APPLIED STOCHASTIC PROCESSES

G.A. Pavliotis
Department of Mathematics
Imperial College London
London SW7 2AZ, UK

January 18, 2009

2
Contents

1 Introduction 7
1.1 Historical Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.2 The One-Dimensional Random Walk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

2 Elements of Probability Theory 13

2.1 Basic Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.2 Random Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

3 Basics of the theory of Stochastic Processes 19

3.1 Definition of a Stochastic Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.2 Stationary Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.2.1 Strictly Stationary Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.2.2 Second Order Stationary Processes . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.2.3 Ergodic Properties of Stationary Processes . . . . . . . . . . . . . . . . . . . . . . 24
3.3 Brownian Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.4 Other Examples of Stochastic Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.4.1 Brownian Bridge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.4.2 Fractional Brownian Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.4.3 The Poisson Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.5 The Karhunen-Loéve Expansion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.6 The Path Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.7 Discussion and Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

4 Markov Processes 39
4.1 Introduction and Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
4.2 Definition of a Markov Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
4.3 The Chapman-Kolmogorov Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
4.4 The Generator of a Markov Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
4.5 The Adjoint Semigroup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.6 Ergodic Markov processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4.7 Stationary Markov Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

3
5 Diffusion Processes 53
5.1 Definition of a Diffusion Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
5.2 The Backward and Forward Kolmogorov Equations . . . . . . . . . . . . . . . . . . . . . . 54
5.2.1 The Backward Kolmogorov Equation . . . . . . . . . . . . . . . . . . . . . . . . . 54
5.2.2 The Forward Kolmogorov Equation . . . . . . . . . . . . . . . . . . . . . . . . . . 56
5.3 The Fokker-Planck Equation in Arbitrary Dimensions . . . . . . . . . . . . . . . . . . . . . 58
5.4 Connection with Stochastic Differential Equations . . . . . . . . . . . . . . . . . . . . . . . 59
5.5 Discussion and Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

6 The Fokker-Planck Equation 61

6.1 Basic Properties of the FP Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
6.1.1 Existence and Uniqueness of Solutions . . . . . . . . . . . . . . . . . . . . . . . . 61
6.1.2 The FP equation as a conservation law . . . . . . . . . . . . . . . . . . . . . . . . . 63
6.1.3 Boundary conditions for the Fokker–Planck equation . . . . . . . . . . . . . . . . . 63
6.2 Examples of Diffusion Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
6.2.1 Brownian Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
6.2.2 The Onrstein-Uhlenbeck Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
6.2.3 The Geometric Brownian Motin . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
6.3 The OU Process and Hermite Polynomials . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
6.4 Gradient Flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
6.5 Eigenfunction Expansions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
6.6 Self-adjointness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
6.7 Reduction to a Schrödinger Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
6.8 The Klein-Kramers-Chandrasekhar Equation . . . . . . . . . . . . . . . . . . . . . . . . . 82
6.9 Brownian Motion in a Harmonic Potential . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
6.10 Discussion and Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

7 Stochastic Differential Equations 97

7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
7.2 The Itô Stochastic Integral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
7.2.1 The Stratonovich Stochastic Integral . . . . . . . . . . . . . . . . . . . . . . . . . . 99
7.3 Existence and Uniqueness of solutions for SDEs . . . . . . . . . . . . . . . . . . . . . . . . 99
7.4 Derivation of the Stratonovich SDE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
7.5 Itô versus Stratonovich . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
7.6 The Generator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
7.7 The Itô Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
7.8 Examples of SDEs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
7.8.1 The Stochastic Landau Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
7.9 The Backward Kolmogorov Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
7.9.1 Derivation of the Fokker-Planck Equation . . . . . . . . . . . . . . . . . . . . . . . 110

4
8 The Smoluchowski and Freidlin-Wentzell Limits 113
8.1 Asymptotics for the Langevin equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
8.2 The Kramers to Smoluchowski Limit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
8.3 The Freidlin–Wentzell Limit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120

9 Exit Time Problems and Reaction Rate Theory 127

9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
9.2 Kramers’ Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
9.3 The Mean First Passage Time Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
9.3.1 The Boundary Value Problem for the MFPT . . . . . . . . . . . . . . . . . . . . . . 129
9.4 Escape from a potential barrier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
9.4.1 Calculation of the Reaction Rate in the Smoluchowski regime . . . . . . . . . . . . 131
9.4.2 The Intermediate Regime: γ = O(1) . . . . . . . . . . . . . . . . . . . . . . . . . 133
9.4.3 Calculation of the Reaction Rate in the energy-diffusion-limited regime . . . . . . . 133
9.5 Extension to Higher Dimensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
9.6 Discussion and Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

5
6
Chapter 1

Introduction

1.1 Historical Overview

• The theory of stochastic processes started with Einstein’s work on the theory of Brownian motion:
Concerning the motion, as required by the molecular-kinetic theory of heat, of particles suspended in
liquids at rest (1905).

– explanation of Brown’s observation (1827): when suspended in water, small pollen grains are
found to be in a very animated and irregular state of motion.
– Einstein’s theory is based on
∗ A Markov chain model for the motion of the particle (molecule, pollen grain...).
∗ 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 processes can be found:

– The assumption of Markovianity (no memory) expressed through the Chapman-Kolmogorov

equation.
– The Fokker–Planck equation (in this case, the diffusion equation).
– The derivation of the Fokker-Planck equation from the master (Chapman-Kolmogorov) equation
through a Kramers-Moyal expansion.
– The calculation of a transport coefficient (the diffusion equation) using macroscopic (kinetic
theory-based) considerations:
kB T
D= .
6πηa
– kB is Boltzmann’s constant, T is the temperature, η is the viscosity of the fluid and a is the
diameter of the particle.

• Einstein’s theory is based on the Fokker-Planck equation. Langevin (1908) developed a theory based
on a stochastic differential equation. The equation of motion for a Brownian particle is
d2 x dx
m 2
= −6πηa + ξ,
dt dt

7
• where ξ is a random force.

• There is complete agreement between Einstein’s theory and Langevin’s theory.

• The theory of Brownian motion was developed independently by Smoluchowski.

• The approaches of Langevin and Einstein represent the two main approaches in the theory of stochas-
tic processes:

– Study individual trajectories of Brownian particles. Their evolution is governed by a stochastic

differential equation:
dX
= F (X) + Σ(X)ξ(t),
dt

– where ξ(t) is a random force.

• Study the probability ρ(x, t) of finding a particle at position x at time t. This probability distribution
satisfies the Fokker-Planck equation:

∂ρ 1
= ∇ · (F (x)ρ) + ∇∇ : (A(x)ρ),
∂t 2

• where A(x) = Σ(x)Σ(x)T .

• The theory of stochastic processes was developed during the 20th century:

– Physics:
∗ Smoluchowksi.
∗ Planck (1917).
∗ Klein (1922).
∗ Ornstein and Uhlenbeck (1930).
∗ Kramers (1940).
∗ Chandrasekhar (1943).
∗ ...
– Mathematics:
∗ Wiener (1922).
∗ Kolmogorov (1931).
∗ Itô (1940’s).
∗ Doob (1940’s and 1950’s).
∗ ...

