p)i = p eLt p (p, q) dpdq. (8.101)
In this subsection we solve the Poisson equation (8.88) in one dimension perturba-
tively for small . We shall use singular perturbation theory for partial differential
equations. The operator L0 that appears in (8.88) can be written in the form
L0 = LH + LOU
where LH stands for the (backward) Liouville operator associated with the Hamil-
tonian H(p, q) and LOU for the generator of the OU process, respectively:
LH = pq q V p , LOU = pp + 1 p2 .
We expect that the solution of the Poisson equation scales like 1 when 1.
Thus, we look for a solution of the form
1
= 0 + 1 + 2 + . . . (8.102)
We substitute this ansatz in (8.88) to obtain the sequence of equations
LH 0 = 0, (8.103a)
LH 1 = p + LOU 0 , (8.103b)
LH 2 = LOU 1 . (8.103c)
From equation (8.103a) we deduce that, since the 0 is in the null space of the
Liouville operator, the first term in the expansion is a function of the Hamiltonian
z(p, q) = 12 p2 + V (q):
0 = 0 (z(p, q)).
Now we want to obtain an equation for 0 by using the solvability condition for
(8.103b). To this end, we multiply this equation by an arbitrary function of z,
8.6. THE UNDERDAMPED AND OVERDAMPED LIMITS OF THE DIFFUSION COEFFICIENT177
2
LOU = ( 1 p2 ) + 1 p2 2 .
z z
Hence, the integral equation for 0 (z) becomes
Z + Z 2
1 2 1 2 1
g(z) p(z, q) + ( p ) + p 2
0 (z) dzdq = 0.
Emin z z p(z, q)
Let E0 denote the critical energy, i.e. the energy along the separatrix (homoclinic
orbit). We set
Z x2 (z) Z x2 (z)
1
S(z) = p(z, q) dq, T (z) = dq,
x1 (z) x1 (z) p(z, q)
where Riskens notation [64, p. 301] has been used for x1 (z) and x2 (z).
We need to consider the cases z > E0 , p > 0 , z > E0 , p < 0 and
Emin < z < E0 separately.
We consider first the case E > E0 , p > 0. In this case x1 (x) = , x2 (z) =
. We can perform the integration with respect to q to obtain
Z +
1 1 2
g(z) 2 + ( T (z) S(z)) + S(z) 2 0 (z) dz = 0,
E0 z z
This equation is valid for every test function g(z), from which we obtain the fol-
lowing differential equation for 0 :
1 1 2
L := 1 S(z) + S(z) 1 = , (8.104)
T (z) T (z) T (z)
where primes denote differentiation with respect to z and where the subscript 0 has
been dropped for notational simplicity.
178 CHAPTER 8. THE LANGEVIN EQUATION
A similar calculation shows that in the regions E > E0 , p < 0 and Emin <
E < E0 the equation for 0 is
2
L = , E > E0 , p < 0 (8.105)
T (z)
and
L = 0, Emin < E < E0 . (8.106)
Equations (8.104), (8.105), (8.106) are augmented with condition (8.85) and a con-
tinuity condition at the critical energy [18]
where 1 , 2 , 3 are the solutions of equations (8.104), (8.105) and (8.106), re-
spectively.
The average of a function h(q, p) = h(q, p(z, q)) can be written in the form [64,
p. 303]
Z Z
hh(q, p)i := h(q, p) (q, p) dqdp
Z + Z x2 (z)
= Z1 h(q, p(z, q)) + h(q, p(z, q)) (p(q, z))1 ez dzdq,
Emin x1 (z)
to leading order in , and where 0 (z) is the solution of the two point boundary
value problem (8.104). We remark that if we start with formula D = 1 h|p |2 i
8.6. THE UNDERDAMPED AND OVERDAMPED LIMITS OF THE DIFFUSION COEFFICIENT179
for the diffusion coefficient, we obtain the following formula, which is equivalent
to (8.109): Z
4 1 +
D= Z |z 0 (z)|2 ez dz.
E0
Now we solve the equation for 0 (z) (for notational simplicity, we will drop the
subscript 0 ). Using the fact that S (z) = T (z), we rewrite (8.104) as
1 (S ) + S = 2.
Condition (8.85) implies that the derivative of the unique solution of (8.104) is
(z) = S 1 (z).
We use this in (8.109), together with an integration by parts, to obtain the following
formula for the diffusion coefficient:
Z
1 2 1 1 + ez
D = 8 Z dz. (8.110)
E0 S(z)
We emphasize the fact that this formula is exact in the limit as 0 and is valid
for all periodic potentials and for all values of the temperature.
Consider now the case of the nonlinear pendulum V (q) = cos(q). The
partition function is
(2)3/2
Z = J0 (),
1/2
where J0 () is the modified Bessel function of the first kind. Furthermore, a simple
calculation yields !
r
2
S(z) = 25/2 z + 1E ,
z+1
where E() is the complete elliptic integral of the second kind. The formula for the
diffusion coefficient becomes
Z +
1 ez
D= p dz. (8.111)
2 1/2 J0 () 1 z + 1E( 2/(z + 1))
180 CHAPTER 8. THE LANGEVIN EQUATION
= 0 + 1 + 2 2 + 3 3 + . . . (8.114)
We substitute this into (8.88) and obtain the following sequence of equations:
LOU 0 = 0, (8.115a)
LOU 1 = p + LH 0 , (8.115b)
LOU 2 = LH 1 , (8.115c)
LOU 3 = LH 2 . (8.115d)
8.6. THE UNDERDAMPED AND OVERDAMPED LIMITS OF THE DIFFUSION COEFFICIENT181
LOU 1 = p(1 + q ).