8
1.2 The One-Dimensional Random Walk
We let time be discrete, i.e. t = 0, 1, . . . . Consider the following stochastic process Sn :

• S0 = 0;

• at each time step it moves to ±1 with equal probability 12 .

In other words, at each time step we flip a fair coin. If the outcome is heads, we move one unit to the right.
If the outcome is tails, we move one unit to the left.
Alternatively, we can think of the random walk as a sum of independent random variables:
n
X
Sn = Xj ,
j=1

where Xj ∈ {−1, 1} with P(Xj = ±1) = 12 .

We can simulate the random walk on a computer:

• We need a (pseudo)random number generator to generate n independent random variables which

are uniformly distributed in the interval [0,1].
1
• If the value of the random variable is > 2 then the particle moves to the left, otherwise it moves to the
right.

• We then take the sum of all these random moves.

• The sequence {Sn }Nn=1 indexed by the discrete time T = {1, 2, . . . N } is the path of the random
walk. We use a linear interpolation (i.e. connect the points {n, Sn } by straight lines) to generate a
continuous path.

• Every path of the random walk is different: it depends on the outcome of a sequence of independent
random experiments.

• We can compute statistics by generating a large number of paths and computing averages. For exam-
ple, 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.

9
 50−step random walk

−2

−4

−6

0 5 10 15 20 25 30 35 40 45 50

Figure 1.1: Three paths of the random walk of length N = 50.

1000−step random walk

20

10

−10

−20

−30

−40

−50

0 100 200 300 400 500 600 700 800 900 1000

Figure 1.2: Three paths of the random walk of length N = 50.

10
 First 50 steps First 1000 steps
2 2

0 0

−2 −2
0 0.5 1 0 0.5 1
First 100 steps First 5000 steps
2 2

0 0

−2 −2
0 0.5 1 0 0.5 1
First 200 steps First 25000 steps
2 2

0 0

−2 −2
0 0.5 1 0 0.5 1

Figure 1.3: Space-time rescaled random walks.

• 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) to a Brownian
2
motion with diffusion coefficient D = ∆x 1
2∆t = 2 .

• Brownian motion W (t) is a continuous time stochastic processes with continuous paths that starts at
0 (W (0) = 0) and has independent, normally. distributed Gaussian increments.

• We can simulate the Brownian motion on a computer using a random number generator that generates
normally distributed, independent random variables.

Why introduce randomness in the description of physical systems?

• To describe outcomes of a repeated set of experiments. Think of tossing a coin repeatedly or of
throwing a dice.

• To describe a deterministic process for which we have incomplete information. Think of an ODE for
which we don’t know exactly the initial conditions (weather prediction).

• To describe a complicated deterministic system with many degrees of freedom using a simpler, low
dimensional stochastic system. Think of the physical model for Brownian motion (a heavy particle
colliding with many small particles).

• To describe a system that is inherently random. Think of quantum mechanics.

11
 2
mean of 1000 paths
5 individual paths
1.5

0.5
U(t)
0

−0.5

−1

−1.5
0 0.2 0.4 0.6 0.8 1
t

Figure 1.4: Sample Brownian paths.

12
Chapter 2

Elements of Probability Theory

2.1 Basic Definitions

Definition 2.1.1. The set of all possible outcomes of an experiment is called the sample space and is denoted
by Ω.

Example 2.1.2. • The possible outcomes of the experiment of tossing a coin are H and T . The sample

space is Ω = H, T .

• The possible outcomes of the experiment of throwing a die are 1, 2, 3, 4, 5 and 6. The sample space

is Ω = 1, 2, 3, 4, 5, 6 .

• Events are subsets of the sample space.

• We would like the unions and intersections of events to also be events.

Definition 2.1.3. A collection F of Ω is called a field on Ω if

i. ∅ ∈ F;

ii. if A ∈ F then Ac ∈ F;

iii. If A, B ∈ F then A ∪ B ∈ cF .

From the definition of a field we immediately deduce that F is closed under finite unions and finite intersec-
tions:
A1 , . . . An ∈ F ⇒ ∪ni=1 Ai ∈ F, ∩ni=1 Ai ∈ F.
When Ω is infinite dimensional then the above definition is not appropriate since we need to consider
countable unions of events.

Definition 2.1.4. A collection F of Ω is called a σ-field or σ-algebra on Ω if

i. ∅ ∈ F;

ii. if A ∈ F then Ac ∈ F;

13
 iii. If A1 , A2 , · · · ∈ F then ∪∞
i=1 Ai ∈ F.

A σ-algebra is closed under the operation of taking countable intersections.


Example 2.1.5. • F = ∅, Ω .

• F = ∅, A, Ac , Ω where A is a subset of Ω.

• The power set of Ω, denoted by {0, 1}Ω which contains all subsets of Ω.

A sub-σ–algebra is a collection of subsets of a σ–algebra which satisfies the axioms of a σ–algebra.

Definition 2.1.6. A probability measure P on the measurable space (Ω, F) is a function P : F 7→ [0, 1]
satisfying

i. P(∅) = 0, P(Ω) = 1;

ii. For A1 , A2 , . . . with Ai ∩ Aj = ∅, i 6= j then

∞
X
P(∪∞
i=1 Ai ) = P(Ai ).
i=1

Definition 2.1.7. The triple Ω, F, P comprising a set Ω, a σ-algebra F of subsets of Ω and a probability
measure P on (Ω, F) is a called a probability space.

• Assume that E|X| < ∞ and let G be a sub–σ–algebra of F. The conditional expectation of X with
respect to G is defined to be the function E[X|G] : Ω 7→ E which is G–measurable and satisfies
Z Z
E[X|G] dµ = X dµ ∀ G ∈ G.
G G

• 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.

2.2 Random Variables

Definition 2.2.1. Let (Ω, F) and (E, G) be two measurable spaces. A function X : Ω 7→ E such that the
event
{ω ∈ Ω : X(ω) ∈ A} =: {X ∈ A} (2.1)

belongs to F for arbitrary A ∈ G is called a measurable function or random variable.

When E is R equipped with its Borel σ-algebra, then (2.1) can by replaced with

{X 6 x} ∈ F ∀x ∈ R.

14
 • Let X be a random variable (measurable function) from (Ω, F, µ) to (E, G). If E is a metric space
then we may define expectation with respect to the measure µ by
Z
E[X] = X(ω) dµ(ω).
Ω

• More generally, let f : E 7→ R be G–measurable. Then,

Z
E[f (X)] = f (X(ω)) dµ(ω).
Ω

• Let U be a topological space. We will use the notation B(U ) to denote the Borel σ–algebra of U :
the smallest σ–algebra containing all open sets of U . Every random variable from a probability space
(Ω, F, µ) to a measurable space (E, B(E)) induces a probability measure on E:

µX (B) = PX −1 (B) = µ(ω ∈ Ω; X(ω) ∈ B), B ∈ B(E).

The measure µX is called the distribution (or sometimes the law) of X.

Example 2.2.2. Let I denote a subset of the positive integers. A vector ρ0 = {ρ0,i , i ∈ I} is a distribution
P
on I if it has nonnegative entries and its total mass equals 1: i∈I ρ0,i = 1.

In the case where E = R, the distribution function of the random variable X is the function F : [0, 1] 7→
[0, 1] given by
FX (x) = P(X 6 x). (2.2)
In this case, (R, B(R), FX ) becomes a probability space.
The distribution function F (x) of a random variable has the properties that limx→−∞ F (x) = 0, limx→+∞ F (x) =
1 and is right continuous.

Definition 2.2.3. A random variable X with values on R is called discrete if it takes values in some count-
able subset {x0 , x1 , x2 , . . . } of R. i.e.: P(X = x) 6= x only for x = x0 , x1 , . . . .

With a random variable we can associate the probability mass function pk = P(X = xk ). We will con-
sider nonnegative integer valued discrete random variables. In this case pk = P(X = k), k = 0, 1, 2, . . . .

Example 2.2.4. The Poisson random variable is the nonnegative integer valued random variable with prob-
ability mass function
λk −k
pk = P(X = k) = e , k = 0, 1, 2, . . . ,
k!
where λ > 0.

Example 2.2.5. The binomial random variable is the nonnegative integer valued random variable with
probability mass function
N!
pk = P(X = k) = pn q N −n k = 0, 1, 2, . . . N,
n!(N − n)!
where p ∈ (0, 1), q = 1 − p.

15
Definition 2.2.6. A random variable X with values on R is called continuous if P(X = x) = 0 ∀x ∈ R.
Let X be a random variable from (Ω, F, µ) to (Rd , B(Rd )). The (joint) distribution function FX Rd mapsto[0, 1]
is defined ass
FX (x) = P(X 6 x).
• We can use the distribution of a random variable to compute expectations and probabilities:
Z
E[f (X)] = f (x) dµX (x)
S
and Z
P[X ∈ G] = dµX (x), G ∈ B(E).
G

• When E = Rd and we can write dµX (x) = ρ(x) dx, then we refer to ρ(x) as the probability density
function (pdf), or density with respect to Lebesque measure for X.

• When E = Rd then by Lp (Ω; Rd ), or sometimes Lp (Ω; µ) or even simply Lp (µ), we mean the Banach
space of measurable functions on Ω with norm
 1/p
kXkLp = E|X|p .

• Let X be a nonnegative integer valued random variable with probability mass function pk . We can
compute the expectation of an arbitrary function of X using the formula
∞
X
E(f (X)) = f (k)pk .
k=0

Example 2.2.7. • Consider the random variable X : Ω 7→ R with pdf

 
− 12 (x − m)2
γσ,m (x) := (2πσ) exp − .
2σ
Such an X is termed a Gaussian or normal random variable. The mean is
Z
EX = xγσ,m (x) dx = m
R
and the variance is Z
2
E(X − m) = (x − m)2 γσ,m (x) dx = σ.
R

• Let m ∈ Rd and Σ ∈ Rd×d be symmetric and positive definite. The random variable X : Ω 7→ Rd
with pdf  
 − 1 1 −1
d 2
γΣ,m (x) := (2π) detΣ exp − hΣ (x − m), (x − m)i
2
is termed a multivariate Gaussian or normal random variable. The mean is
E(X) = m (2.3)
and the covariance matrix is
 
E (X − m) ⊗ (X − m) = Σ. (2.4)

16
 • Since the mean and variance specify completely a Gaussian random variable on R, the Gaussian is
commonly denoted by N (m, σ). The standard normal random variable is N (0, 1).

• Since the mean and covariance matrix completely specify a Gaussian random variable on Rd , the
Gaussian is commonly denoted by N (m, Σ).

Example 2.2.8. An exponential random variable T : Ω → R+ with rate λ > 0 satisfies

P(T > t) = e−λt , ∀t > 0.

We write T ∼ exp(λ). The related pdf is

n λe−λt , t > 0,
fT (t) = (2.5)
0, t < 0.

Notice that Z ∞ Z ∞
1 1
ET = tfT (t)dt = (λt)e−λt d(λt) = .
−∞ λ 0 λ
If the times τn = tn+1 − tn are i.i.d random variables with τ0 ∼ exp(λ) then, for t0 = 0,
n−1
X
tn = τk
k=0

and it is possible to show that

e−λt (λt)k
P(0 6 tk 6 t < tk+1 ) = . (2.6)
k!

17
18
Chapter 3

Basics of the theory of Stochastic Processes

3.1 Definition of a Stochastic Process

• Let T be an ordered set. A stochastic process is a collection of random variables X = {Xt ; t ∈ T }
where, for each fixed t ∈ T , Xt is a random variable from (Ω, F) to (E, G).

• The measurable space {Ω, F} is called the sample space. The space (E, G) is called the state space
.

• In this course we will take the set T to be [0, +∞).

• The state space E will usually be Rd equipped with the σ–algebra of Borel sets.

• A stochastic process X may be viewed as a function of both t ∈ T and ω ∈ Ω. We will sometimes

write X(t), X(t, ω) or Xt (ω) instead of Xt . For a fixed sample point ω ∈ Ω, the function Xt (ω) :
T 7→ E is called a sample path (realization, trajectory) of the process X.

• The finite dimensional distributions (fdd) of a stochastic process are the distributions of the E k –
valued random variables (X(t1 ), X(t2 ), . . . , X(tk )) for arbitrary positive integer k and arbitrary times
ti ∈ T, i ∈ {1, . . . , k}:
F (x) = P(X(ti ) 6 xi , i = 1, . . . , k)
with x = (x1 , . . . , xk ).

• We will say that two processes Xt and Yt are equivalent if they have same finite dimensional distribu-
tions.

• From experiments or numerical simulations we can only obtain information about the (fdd) of a pro-
cess.

Definition 3.1.1. A Gaussian process is a stochastic processes for which E = Rd and all the finite dimen-
sional distributions are Gaussian

F (x) = P(X(ti ) 6 xi , i = 1, . . . , k)
 
−n/2 −1/2 1 −1
= (2π) (detKk ) exp − hKk (x − µk ), x − µk i ,
2

19
for some vector µk and a symmetric positive definite matrix Kk .

• A Gaussian process x(t) is characterized by its mean

m(t) := Ex(t)
and the covariance function



C(t, s) = E x(t) − m(t) ⊗ x(s) − m(s) .

• Thus, the first two moments of a Gaussian process are sufficient for a complete characterization of the
process.