Let
1
2 2 p2
(p) = e 2 ,
be the invariant distribution of the OU process (i.e. LOU (p) = 0). The solvabil-
ity condition for an equation of the form LOU = f requires that the right hand
side averages to 0 with respect to (p), i.e. that the right hand side of the equation
is orthogonal to the null space of the adjoint of LOU . This condition is clearly
satisfied for the equation for 1 . Thus, by Fredholm alternative, this equation has
a solution which is
1 (p, q) = (1 + q )p + 1 (q),
where the function 1 (q) of is to be determined. We substitute this into the right
hand side of the third equation to obtain
LOU 2 = p2 q2 q V (1 + q ) + pq 1 (q).
From the solvability condition for this we obtain an equation for (q):
1 q2 q V (1 + q ) = 0, (8.116)
together with the periodic boundary conditions. The derivative of the solution of
this two-point boundary value problem is
2
q + 1 = R V (q) dq
eV (q) . (8.117)
e
The first two terms in the large expansion of the solution of equation (8.88) are
1 1
(p, q) = (q) + (1 + q ) + O ,
2
where (q) is the solution of (8.116). Substituting this in the formula for the diffu-
sion coefficient and using (8.117) we obtain
Z Z
4 2 1
D = p (p, q) dpdq = +O ,
Z Zb 3
182 CHAPTER 8. THE LANGEVIN EQUATION
R R
where Z = eV (q) , Zb = eV (q) . This is, of course, the Lifson-Jackson
formula which gives the diffusion coefficient in the overdamped limit [43]. Con-
tinuing in the same fashion, we can also calculate the next two terms in the expan-
sion (8.114), see Exercise 4. From this, we can compute the next order correction
to the diffusion coefficient. The final result is
4 2 4 2 Z1 1
D= +O , (8.118)
Z Z b 3
ZZ b 2 5
R
where Z1 = |V (q)|2 eV (q) dq.
In the case of the nonlinear pendulum, V (q) = cos(q), formula (8.118) gives
1 2 J2 () 2 1
D= J () 3 J0 () + O , (8.119)
0 3
J0 () 5
where
LV = q V q + 1 q
and (q) is the solution of the PDE LV = q V with periodic boundary condi-
tions.
Now we prove several properties of the effective diffusion tensor in the over-
damped limit. For this we will need the following integration by parts formula
Z Z Z
y dy = y () y dy = ( y ) dy. (8.121)
Td Td Td
Theorem 8.6.1. The effective diffusion tensor D0 (8.120) satisfies the upper and
lower bounds
D
6 h, Ki 6 D||2 Rd , (8.122)
ZZb
8.6. THE UNDERDAMPED AND OVERDAMPED LIMITS OF THE DIFFUSION COEFFICIENT183
where
Z
Zb = eV (y)/D dy.
Td
Proof. The lower bound follows from the general lower bound (??), equation (??)
and the formula for the Gibbs measure. To establish the upper bound, we use
(8.121) and (??) to obtain
Z Z
T
K = DI + 2D () dy + y V dy
d d
ZT ZT
= DI 2D y dy + y V dy
Td Td
Z Z
= DI 2 y V dy + y V dy
Td Td
Z
= DI y V dy
d
ZT
= DI L0 dy
Td
Z
= DI D y y dy. (8.123)
Td
Hence, for = ,
Z
h, Ki = D||2 D |y |2 dy
Td
6 D||2 .
The one dimensional case is always in gradient form: b(y) = y V (y). Further-
more in one dimension we can solve the cell problem (??) in closed form and
calculate the effective diffusion coefficient explicitlyup to quadratures. We start
184 CHAPTER 8. THE LANGEVIN EQUATION
with the following calculation concerning the structure of the diffusion coefficient.
Z 1 Z 1
K = D + 2D y dy + y V dy
0 0
Z 1 Z 1
= D + 2D y dy + D y dy
0 0
Z 1 Z 1
= D + 2D y dy D y dy
0 0
Z 1
= D 1 + y dy. (8.124)
0
Dyy y V y = y V. (8.125)
y (y) = 1 + c1 eV (y)/D .
The periodic boundary conditions imply that (0) = (1), from which we con-
clude that
Z 1
1 + c1 eV (y)/D dy = 0.
0
Hence
Z 1
1
c1 = , b=
Z eV (y)/D dy.
Zb 0
We deduce that
1 V (y)/D
y = 1 + e .
Zb
8.6. THE UNDERDAMPED AND OVERDAMPED LIMITS OF THE DIFFUSION COEFFICIENT185
The Cauchy-Schwarz inequality shows that Z Z b > 1. Notice that in the one
dimensional case the formula for the effective diffusivity is precisely the lower
bound in (8.122). This shows that the lower bound is sharp.
Example 8.6.2. Consider the potential
a1 : y [0, 12 ],
V (y) = (8.128)
a2 : y ( 21 , 1],
In Figure 8.2 we plot the effective diffusivity given by (8.129) as a function of the
molecular diffusivity D. We observe that K decays exponentially fast in the limit
as D 0.
0
10
2
10
4
10
K
6
10
8
10
10
10
1 0 1
10 10 10
D
Figure 8.2: Effective diffusivity versus molecular diffusivity for the potential
(8.128).
where V (x) is a smooth periodic function with period L, F and D > 0 constants
and (t) standard white noise in one dimension. To simplify the notation we have
set = 1.
The stationary FokkerPlanck equation corresponding to(8.130) is
x V (x) F (x) + Dx (x) = 0, (8.131)
with periodic boundary conditions. Formula (??) for the effective drift now be-
comes Z L
Uef f = (V (x) + F )(x) dx. (8.132)
0
with
1
Z (x) := e D (V (x)F x) ,
8.6. THE UNDERDAMPED AND OVERDAMPED LIMITS OF THE DIFFUSION COEFFICIENT187
and Z Z
L x+L
Z= dx dyZ+ (y)Z (x). (8.134)
0 x
Upon using (8.133) in (8.132) we obtain [60, Ch. 9]
DL FL
Uef f = 1 e D . (8.135)
Z
Our goal now is to calculate the effective diffusion coefficient. For this we first
need to solve the Poisson equation (8.94a) which now becomes
with periodic boundary conditions. Then we need to evaluate the integrals in (??):
Z L Z L
Def f = D + (V (x) + F Uef f )(x) dx + 2D x (x)(x) dx.
0 0
It will be more convenient for the subsequent calculation to rewrite the above for-
mula for the effective diffusion coefficient in a different form. The fact that (x)
solves the stationary FokkerPlanck equation, together with elementary integra-
tions by parts yield that, for all sufficiently smooth periodic functions (x),
Z L Z L
(x)(L(x))(x) dx = D (x (x))2 (x) dx.
0 0
Now we have
Z L Z L
Def f = D+ (V (x) + F Uef f )(x)(x) dx + 2D x (x)(x) dx
0 0
Z L Z L
= D+ (L(x))(x)(x) dx + 2D x (x)(x) dx
0 0
Z L Z L
2
= D+D (x (x)) (x) dx + 2D x (x)(x) dx
0 0
Z L
= D (1 + x (x))2 (x) dx. (8.137)
0
Now we solve the Poisson equation (8.136) with periodic boundary conditions. We
multiply the equation by Z (x) and divide through by D to rewrite it in the form
Uef f
x (x (x)Z (x)) = x Z (x) + Z (x).
D
188 CHAPTER 8. THE LANGEVIN EQUATION
We integrate this equation from x L to x and use the periodicity of (x) and
V (x) together with formula (8.135) to obtain
FL
FL
L FL
Z x
x (x)Z (x) 1 e D = Z (x) 1 e D + 1 e D Z (y) dy,
Z xL
Substituting this into (8.137) and using the formula for the invariant distribution
(8.133) we finally obtain
Z L
D
Def f = 3 (I+ (x))2 I (x) dx, (8.138)
Z 0
with
Z x Z x+L
I+ (x) = Z (y)Z+ (x) dy and I (x) = Z+ (y)Z (x) dy.
xL x
Formula (8.138) for the effective diffusion coefficient (formula (22) in [61]) is the
main result of this section.
8.9 Exercises
1. Let Lb be the generator of the two-dimensional Ornstein-Uhlenbeck operator (8.17).
Calculate the eigenvalues and eigenfunctions of L. b Show that there exists a
transformation that transforms Lb into the Schrodinger operator of the two-dimensional
quantum harmonic oscillator.
Lb = 1 (c ) (c+ ) 2 (d ) (d+ ) .
3. Show that the operators a , b defined in (8.15) and (8.16) satisfy the commu-
tation relations
[a+ , a ] = 1, (8.139a)
[b+ , b ] = 1, (8.139b)
[a , b ] = 0. (8.139c)
9.1 Introduction
How long does it take for a system to switch spontaneously from one state
to another?
How is the transfer made, i.e. through what path in the relevant state space?
There is a lot of important current work on this problem by E, Vanden Eijn-
den etc.
We can separate between the 1d problem, the finite dimensional problem and the
infinite dimensional problem (SPDEs). We we will solve completely the one di-
mensional problem and discuss in some detail about the finite dimensional prob-
lem. The infinite dimensional situation is an extremely hard problem and we will
191
192CHAPTER 9. THE MEAN FIRST PASSAGE TIME AND EXIT TIME PROBLEMS
3
0 50 100 150 200 250 300 350 400 450 500
only make some remarks. The study of bistability and metastability is a very active
research area, in particular the development of numerical methods for the calcula-
tion of various quantities such as reaction rates, transition pathways etc.
We will mostly consider the dynamics of a particle moving in a bistable poten-
tial, under the influence of thermal noise in one dimension:
p
x = V (x) +
2kB T . (9.1)
1 1 1
V (x) = x4 x2 + . (9.2)
4 2 4
It is easily checked that this potential has three local minima, a local maximum at
x = 0 and two local minima at x = 1. The values of the potential at these three
points are:
1
V (1) = 0, V (0) = .
4
We will say that the height of the potential barrier is 14 . The physically (and math-
ematically!) interesting case is when the thermal fluctuations are weak when com-
pared to the potential barrier that the particle has to climb over.
9.2. BROWNIAN MOTION IN A BISTABLE POTENTIAL 193
More generally, we assume that the potential has two local minima at the points
a and c and a local maximum at b. Let us consider the problem of the escape of the
particle from the left local minimum a. The potential barrier is then defined as
E = V (b) V (a).
Our assumption that the thermal fluctuations are weak can be written as
kB T
1.
E
In this limit, it is intuitively clear that the particle is most likely to be found at either
a or c. There it will perform small oscillations around either of the local minima.
This is a result that we can obtain by studying the small temperature limit by using
perturbation theory. The result is that we can describe locally the dynamics of
the particle by appropriate OrnsteinUhlenbeck processes. Of course, this result is
valid only for finite times: at sufficiently long times the particle can escape from
the one local minimum, a say, and surmount the potential barrier to end up at c.
It will then spend a long time in the neighborhood of c until it escapes again the
potential barrier and end at a. This is an example of a rare event. The relevant
time scale, the exit time or the mean first passage time scales exponentially in
:= (kB T )1 :
= 1 exp(E).