3.2 Stationary Processes

3.2.1 Strictly Stationary Processes
Definition 3.2.1. A stochastic process is called (strictly) stationary if all finite dimensional distributions are
invariant under time translation: for any integer k and times ti ∈ T , the distribution of (X(t1 ), X(t2 ), . . . , X(tk ))
is equal to that of (X(s+t1 ), X(s+t2 ), . . . , X(s+tk )) for any s such that s+ti ∈ T for all i ∈ {1, . . . , k}.
In other words,
P(Xt1 +t ∈ A1 , Xt2 +t ∈ A2 . . . Xtk +t ∈ Ak )
= P(Xt1 ∈ A1 , Xt2 ∈ A2 . . . Xtk ∈ Ak ), ∀t ∈ T.

• Let Xt be a strictly stationary stochastic process with finite second moment (i.e. Xt ∈ L2 ). The
definition of strict stationarity implies that EXt = µ, a constant, and E((Xt −µ)(Xs −µ)) = C(t−s).
Hence, a strictly stationary process with finite second moment is also stationary in the wide sense. The
converse is not true.

Example 3.2.2. 1. A sequence Y0 , Y1 , . . . of independent, identically distributed random variables with

mean µ and variance σ 2 is a stationary process with R(k) = σ 2 δk0 , σ 2 = E(Yk )2 . Let Zn = Yk , k =
0, 1, . . . . We have
N −1 N −1
1 X 1 X
Zj = Yj → EY0 = µ,
N N
j=0 j=0
by the strong law of large numbers. Hence, we can calculate the mean (as well as higher order moments)
using a single sample path of our stochastic process, provided that it is long enough (N ≫ 1). This is an
example of an ergodic stationary processes.

2. Let Z be a single random variable with mean µ and variance σ 2 and set Zj = Z, j = 0, 1, 2, . . . . Then
the sequence Z0 , Z1 , Z2 , . . . is a stationary sequence with R(k) = σ 2 . In this case we have that
N −1 N −1
1 X 1 X
Zj = Z = Z,
N N
j=0 j=0

which is independent of N and does not converge to the mean of the stochastic processes Zn . This is an
example of a non-ergodic processes.

20
 We will see in the next section that, for stationary processes, ergodicity is related to fast dacay of cor-
relations. In the first of the examples above, there was no correlation between our stochastic processes
at different times. On the contrary, in our second example there was very strong correlation between the
stochastic process at different times (in fact the stochastic process is given by the same random variable at
all times).

3.2.2 Second Order Stationary Processes

Let Ω, F, P be a probability space. Let Xt , t ∈ T (with T = R or Z) be a real-valued random process on
this probability space with finite second moment, E|Xt |2 < +∞ (i.e. Xt ∈ L2 ).

Definition 3.2.3. A stochastic process Xt ∈ L2 is called second-order stationary or wide-sense stationary

if the first moment EXt is a constant and the covariance function E(Xt − µ)(Xs − µ) depends only on the
difference t − s:
EXt = µ, E((Xt − µ)(Xs − µ)) = C(t − s).

The constant µ is the expectation of the process Xt . Without loss of generality, we can set µ = 0,
since if EXt = µ then the process Yt = Xt − µ is mean zero. A mean zero process with be called a
centered process. The function C(t) is the covariance or the autocorrelation function of the Xt . Notice
that C(t) = E(Xt X0 ), whereas C(0) = E(Xt2 ), which is finite, by assumption. Since we have assumed
that Xt is a real valued process, we have that C(t) = C(−t), t ∈ R.

Remark 3.2.4. The first two moments of a Gaussian process are sufficient for a complete characterization
of the process. A corollary of this is that a second order stationary Gaussian process is also a (strictly)
stationary process.

Continuity properties of the covariance function are equivalent to continuity properties of the paths of
Xt in the L2 sense, i.e.
lim E|Xt+h − Xt |2 = 0.
h→0

Lemma 3.2.5. Assume that the covariance function C(t) of a second order stationary process is continuous
at t = 0. Then it is continuous for all t ∈ R. Furthermore, the continuity of C(t) is equivalent to the
continuity of the process Xt in the L2 -sense.

Proof. Fix t ∈ R. We calculate:

|C(t + h) − C(t)|2 = |E(Xt+h X0 ) − E(Xt X0 )|2 = E|((Xt+h − Xt )X0 )|2

6 E(X0 )2 E(Xt+h − Xt )2
2
= C(0)(EXt+h + EXt2 − 2EXt Xt+h )
= 2C(0)(C(0) − C(h)) → 0,

as h → 0.
Assume now that C(t) is continuous. From the above calculation we have

E|Xt+h − Xt |2 = 2(C(0) − C(h)), (3.1)

21
which converges to 0 as h → 0. Conversely, assume that Xt is L2 -continuous. Then, from the above
equation we get limh→0 C(h) = C(0).

Notice that form (3.1) 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 stationary processes and Fourier analysis we will use Bochner’s theorem, which
applies to all nonnegative functions.

Definition 3.2.6. A function f (x) : R 7→ R is called nonnegative definite if

n
X
f (ti − tj )ci c̄j > 0 (3.2)
i,j=1

for all n ∈ N, t1 , . . . tn ∈ R, c1 , . . . cn ∈ C.

Lemma 3.2.7. The covariance function of second order stationary process is a nonnegative definite function.
P
Proof. We will use the notation Xtc := ni=1 Xti ci . We have.
n
X n
X
C(ti − tj )ci c̄j = EXti Xtj ci c̄j
i,j=1 i,j=1
 
Xn n
X

= E Xti ci Xtj c̄j  = E Xtc X̄tc
i=1 j=1

= E|Xtc |2 > 0.

Theorem 3.2.8. (Bochner) There is a one-to-one correspondence between the set of continuous nonnegative
definite functions and the set of finite measures on the Borel σ-algebra of R: if ρ is a finite measure, then
Z
C(t) = eixt ρ(dx) (3.3)
R

in nonnegative definite. Conversely, any nonnegative definite function can be represented in the form (3.3).

Definition 3.2.9. Let Xt be a second order stationary process with covariance C(t) whose Fourier trans-
form is the measure ρ(dx). The measure ρ(dx) is called the spectral measure of the process Xt .

In the following we will assume that the spectral measure is absolutely continuous with respect to the
Lebesgue measure on R with density f (x), dρ(x) = f (x)dx. The Fourier transform f (x) of the covariance
function is called the spectral density of the process:
Z ∞
1
f (x) = e−itx C(t) dt.
2π −∞

22
From (3.3) it follows that that the covariance 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.
−∞

There are various cases where the experimentally measured quantity is the spectral density (or power spec-
trum) of the stochastic process.

Remark 3.2.10. The correlation function of a second order stationary process enables us to associate a
time scale to Xt , the correlation time τcor :
Z ∞ Z ∞
1
τcor = C(τ ) dτ = E(Xτ X0 )/E(X02 ) dτ.
C(0) 0 0

The slower the decay of the correlation function, the larger the correlation time is. Of course, we have to
assume sufficiently fast decay of correlations so that the correlation time is finite.