It is more customary to calculate the reaction rate := 1 which gives the rate
with which particles escape from a local minimum of the potential:
= exp(E). (9.3)
It is very important to notice that the escape from a local minimum, i.e. a state of
local stability, can happen only at positive temperatures: it is a noise assisted event.
Indeed, consider the case T = 0. The equation of motion becomes
x = V (x), x(0) = x0 .
On the other hand, at high temperatures the particle does not see the potential
barrier: it essentially jumps freely from one local minimum to another.
To get a better understanding of the dependence of the dynamics on the depth of
the potential barrier relative to temperature, we solve the equation of motion (9.1)
numerically. In Figure we present the time series of the particle position. We
observe that at small temperatures the particle spends most of its time around x =
1 with rapid transitions from 1 to 1 and back.
Clearly, this is a random variable which is called the first passage time. The
average of this random variable is called the mean first passage time MFPT or the
first exit time:
x
(x) := ED = E inf {t > 0 : Xtx / D} X0x = x .
9.3. THE MEAN FIRST PASSAGE TIME 195
We have written the second equality in the above in order to emphasize the fact
that the mean first passage time is defined in terms of a conditional expectation, i.e.
the MFPT is defined as the expectation of the first time the diffusion processes Xt
leaves the domain, conditioned on Xt starting at x . Consequently, the MFPT
is a function of the starting point x. Consider now an ensemble of initial conditions
distributed according to a distribution p0 (x). The confinement time is defined as
Z Z
= (x)p0 (x) dx = E inf {t > 0 : Xtx / D} X0x = x p0 (x) dx.
(9.5)
We can calculate the MFPT by solving an appropriate boundary value problem.
The calculation of the confinement time follows then by calculating the integral
in [?].
Theorem 9.3.1. The MFPT is the solution of the boundary value problem
L = 1, x D, (9.6a)
= 0, x D, (9.6b)
where L is the generator of the SDE 9.6.
The homogeneous Dirichlet boundary conditions correspond to an absorbing
boundary: the particles are removed when they reach the boundary. Other choices
of boundary conditions are also possible. The rigorous proof of Theorem 9.3.1 is
based on Itos formula.
Proof. Let (X, x, t) be the probability distribution of the particles that have not
left the domain D at time t. It solves the FP equation with absorbing boundary
conditions.
= L , (X, x, 0) = (X x), |D = 0. (9.7)
t
We can write the solution to this equation in the form
(X, x, t) = eL t (X x),
where the absorbing boundary conditions are included in the definition of the semi-
group eL t . The homogeneous Dirichlet (absorbing) boundary conditions imply
that
lim (X, x, t) = 0.
t+
196CHAPTER 9. THE MEAN FIRST PASSAGE TIME AND EXIT TIME PROBLEMS
That is: all particles will eventually leave the domain. The (normalized) number of
particles that are still inside D at time t is
Z
S(x, t) = (X, x, t) dx.
D
S
= f (x, t),
t
where f (x, t) is the first passage times distribution. The MFPT is the first mo-
ment of the distribution f (x, t):
Z + Z +
dS
(x) = f (s, x)s ds = s ds
0 0 ds
Z + Z + Z
= S(s, x) ds = (X, x, s) dXds
0 0 D
Z + Z
= eL s (X x) dXds
0 D
Z + Z Z +
= (X x) eLs 1 dXds = eLs 1 ds.
0 D 0
In the case where a part of the boundary is absorbing and a part is reflecting,
then we end up with a mixed boundary value problem for the MFPT:
L = 1, x D, (9.8a)
= 0, x DA , (9.8b)
J = 0, x DR . (9.8c)
1.8
1.6
1.4
1.2
(x)
0.8
0.6
0.4
0.2
0
1 0.8 0.6 0.4 0.2 0 0.2 0.4 0.6 0.8 1
x
Figure 9.1: The mean first passage time for Brownian motion with one absorbing
and one reflecting boundary.
9.3.2 Examples
In this section we consider a few simple examples for which we can calculate the
mean first passage time in closed form.
We consider the problem of Brownian motion moving in the interval [a, b]. We
assume that the left boundary is absorbing and the right boundary is reflecting.
The boundary value problem for the MFPT time becomes
d2 d
= 1, (a) = 0, (b) = 0. (9.9)
dx2 dx
x2 a
(x) = + bx + a b .
2 2
The MFPT time for Brownian motion with one absorbing and one reflecting bound-
ary in the interval [1, 1] is plotted in Figure 9.3.2.
198CHAPTER 9. THE MEAN FIRST PASSAGE TIME AND EXIT TIME PROBLEMS
0.5
0.45
0.4
0.35
0.3
(x)
0.25
0.2
0.15
0.1
0.05
0
1 0.8 0.6 0.4 0.2 0 0.2 0.4 0.6 0.8 1
x
Figure 9.2: The mean first passage time for Brownian motion with two absorbing
boundaries.
Consider again the problem of Brownian motion moving in the interval [a, b], but
now with both boundaries being absorbing. The boundary value problem for the
MFPT time becomes
d2
= 1, (a) = 0, (b) = 0. (9.10)
dx2
The solution of this equation is
x2 a
(x) = + bx + a b .
2 2
The MFPT time for Brownian motion with two absorbing boundaries in the interval
[1, 1] is plotted in Figure 9.3.2.
Consider now the mean exit time problem from an interval [a, b] for a general one-
dimensional diffusion process with generator
d 1 d2
L = a(x) + b(x) 2 ,
dx 2 dx
9.4. ESCAPE FROM A POTENTIAL BARRIER 199
where the drift and diffusion coefficients are smooth functions and where the dif-
fusion coefficient b(x) is a strictly positive function (uniform ellipticity condition).