Example 3.2.11. • Consider the mean zero, second order stationary process with covariance function

D −α|t|
R(t) = e . (3.4)
α

• The spectral density of this process is:

Z +∞
1 D
f (x) = e−ixt e−α|t| dt
2π α −∞
Z 0 Z +∞ 
1 D −ixt αt −ixt −αt
= e e dt + e e dt
2π α −∞ 0
 
1 D 1 1
= +
2π α ix + α −ix + α
D 1
= .
π x + α2
2

• This function is called the Cauchy or the Lorentz distribution.

• The correlation time is (we have that R(0) = D/α)

Z ∞
τcor = e−αt dt = α−1 .
0

A Gaussian process with an exponential correlation function is of particular importance in the theory
and applications of stochastic processes.

Definition 3.2.12. A Gaussian stochastic process with mean zero and covariance function (3.4) is called the
stationary Ornstein-Uhlenbeck process.

23
 The OU process was introduced by Ornstein and Uhlenbeck in 1930 (G.E. Uhlenbeck, L.S. Ornstein,
Phys. Rev. 36, 823 (1930)) as a model for the velocity of a Brownian 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, (3.5)
0

where Y (t) denotes the stationary OU process.

Lemma 3.2.13. Let Y (t) denote the stationary OU process with covariance function (3.4). Then the position
process (3.5) is a mean zero Gaussian process with covariance function

E(X(t)X(s)) = 2 min(t, s) + e− min(t,s) + e−max(t,s) − e−|t−s| − 1.

Exercise 3.2.14. Prove Lemma (3.2.13).

3.2.3 Ergodic Properties of Stationary Processes

Second order stationary processes are ergodic: time averages equal phase space (ensemble) averages. An
example of an ergodic theorem for a stationary processes is the following L2 (mean-square) ergodic theorem.

Theorem 3.2.15. Let {Xt }t>0 be a second order stationary process on a probability space Ω, F, P with
mean µ and covariance R(t), and assume that R(t) ∈ L1 (0, +∞). Then
 Z T 2
1 
lim E  X(s) ds − µ = 0. (3.6)
T →+∞ T 0
For the proof of this result we will first need an elementary lemma.

Lemma 3.2.16. Let R(t) be an integrable symmetric function. Then

Z TZ T Z T
R(t − s) dtds = 2 (T − s)R(s) ds. (3.7)
0 0 0

Proof. We make the change of variables u = t − s, v = t + s. The domain of integration in the t, s

variables is [0, T ] × [0, T ]. In the u, v variables it becomes [−T, T ] × [0, 2(T − |u|)]. The Jacobian of the
transformation is
∂(t, s) 1
J= = .
∂(u, v) 2
The integral becomes
Z T Z T Z T Z 2(T −|u|)
R(t − s) dtds = R(u)J dvdu
0 0 −T 0
Z T
= (T − |u|)R(u) du
−T
Z T
= 2 (T − u)R(u) du,
0

where the symmetry of the function R(u) was used in the last step.

24
 Proof of Theorem 3.2.15. We use Lemma (3.2.16) to calculate:
 Z T 2 Z 2
1


 1  T 

E Xs ds − µ = E (X s − µ) ds
T 0 T2  0 
Z TZ T
1
= E (R(t) − µ)(R(s) − µ) dtds
T2 0 0
Z TZ T
1
= R(t − s) dtds
T2 0 0
Z T
2
= (T − u)R(u) du
T2 0
Z Z
2 +∞  u 2 +∞
6 1− R(u) 6 R(t) dt → 0,
T 0 T T 0
using the dominated convergence theorem and the assumption R(·) ∈ L1 .
From the above calculation we conclude that
Z T Z ∞
u
lim 1− R(u) du = R(t) dt =: D,
T →+∞ 0 T 0

which is a finite quantity. Assume that µ = 0. The above calculation suggests that
Z t 2
1
lim E X(t) dt = 2D.
T →+∞ T 0

Hence, 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
stationary process Xt scales linearly in time, with proportionality coefficient 2D.
Assume that Xt is the velocity of a (Brownian) particle. In this case, the integral 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 (3.8)
0 0
is the diffusion coefficient. Thus, one expects that at sufficiently long times and under appropriate as-
sumptions 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 (3.8) 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.

25
Example 3.2.17. Consider the stochastic processes with an exponential correlation function from Exam-
ple 3.2.11, and assume that this stochastic process describes the velocity of a Brownian particle. Since
R(t) ∈ L1 (0, +∞) Theorem 3.2.15 applies. Furthermore, the diffusion coefficient of the Brownian particle
is given by Z +∞
D
R(t) dt = R(0)τc−1 = 2 .
0 α

3.3 Brownian Motion

• The most important continuous time stochastic process is Brownian motion. Brownian motion is
a mean zero, continuous (i.e. it has continuous sample paths: for a.e ω ∈ Ω the function Xt is a
continuous function of time) process with independent Gaussian increments.

• A process Xt has independent increments if for every sequence t0 < t1 < . . . tn the random
variables
Xt1 − Xt0 , Xt2 − Xt1 , . . . , Xtn − Xtn−1
are independent.

• If, furthermore, for any t1 , t2 , s ∈ T and Borel set B ⊂ R

P(Xt2 +s − Xt1 +s ∈ B) = P(Xt2 − Xt1 ∈ B)

then the process Xt has stationary independent increments.

Definition 3.3.1. • A one dimensional standard Brownian motion W (t) : R+ → R is a real valued
stochastic process such that

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) = 2π(t − s) exp − ; (3.9)
2(t − s)

• A d–dimensional standard Brownian motion W (t) : R+ → Rd is a collection of d independent one

dimensional Brownian motions:

W (t) = (W1 (t), . . . , Wd (t)),

where Wi (t), i = 1, . . . , d are independent one dimensional Brownian motions. The density of the
Gaussian random vector W (t) − W (s) is thus
 −d/2  
kxk2
g(x; t, s) = 2π(t − s) exp − .
2(t − s)

26
 2
mean of 1000 paths
5 individual paths
1.5

0.5
U(t)
0

−0.5

−1

−1.5
0 0.2 0.4 0.6 0.8 1
t

Figure 3.1: Brownian sample paths

Brownian motion is sometimes referred to as the Wiener process .

Brownian motion has continuous paths. More precisely:

Definition 3.3.2. Let Xt and Yt , t ∈ T , be two stochastic processes defined on the same probability space
{Ω, F, P}. The process Yt is said to be a modification of Xt if P(Xt = Yt ) = 1 ∀t ∈ T . Then there is a
modification of Brownian motion such that all paths are continuous.

Lemma 3.3.3. There is a continuous modification of Brownian motion.

This follows from a theorem due to Kolmogorov.

Theorem 3.3.4. (Kolmogorov) Let Xt , t ∈ [0, ∞) be a stochastic process on a probability space {Ω, F, P}.
Suppose that there are positive constants α and β, and for each T > 0 there is a constant C(T ) such that

E|Xt − Xs |α 6 C(T )|t − s|1+β , 0 6 s, t 6 T. (3.10)

There there exists a continuous modification Yt of the process Xt .

Exercise 3.3.5. Use Theorem (3.3.4) to show that Brownian motion has a continuous modification.

Remark 3.3.6. Equivalently, we could have defined the one dimensional standard Brownian motion as
a stochastic process on a probability space {Ω, F, P} with continuous paths for almost all ω ∈ Ω, and
Gaussian finite dimensional distributions with zero mean and covariance E(Wti Wtj ) = min(ti , tj ). One
can then show that Definition 3.3.1 and continuity of paths are equivalent to the above definition.

It is possible to prove rigorously the existence of the Wiener process (Brownian motion):

Theorem 3.3.7. (Wiener) There exists an almost-surely continuous process Wt with independent increments
such and W0 = 0, such that for each t > the random variable Wt is N (0, t). Furthermore, Wt is almost
surely locally Hölder continuous with exponent α for any α ∈ (0, 12 ).

27
 Notice that Brownian paths are not differentiable.
We can also construct Brownian motion through the limit of the random walk: let X1 , X2 , . . . be iid
random variables on a probability space {Ω, F, P} with mean 0 and variance 1. Define the discrete time
P n
stochastic process Sn with S0 = 0, Sn = j=1 Xj , n > n. Define not a continuous time stochastic
process with continuous paths as the linearly interpolated, appropriately rescaled random walk:
1 1
Wtn = √ S[nt] + (nt − [nt]) √ X[nt]+1 ,
n n

where [·] denotes the integer part of a number. Then Wtn converges weakly to a one dimensional standard
Brownian motion.
Brownian motion is a Gaussian process. For the d–dimensional Brownian motion, and for I the d × d
dimensional identity, we have (see (2.3) and (2.4))

EW (t) = 0 ∀t > 0

and  
E (W (t) − W (s)) ⊗ (W (t) − W (s)) = (t − s)I. (3.11)

Moreover,  
E W (t) ⊗ W (s) = min(t, s)I. (3.12)

From the formula for the Gaussian density g(x, t−s), eqn. (3.9), we immediately 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 .
2πt
We can easily calculate all moments of the Brownian motion:
Z +∞
n 1 2
E(x (t)) = √ xn e−x /2t dx
2πt −∞
n
1.3 . . . (n − 1)tn/2 , n even,
=
0, n odd.

Exercise 3.3.8. Let a1 , . . . an and s1 , . . . sn be positive real numbers. Calculate the mean and variance of
the random variable
X n
X= ai W (si ).
i=1

Exercise 3.3.9. Let W (t) be the standard one-dimensional Brownian motion and let σ, s1 , s2 > 0. Calcu-
late

i. EeσW (t) .


ii. E sin(σW (s1 )) sin(σW (s2 )) .

28
 Notice that if Wt is a standard Brownian motion, then so is the rescaled processes
1
Xt = √ Wct .
c

To prove this we show that the two processes have the same distribution function:
√

P (Xt 6 x) = P Wct 6 cx
Z √cx  
1 z2
= √ exp − dz
−∞ 2πct 2ct
Z x  2
1 y √
= √ exp − c dy
−∞ 2πct 2t
= P (Wt 6 x) .

We can also add a drift and change the diffusion coefficient of the Brownian motion: we will define a
Brownian motion with drift µ and variance σ 2 as the process

Xt = µt + σWt .

Notice that
EXt = µt, E(Xt − EXt )2 = σ 2 t.

We can define the OU process through the Brownian motion via a time change.

Lemma 3.3.10. Let W (t) be a standard Brownian motion and consider the process

V (t) = e−t W (e2t ).

Then V (t) is a Gaussian second order stationary process with mean 0 and covariance

R(s, t) = e−|t−s| . (3.13)

For the proof of this result we first need to show that time changed Gaussian processes are also Gaussian.

Lemma 3.3.11. 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
{Y (t1 ), Y (t2 ), . . . Y (tN )} (3.14)

is has is a multivariate Gaussian random variable. Since f (t) is strictly increasing, it is invertible and hence,
there exist si , i = 1, . . . N such that si = f −1 (ti ). Thus, the random vector (3.14) can be rewritten as

{X(s1 ), X(s2 ), . . . X(sN )},

which is Gaussian for all N and all choices of times s1 , s2 , . . . sN . Hence Y (t) is also Gaussian.

29
 Proof of Lemma 3.3.10. 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 (3.13), we calculate

E(V (t)V (s)) = e−t−s E(W (e2t )W (e2s )) = e−t−s min(e2t , e2s )
= e−|t−s| .

The Gaussianity of the V (t) follows from Lemma 3.3.11 (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))).

Exercise 3.3.12. Use Lemma 3.3.10 to obtain the probability density function of the stationary Ornstein-
Uhlenbeck process.

Exercise 3.3.13. Calculate the mean and the correlation function of the integral of a standard Brownian
motion Z t
Yt = Ws ds.
0

Exercise 3.3.14. Show that the process

Z t+1
Yt = (Ws − Wt ) ds, t ∈ R,
t

is second order stationary.

3.4 Other Examples of Stochastic Processes

3.4.1 Brownian Bridge
Let W (t) be a standard one dimensional Brownian motion. We define the Brownian bridge (from 0 to 0)
to be the process
Bt = Wt − tW1 , t ∈ [0, 1]. (3.15)
Notice that B0 = B1 = 0. Equivalently, we can define the Brownian bridge to be the continuous Gaussian
process {Bt : 0 6 t 6 1} such that

EBt = 0, E((Bt − EBt )(Bs − EBs )) = min(s, t) − st, s, t ∈ [0, 1]. (3.16)

Another, equivalent definition of the Brownian bridge is through an appropriate time change of the Brownian
motion:  
t
Bt = t(1 − t)W , t ∈ [0, 1). (3.17)
1−t
Conversely, we can write the Brownian motion as a time change of the Brownian bridge:
 
t
Wt = (t + 1)B , t > 0.
1+t

Exercise 3.4.1. Show the equivalence between definitions (3.15), (3.16) and (3.17).

30
Exercise 3.4.2. Use (3.15) to show that the probability density function of the Brownian bridge is
 
1/2 |x − 1|2
f (x, t) = (2π(1 − t)) exp − .
2(1 − t)

Exercise 3.4.3. Give the definition of the Brownian Bridge from 0 to a, where a ∈ R. Calculate the mean
and covariance of this Brownian bridge.

3.4.2 Fractional Brownian Motion

Definition 3.4.4. A (normalized) fractional Brownian motion WtH , t > 0 with Hurst parameter H ∈ (0, 1)
is a centered Gaussian process with continuous sample paths whose covariance is given by

1 2H

E(WtH WsH ) = s + t2H − |t − s|2H . (3.18)
2
Fractional Brownian motion has the following properties.
1
i. When H = 12 , Wt2 becomes the standard Brownian motion.

ii. W0H = 0, EWtH = 0, E(WtH )2 = |t|2H , t > 0.

iii. It has stationary increments, E(WtH − WsH )2 = |t − s|2H .

iv. It has the following self similarity property

H
(Wαt , t > 0) = (αH WtH , t > 0), α > 0,

where the equivalence is in law.