In order to calculate the mean first passage time we need to solve the differential
equation
d 1 d2
a(x) + b(x) 2 = 1, (9.11)
dx 2 dx
together with appropriate boundary conditions, depending on whether we have one
absorbing and one reflecting boundary or two absorbing boundaries. To solve
this equation we first define the function (x) through (x) = 2a(x)/b(x) to
write (9.11) in the form
2 (x)
e(x) (x) = e
b(x)
where the constants c1 and c2 are to be determined from the boundary conditions.
When both boundaries are absorbing we get
Z x Z z Z x
(z) e(y) 2Zb
(x) = 2 e dz dy + e(y) dy. (9.12)
a a b(y) Z a
= () exp(E)
and we calculate the escape rate = (). In particular, we analyze the depen-
dence of the escape rate on the friction coefficient. We will see that the we need to
distinguish between the cases of large and small friction coefficients.
200CHAPTER 9. THE MEAN FIRST PASSAGE TIME AND EXIT TIME PROBLEMS
We want to calculate the rate of escape from the potential barrier in this case. We
assume that the particle is initially at x0 which is near a, the left potential mini-
mum. Consider the boundary value problem for the MFPT of the one dimensional
diffusion process (9.1) from the interval (a, b):
1 eV x eV = 1 (9.14)
Now we can solve the problem of the escape from a potential well: the reflecting
boundary is at x = a, the left local minimum of the potential, and the absorbing
boundary is at x = b, the local maximum. We can replace the B.C. at x = a by a
repelling B.C. at x = :
Z b Z y
1 V (y)
(x) = dye dzeV (z) .
x
When Eb 1 the integral wrt z is dominated by the value of the potential near
a. Furthermore, we can replace the upper limit of integration by :
Z z Z +
02 2
exp(V (z)) dz exp(V (a)) exp (z a) dz
2
s
2
= exp (V (a)) ,
02
where we have used the Taylor series expansion around the minimum:
1
V (z) = V (a) + 02 (z a)2 + . . .
2
9.4. ESCAPE FROM A POTENTIAL BARRIER 201
Similarly, the integral wrt y is dominated by the value of the potential around the
saddle point. We use the Taylor series expansion
1
V (y) = V (b) b2 (y b)2 + . . .
2
Assuming that x is close to a, the minimum of the potential, we can replace the
lower limit of integration by . We finally obtain
Z b Z b
b2 2
exp(V (y)) dy exp(V (b)) exp (y b) dy
x 2
s
1 2
= exp (V (b)) .
2 b2
The reaction rate depends on the fiction coefficient and the temperature. In
the overdamped limit ( 1) we retrieve (??), appropriately rescaled with
:
0 b
= exp (Eb ) . (9.17)
2
We can also obtain a formula for the reaction rate for = O(1):
q
2 2
4 b 2 0
= exp (Eb ) . (9.18)
b 2
first eigenvalue) of the generator of the Markov process x(t) which is the solution
of p
x = V (x) + 2kB T W .
The theory of Freidlin and Wentzell has also been extended to infinite dimensional
problems. This is a very important problem in many applications such as micro-
magnetics...We refer to CITE... for more details.
A systematic study of the problem of the escape from a potential well was
developed by Matkowsky, Schuss and collaborators [67, 50, 51]. This approach
is based on a systematic use of singular perturbation theory. In particular, the
calculation of the transition rate which is uniformly valid in the friction coefficient
is presented in [51]. This formula is obtained through a careful analysis of the PDE
pq q V p + (pp + kB T p2 ) = 1,
for the mean first passage time . The PDE is equipped, of course, with the appro-
priate boundary conditions. Singular perturbation theory is used to study the small
temperature asymptotics of solutions to the boundary value problem. The formula
derived in this paper reduces to the formulas which are valid at large and small
values of the friction coefficient at the appropriate asymptotic limits.
The study of rare transition events between long lived metastable states is a
key feature in many systems in physics, chemistry and biology. Rare transition
events play an important role, for example, in the analysis of the transition between
different conformation states of biological macromolecules such as DNA [68]. The
study of rare events is one of the most active research areas in the applied stochastic
processes. Recent developments in this area involve the transition path theory of W.
E and Vanden Eijnden. Various simple applications of this theory are presented in
Metzner, Schutte et al 2006. As in the mean first passage time approach, transition
path theory is also based on the solution of an appropriate boundary value problem
for the so-called commitor function.
9.6 Exercises
204CHAPTER 9. THE MEAN FIRST PASSAGE TIME AND EXIT TIME PROBLEMS
Chapter 10
10.1 Introduction
In this final chapter we study the connection between stochastic processes and
non-equilibrium statistical mechanics. In particular, we derive stochastic equations
of evolution for a particle (or more generally, a low-dimensional deterministic
Hamiltoniandynamical system) that is in contact with a heat bath. This derivation
provides a justification for the use of stochastic differential equations in physics
and chemistry. We also develop some additional tools that are useful in the study
of systems far from equilibrium such linear response theory and projection operator
techniques.
In Section 10.2 we study the Kac-Zwanzig model and we derive the generalized
Langevin equation (GLE), together with the fluctuation-dissipation theorem . The
generalized Langevin equation is studied in Section 10.3. More general classes
of models that describe the dynamics of a particle interacting with a heat bath
are studied in Section 10.4. Linear response theory , one of the most important
techniques that are used in the study of systems far from equilibrium, is developed
in Section 10.5. Projection operator techniques, another extremely useful tool in
non-equilibrium statistical mechanics, are studied in Section 10.6. Discussion and
bibliographical remarks are included in Section 10.7. Exercises can be found in
Section 10.8.