3.4.3 The Poisson Process

• Another fundamental continuous time process is the Poisson process :

Definition 3.4.5. The Poisson process with intensity λ, denoted by N (t), is an integer-valued, contin-
uous time, stochastic process with independent increments satisfying

k
e−λ(t−s) λ(t − s)
P[(N (t) − N (s)) = k] = , t > s > 0, k ∈ N.
k!

• Both Brownian motion and the Poisson process are homogeneous (or time-homogeneous): the in-
crements between successive times s and t depend only on t − s.

Exercise 3.4.6. Use Theorem (3.3.4) to show that there does not exist a continuous modification of the
Poisson process.

31
3.5 The Karhunen-Loéve Expansion
Let f ∈ L2 (Ω) where Ω is a subset of Rd and let {en }∞ 2
n=1 be an orthonormal basis in L (Ω). Then, it is
well known that f can be written as a series expansion:
∞
X
f= fn en ,
n=1

where Z
fn = f (x)en (x) dx.
Ω

The convergence is in L2 (Ω):

N
X
lim kf (x) − fn en (x)kL2 (Ω) = 0.
N →∞
n=1

It turns out that we can obtain a similar expansion for an L2 mean zero process which is continuous in the
L2 sense:
EXt2 < +∞, EXt = 0, lim E|Xt+h − Xt |2 = 0. (3.19)
h→0

For simplicity we will take T = [0, 1]. Let R(t, s) = E(Xt Xs ) be the correlation function.

Exercise 3.5.1. Show that the correlation function of a process Xt satisfying (3.19) is continuous in both t
and s.

Let us assume that an expansion of the form

∞
X
Xt (ω) = ξn (ω)en (t), t ∈ [0, 1] (3.20)
n=1

where {en }∞ 2
n=1 is an orthonormal basis in L ([0, 1]). The random variables ξn are calculated as

Z 1 Z ∞
1X
Xt ek (t) dt = ξn en (t)ek (t) dt
0 0 n=1
∞
X
= ξn δnk = ξk ,
n=1

where we assumed that we can interchange the summation and integration. We will assume that these
random variables are orthogonal:
E(ξn ξm ) = λn δnm ,

where {λn }∞
n=1 are positive numbers that will be determined later.

32
 Assuming that an expansion of the form (3.20) exists, we can calculate
∞ X
∞
!
X
R(t, s) = E(Xt Xs ) = E ξk ek (t)ξℓ eℓ (s)
k=1 ℓ=1
∞
XX ∞
= E (ξk ξℓ ) ek (t)eℓ (s)
k=1 ℓ=1
∞
X
= λk ek (t)ek (s).
k=1

Consequently, in order to the expansion (3.20) to be valid we need

∞
X
R(t, s) = λk ek (t)ek (s). (3.21)
k=1

From equation (3.21) it follows that

Z 1 Z ∞
1X
R(t, s)en (s) ds = λk ek (t)ek (s)en (s) ds
0 0 k=1
∞
X Z 1
= λk ek (t) ek (s)en (s) ds
k=1 0

X∞
= λk ek (t)δkn
k=1
= λn en (t).

Hence, in order for the expansion (3.20) to be valid, {λn , en (t)}∞

n=1 have to be the eigenvalues and eigen-
functions of the integral operator whose kernel is the correlation function of Xt :
Z 1
R(t, s)en (s) ds = λn en (t). (3.22)
0

Hence, in order to prove the expansion (3.20) 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 nonnegative (Rf, f > 0 for all f ∈ L2 (0, 1)). Hence, all its eigenvalues are real
and nonnegative. Furthermore, it is a compact operator (if {φn }∞ 2
n=1 is a bounded sequence in L (0, 1), then
{Rφn }∞ n=1 has a weakly convergent subsequence). The spectral 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 Rf0 = 0, {en (t)} are the eigenfunctions of R corresponding to nonzero eigenvalues and the con-
vergence is in L2 . Finally, Spencer’s Theorem states that for R(t, s) continuous on [0, 1] × [0, 1], the
expansion (3.21) is valid, where the series converges absolutely and uniformly.

33
Exercise 3.5.2. Let Xt be a stochastic process satisfying (3.19) and R(t, s) its correlation function. Show
that the integral operator R : L2 [0, 1] 7→ L2 [0, 1]
Z 1
Rf := R(t, s)f (s) ds (3.23)
0

is self-adjoint and nonnegative. Show that all of its eigenvalues are real and nonnegative. Show that eigen-
functions corresponding to different eigenvalues are orthogonal.

Exercise 3.5.3. Let H be a Hilbert space. An operator R : H → H is said to be Hilbert-Schmidt if there

exists a complete orthonormal sequence {φn }∞
n=1 in H such that

∞
X
kRen k2 < ∞.
n=1

Let R : L2 [0, 1] 7→ L2 [0, 1] be the operator defined in (3.23) with R(t, s) being continuous both in t and s.
Show that it is a Hilbert-Schmidt operator.

Now we are ready to prove (3.20).

Theorem 3.5.4. (Karhunen-Loéve). Let {Xt , t ∈ [0, 1]} be an L2 process with zero mean and continuous
correlation function R(t, s). Let {λn , en (t)}∞
n=1 be the eigenvalues and eigenfunctions of the operator R
defined in (3.23). Then
X∞
Xt = ξn en (t), t ∈ [0, 1], (3.24)
n=1

where
Z 1
ξn = Xt en (t) dt, Eξn = 0, E(ξn ξm ) = λδnm . (3.25)
0

The series converges in L2 to X(t), uniformly in t.

Proof. The fact that Eξn = 0 follows from the fact that Xt is mean zero. The orthogonality of the random
variables {ξn }∞
n=1 follows from the orthogonality of the eigenfunctions of R:

Z 1Z 1
E(ξn ξm ) = 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
= λn en (s)em (s) ds
0
= λn δnm .

34
 PN
Consider now the partial sum SN = n=1 ξn en (t).

E|Xt − SN |2 = EXt2 + ESN

2
− 2E(Xt SN )
N N
!
X X
= R(t, t) + E ξk ξℓ ek (t)eℓ (t) − 2E Xt ξn en (t)
k,ℓ=1 n=1
N
X N Z
X 1
= R(t, t) + λk |ek (t)|2 − 2E Xt Xs ek (s)ek (t) ds
k=1 k=1 0
N
X
= R(t, t) − λk |ek (t)|2 → 0,
k=1

by Spencer’s theorem.

Exercise 3.5.5. Let Xt a mean zero second order stationary process with continuous covariance R(t). Show
that

∞
X
λn = T R(0).
n=1

Remark 3.5.6. Let Xt be a Gaussian second order process with continuous covariance R(t, s). Then the
random variables {ξk }∞
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-Loéve expansion becomes:
+∞ p
X
Xt = λk ξk ek (t), (3.26)
k=1

where {ξk }∞
k=1 are independent N (0, 1) random variables.