205
206CHAPTER 10. STOCHASTIC PROCESSES AND STATISTICAL MECHANICS
{Q, P } are the coordinates of the Brownian particle and {{q}, {p}} the coordi-
nates of particles in the heat bath. The last term in the Hamiltonian function (10.1),
HI (Q, q) describes the interaction between the particle and the heat bath. The heat
bath is assumed to be in equilibrium at temperature 1 . For this, we need to pre-
pare the system appropriately, i.e. we need to assume that the initial conditions for
the particles in the heat bath are random variables that are distributed according to
an appropriate probability distribution, an appropriate Gibbs measure.
For simplicity we will restrict ourselves to the one dimensional case. We will
also consider the simplest possible model for the heat bath as well as the simplest
possible coupling between the particle and the heat bath: the heat bath will taken
to consists of N harmonic oscillators and the coupling will be taken to be linear:
X N
P2 p2n 1
H(QN , PN , q, p) = N + V (QN ) + + mn n2 qn2 n qn QN (10.2)
,
2 2mn 2
n=1
where we have introduced the subscript N in the notation for the position and mo-
mentum of the distinguished particle, QN and PN to emphasize their dependence
on the number N of the harmonic oscillators in the heat bath. V (Q) denotes the
potential experienced by the Brownian particle. For notational simplicity we have
assumed that the Brownian particle has unit mass. Notice also that we have intro-
duced a parameter that measures the strength of the coupling between the particle
and the thermal reservoir and that we have also introduced a family of constants
{n }N
n=1 .
Hamiltons equations of motion are:
N
X
N + V (QN ) =
Q n q n , (10.3a)
n=1
n
qn + n2 qn QN = 0, n = 1, . . . N. (10.3b)
mn
10.2. THE KAC-ZWANZIG MODEL 207
The equations for the particles in the harmonic heat bath are second order linear
inhomogeneous equations with constant coefficients. Our plan is to solve them and
then to substitute the result in the equations of motion for the Brownian particle.
We can solve the equations of motion for the heat bath variables using the variation
of constants formula. Set zn = (qn vn )T , vn = qn . Then equations (10.3b) can be
written as
dzn
= An zn + hN (t), (10.4)
dt
where
0 1 0
An = and F (t) = n
n2 0 mn QN (t)
The solution of (10.4) is
Z t
An t
zn (t) = e zn (0) + eAn (ts) hN (s) ds.
0
where
N
X 2n
EN = , (10.8a)
mn n2
n=1
XN
2n
RN (t) = cos(n t), (10.8b)
mn n2
n=1
XN h
2n
FN (t) = n qn (0) QN (0) cos(n t) (10.8c)
n=1
mn n2
n pn (0) i
+ sin(n t) (10.8d)
mn n
It is important to note that equation (10.7) with EN , RN (t) and FN (t) given
by (10.8a) is equivalent to the original Hamiltonian system (10.2): so far no ap-
proximation or particular assumption has been made. Notice also that the above
calculation is valid for any number of harmonic oscillators in the heat bath, even
for N = 1!
Equation (10.7) can be also written in the form
Z t
QN + Veff (QN ) = 2
RN (t s)Q N (s) ds + FN (t), (10.9)
0
initially in equilibrium in the absence of the Brownian particle or that the heat bath
is initially in equilibrium in the presence of the distinguished particle, i.e. that the
initial positions and momenta of the heat bath particles are distributed according to
a Gibbs distribution, conditional on the knowledge of {Q0 , P0 }:
where
N
" 2 #
X p2n 1 n
Heff(q, p, QN ) = + mn n2 qn QN ,2 (10.12)
n=1
2mn 2 mn n2
where the n n are mutually independent sequences of i.i.d. N (0, 1) random vari-
ables and we have used the notation kn = mn n2 . We reiterate that we actually
consider the Gibbs measure of an effective Hamiltonian. If we assume that the
p at t = 0 in the absence of the distinguished particle,
heat bath is in equilibrium
then we have qn (0) = 1 kn1 n . Our choice of the initial conditions (10.13)
ensures that the forcing term in the generalized Langevin equation that we will
derive is mean zero (see below).
Now we use (10.13) into (10.8c) to obtain
p N
X q
FN (t) = 1 n kn1 n cos(n t) + n sin(n t) . (10.14)
n=1
Equation (10.9) is called the generalized Langevin equation , FN (t) the noise
and RN (t)
XN
2n
RN (t) = cos(n t) (10.15)
k
n=1 n
2
Notice that if we add the quadratic term in Q to the Hamiltonian (10.2) then no correction to the
potential V (Q) (eqn. (10.10)) appears.
210CHAPTER 10. STOCHASTIC PROCESSES AND STATISTICAL MECHANICS
is the memory kernel. The noise and memory kernel are related. This is not sur-
Rt
prising, since the dissipation (i.e. the term 0 RN (t s)Q N (s) ds ) and the noise
FN (t) in (10.9) have the same source, namely the interaction between the Brow-
nian particle and the heat bath. In fact, the memory kernel is the autocorrelation
function of the noise (times a constant, the temperature). The following proposition
summarizes the basic properties of the noise term FN (t).
Proposition 10.2.1. The noise term FN (t) is a mean zero Gaussian stationary
process with autocorrelation function
In the writing the above equation we have used the notation hi to denote the
average with respect to the random variables {n , n }N
n=1 .
Proof. The fact that FN (t) is mean zero follows from (10.13). Gaussianity follows
from the fact that the n n are mutually independent Gaussian random variables.
Stationarity is proved in Exercise 3, Chapter 3. The proof of (10.16) follows from
the formulas hn m i = nm , hn m i = nm , hn m i = 0, n, m = 1, . . . N and a
simple trigonometric identity
N
X
hFN (t)FN (s)i = 1 2n kn1 cos(n t) cos(n s)
n=1
+ sin(n t) sin(n s)
= 1 RN (t s).
f 2 (n )
kn = ,
N 2b
10.2. THE KAC-ZWANZIG MODEL 211
where the function f (n ) decays sufficiently fast at infinity. We can rewrite the
dissipation and noise terms in the form
N
X
RN (t) = f 2 (n ) cos(n t)
n=1
and
N
X
FN (t) = f (n ) (n cos(n t) + n sin(n t)) ,
n=1
The time T > 0 if finite but arbitrary. The limiting kernel and noise satisfy the
fluctuation-dissipation theorem (10.16):
QN (t), the solution of (??) converges weakly to the solution of the limiting GLE
Z t
Q = V (Q) 2
R(t s)Q(s) ds + F (t). (10.18)
0
The properties of the limiting dissipation and noise are determined by the function
f (). As an example, consider the Lorentzian function
2/
f 2 () = (10.19)
2 + 2
with > 0. Then
R(t) = e|t| .
The noise process F (t) is a mean zero stationary Gaussian process with continuous
paths and, from (10.17), exponential correlation function:
where F (t) is the OU process (10.20). Q(t), the solution of the GLE (10.18), is not
a Markov process, i.e. the future is not statistically independent of the past, when
conditioned on the present. The stochastic process Q(t) has memory. We can
turn (10.18) into a Markovian SDE by enlarging the dimension of state space, i.e.
introducing auxiliary variables. We might have to introduce infinitely many vari-
ables! For the case of the exponential memory kernel, when the noise is given
by an OU process, it is sufficient to introduce one auxiliary variable. We can
rewrite (10.21) as a system of SDEs:
dQ
= P,
dt
dP
= V (Q) + Z,
dt
dZ p dW
= Z P + 2 1 ,
dt dt
where Z(0) N (0, 1 ).
The process {Q(t), P (t), Z(t)} R3 is Markovian.
It is a degenerate Markov process: noise acts directly only on one of the 3 degrees
of freedom.
We can eliminate the auxiliary process Z by taking an appropriate distinguished
limit.
Set = 1 , = 2 . Equations (10.23) become
dQ
= P,
dt
dP
= V (Q) + Z,
dt r
dZ 1 2 1 dW
= 2Z P+ .
dt 2 dt
10.3. THE GENERALIZED-LANGEVIN EQUATION 213
We can use tools from singular perturbation theory for Markov processes to show
that, in the limit as 0, we have that
1 p dW
Z 2 1 P.
dt
Thus, in this limit we obtain the Markovian Langevin Equation (R(t) = (t))
p dW
= V (Q) Q +
Q 2 1 . (10.24)
dt
and F (t) being a mean zero stationary Gaussian process and where R(t) and F (t)
are related through the fluctuation-dissipation theorem,
n
X q
Q = P, P = V (Q)+ j uj , u j = j uj j pj + 2j 1 , j = 1, . . . n,
j=1
(10.28)
with uj N (0, 1 ) and where Wj (t) are independent standard one dimensional
Brownian motions.
n
X
P = V (Q) +
j uj
j=1
Xn Z t
= V (Q) + j Rj (t s)P (s) ds + j (t)
j=1 0
Z t
= V (Q) R(t s)P (s) ds + F (t)
0
n
X
F (t) = j j (t),
j=1
These additional variables are solutions of a linear system of SDEs. This fol-
lows from results in approximation theory. Consider now the case where the mem-
ory kernel is a bounded analytic function. Its Laplace transform
Z +
b
R(s) = est R(t) dt
0
b 21
R(s) = 22
, i > 0, (10.29)
s + 1 + ...
b
lim R(s) = 0.
s
bN (s) = 0.
lim R
s
with
1 R(t s) = hF1 (t)F1 (s)i.
We can write (10.33) as a Markovian system for the variables {Q, P, Z1 , Z2 }:
Q = P,
P = V (Q) + Z1 (t),
Z 1 = 1 Z1 + Z2 ,
p
Z 2 = 2 Z2 P + 2 1 2 W
2.
Notice that this diffusion process is more degenerate than (10.21): noise acts
on fewer degrees of freedom. It is still, however, hypoelliptic (Hormanders con-
dition is satisfied): there is sufficient interaction between the degrees of freedom
{Q, P, Z1 , Z2 } so that noise (and hence regularity) is transferred from the de-
grees of freedom that are directly forced by noise to the ones that are not. The
corresponding Markov semigroup has nice regularizing properties. There exists a
smooth density. Stochastic processes that can be written as a Markovian process by
adding a finite number of additional variables are called quasimarkovian . Under
appropriate assumptions on the potential V (Q) the solution of the GLE equation
is an ergodic process. It is possible to study the ergodic properties of a quasi-
markovian processes by analyzing the spectral properties of the generator of the
corresponding Markov process. This leads to the analysis of the spectral properties
of hypoelliptic operators.
10.4. OPEN CLASSICAL SYSTEMS 217
(x) denotes the conjugate momentum field. The initial conditions are distributed
according to the Gibbs measure (which in this case is a Gaussian measure) at in-
verse temperature , which we formally write as
where The function (x) models the coupling between the particle and the field.
This coupling is influenced by the dipole coupling approximation from classical
electrodynamics. The Hamiltonian of the particle-field model is
with appropriate definitions for the memory kernel and the noise, which are related
through the fluctuation-dissipation theorem.
LN +1 = i{, H}.
P fN +1 = f, P fN +1 = h.
10.7. DISCUSSION AND BIBLIOGRAPHY 219
f
i = P L(f + h), (10.43a)
t
h
i = (I P )L(f + h). (10.43b)
t
We integrate the second equation and substitute into the first equation. We obtain
Z t
f
i = P Lf i P Lei(IP )Ls (I P )Lf (t s) ds + P Lei(IP )Lt h(0).
t 0
(10.44)
In the Markovian limit (large mass ratio) we obtain the Fokker-Planck equation (??).
10.8 Exercises
1. Prove (10.5). Use this formula to obtain (10.6).
220CHAPTER 10. STOCHASTIC PROCESSES AND STATISTICAL MECHANICS
Index
221
222 INDEX
random variable
Gaussian, 17
uncorrelated, 17
Reversible diffusion, 110
spectral density, 35
stationary process, 31
stationary process
second order stationary, 32
strictly stationary, 31
wide sense stationary, 32
stochastic differential equation, 43
Stochastic Process
quasimarkovian, 216
stochastic process
definition, 29
Gaussian, 30
second-order stationary, 32
stationary, 31
equivalent, 30
Bibliography
[8] A.J. Chorin and O.H. Hald. Stochastic tools in mathematics and science,
volume 1 of Surveys and Tutorials in the Applied Mathematical Sciences.
Springer, New York, 2006.
223
224 BIBLIOGRAPHY
[10] N. Wax (editor). Selected Papers on Noise and Stochastic Processes. Dover,
New York, 1954.
[12] S.N. Ethier and T.G. Kurtz. Markov processes. Wiley Series in Probability
and Mathematical Statistics: Probability and Mathematical Statistics. John
Wiley & Sons Inc., New York, 1986.
[13] L.C. Evans. Partial Differential Equations. AMS, Providence, Rhode Island,
1998.
[15] W. Feller. An introduction to probability theory and its applications. Vol. II.
Second edition. John Wiley & Sons Inc., New York, 1971.
[16] G. W. Ford and M. Kac. On the quantum Langevin equation. J. Statist. Phys.,
46(5-6):803810, 1987.
[20] M.I. Freidlin and A.D. Wentzell. Random Perturbations of dunamical sys-
tems. Springer-Verlag, New York, 1984.
[29] M. Hairer and G.A. Pavliotis. Periodic homogenization for hypoelliptic dif-
fusions. J. Statist. Phys., 117(1-2):261279, 2004.
[33] J. Jacod and A.N. Shiryaev. Limit theorems for stochastic processes, vol-
ume 288 of Grundlehren der Mathematischen Wissenschaften [Fundamental
Principles of Mathematical Sciences]. Springer-Verlag, Berlin, 2003.
[36] S. Karlin and H.M. Taylor. A first course in stochastic processes. Academic
Press [A subsidiary of Harcourt Brace Jovanovich, Publishers], New York-
London, 1975.
[38] H. A. Kramers. Brownian motion in a field of force and the diffusion model
of chemical reactions. Physica, 7:284304, 1940.
[41] A.M. Lacasta, J.M Sancho, A.H. Romero, I.M. Sokolov, and K. Lindenberg.
From subdiffusion to superdiffusion of particles on solid surfaces. Phys. Rev.
E, 70:051104, 2004.
[42] P. D. Lax. Linear algebra and its applications. Pure and Applied Mathematics
(Hoboken). Wiley-Interscience [John Wiley & Sons], Hoboken, NJ, second
edition, 2007.
[45] M. Lo`eve. Probability theory. II. Springer-Verlag, New York, fourth edition,
1978. Graduate Texts in Mathematics, Vol. 46.
[49] P. A. Markowich and C. Villani. On the trend to equilibrium for the Fokker-
Planck equation: an interplay between physics and functional analysis. Mat.
Contemp., 19:129, 2000.
[52] R.M. Mazo. Brownian motion, volume 112 of International Series of Mono-
graphs on Physics. Oxford University Press, New York, 2002.
[53] J. Meyer and J. Schroter. Comments on the Grad procedure for the Fokker-
Planck equation. J. Statist. Phys., 32(1):5369, 1983.
[56] G. A. Pavliotis and A. M. Stuart. Analysis of white noise limits for stochastic
systems with two fast relaxation times. Multiscale Model. Simul., 4(1):135
(electronic), 2005.
[59] G.A. Pavliotis and A.M. Stuart. Multiscale methods, volume 53 of Texts in
Applied Mathematics. Springer, New York, 2008. Averaging and homoge-
nization.
228 BIBLIOGRAPHY
[60] R. L. R. L. Stratonovich. Topics in the theory of random noise. Vol. II. Re-
vised English edition. Translated from the Russian by Richard A. Silverman.
Gordon and Breach Science Publishers, New York, 1967.
[61] P. Reimann, C. Van den Broeck, H. Linke, P. Hanggi, J.M. Rubi, and A. Perez-
Madrid. Diffusion in tilted periodic potentials: enhancement, universality and
scaling. Phys. Rev. E, 65(3):031104, 2002.
[62] P. Reimann, C. Van den Broeck, H. Linke, J.M. Rubi, and A. Perez-Madrid.
Giant acceleration of free diffusion by use of tilted periodic potentials. Phys.
Rev. Let., 87(1):010602, 2001.
[63] Frigyes Riesz and Bela Sz.-Nagy. Functional analysis. Dover Publications
Inc., New York, 1990. Translated from the second French edition by Leo F.
Boron, Reprint of the 1955 original.
[70] R.B. Sowers. A boundary layer theory for diffusively perturbed transport
around a heteroclinic cycle. Comm. Pure Appl. Math., 58(1):3084, 2005.
BIBLIOGRAPHY 229