Example 3.5.7. The Karhunen-Loéve Expansion for Brownian Motion. The correlation function of
Brownian motion is R(t, s) = min(t, s). The eigenvalue problem Rψn = λn ψn becomes
Z 1
min(t, s)ψn (s) ds = λn ψn (t).
0

Let us assume that λn > 0 (it is easy to check that 0 is not an eigenvalue). Upon setting t = 0 we obtain
ψn (0) = 0. The eigenvalue problem can be rewritten in the form
Z t Z 1
sψn (s) ds + t ψn (s) ds = λn ψn (t).
0 t

We differentiate this equation once:

Z 1
ψn (s) ds = λn ψn′ (t).
t

35
We set t = 1 in this equation to obtain the second boundary condition ψn′ (1) = 0. A second differentiation
yields;
−ψn (t) = λn ψn′′ (t),
where primes denote differentiation with respect to t. Thus, in order to calculate the eigenvalues and eigen-
functions of the integral operator whose kernel is the covariance function of Brownian motion, we need to
solve the Sturm-Liouville problem

−ψn (t) = λn ψn′′ (t), ψ(0) = ψ ′ (1) = 0.

It is easy to check that the eigenvalues and (normalized) eigenfunctions are

   2
√ 1 2
ψn (t) = 2 sin (2n + 1)πt , λn = .
2 (2n + 1)π

Thus, the Karhunen-Loéve expansion of Brownian motion on [0, 1] is

∞  
√ X 2 1
Wt = 2 ξn sin (2n + 1)πt . (3.27)
n=1
(2n + 1)π 2

Exercise 3.5.8. Calculate the Karhunen-Loeve expansion for a second order stochastic process with corre-
lation function R(t, s) = ts.

Exercise 3.5.9. Calculate the Karhunen-Loeve expansion of the Brownian bridge on [0, 1].

Exercise 3.5.10. Let Xt , t ∈ [0, T ] be a second order process with continuous covariance and Karhunen-
Loéve expansion
X∞
Xt = ξk ek (t).
k=1
Define the process
Y (t) = f (t)Xτ (t) , t ∈ [0, S],
where f (t) is a continuous function and τ (t) a continuous, nondecreasing function with τ (0) = 0, τ (S) =
T . Find the Karhunen-Loéve expansion of Y (t), in terms of the KL expansion of Xt . Use this in order to
calculate the KL expansion of the Ornstein-Uhlenbeck process.

Exercise 3.5.11. Calculate the Karhunen-Loéve expansion of a centered Gaussian stochastic process with
covariance function R(s, t) = cos(2π(t − s)).

3.6 The Path Space

• Let (Ω, F, µ) be a probability space, (E, ρ) a metric space and let T = [0, ∞). Let {Xt } be a
stochastic process from (Ω, F, µ) to (E, ρ) with continuous sample paths.

• The above means that for every ω ∈ Ω we have that Xt ∈ CE := C([0, ∞); E).

• The space of continuous functions CE is called the path space of the stochastic process.

36
 • We can put a metric on E as follows:
∞
X
1 2 1

ρE (X , X ) := n
max min ρ(Xt1 , Xt2 ), 1 .
2 06t6n
n=1

• We can then define the Borel sets on CE , using the topology induced by this metric, and {Xt } can be
thought of as a random variable on (Ω, F, µ) with state space (CE , B(CE )).

• The probability measure PXt−1 on (CE , B(CE )) is called the law of {Xt }.

• The law of a stochastic process is a probability measure on its path space.

Example 3.6.1. The space of continuous functions CE is the path space of Brownian motion (the
Wiener process). The law of Brownian motion, that is the measure that it induces on C([0, ∞), Rd ),
is known as the Wiener measure.

3.7 Discussion and Bibliography

The kind of analysis presented in Section 3.2.3 was initiated by G.I. Taylor in [25]. The spectral theorem
for compact, self-adjoint operators which was needed in the proof of the Karhunen-Loéve theorem can be
found in [19]. A similar result is valid for random fields. See [23] and the reference therein.

37
38
Chapter 4

Markov Processes

4.1 Introduction and Examples

In this chapter we will study some of the basic properties of Markov stochastic processes. Roughly speaking,
a Markov process is a stochastic process that retains no memory 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 ξi , i = 1, . . . :
XN
XN = ξn , X0 = 0.
n=1

Let i1 , . . . i2 , . . . be a sequence of integers. Then, clearly

P(Xn+1 = in+1 |X1 = i1 , . . . Xn = in ) = P(Xn+1 = in+1 |Xn = in ) (4.1)

In words, the probability that the random walk will be at in+1 at time n + 1 depends only on its current value
and not on how it got there.
Consider the case where the ξn ’s are independent Bernoulli random variables with Eξn = ±1 = 12 .
Then, P(Xn+1 = in+1 |X1 = i1 , . . . Xn = in ) = 12 if in+1 = in ± 1 and 0 otherwise.
The random walk is an example of a discrete time Markov chain:

Definition 4.1.1. A stochastic process {Sn ; n ∈ N} and state space is S = Z is called a discrete time
Markov chain provided that the Markov property (4.1) is satisfied.

Consider now a continuous-time stochastic process Xt with state space S = Z and denote by {Xs , s 6
t} the collection of values of the stochastic process up to time t. We will say that Xt is a Markov processes
provided that
P(Xt+h = it+h |{Xs , s 6 t}) = P(Xt+h = it+h |Xt = it ), (4.2)
for all h > 0. A continuous-time, discrete state space Markov process is called a continuous-time Markov
chain.

39
Example 4.1.2. The Poisson process is a continuous-time Markov chain with
n 0 if j < i,
P(Nt+h = j|Nt = i) = e−λs (λs)j−i
(j−i)! , if j > i.

Similarly, we can define a Markov process for a continuous-time Markov process whose state space is
R. In this case, the above definitions become

P(Xt+h ∈ Γ|{Xs , s 6 t}) = P(Xt+h ∈ Γ|Xt = x) (4.3)

for all Borel sets Γ.

Example 4.1.3. The Brownian motion is a Markov process with conditional probability density
 
1 |x − y|2
p(y, t|x, s) := p(Wt = y|Ws = x) = p exp − . (4.4)
2π(t − s) 2(t − s)

Example 4.1.4. The Ornstein-Uhlenbeck process Vt = e−t W (e2t ) is a Markov process with conditional
probability density
!
1 |y − xe−(t−s) |2
p(y, t|x, s) := p(Vt = y|Vs = x) = p exp − . (4.5)
2π(1 − e−2(t−s) ) 2(1 − e−2(t−s) )

To prove (4.5) we use the formula for the distribution function of the Brownian motion to calculate, for
t > s,

P(Vt 6 y|Vs = x) = P(e−t W (e2t ) 6 y|e−s W (e2s ) = x)

= P(W (e2t ) 6 et y|W (e2s ) = es x)
Z et y t s
1 − |ze −xe |
= p e 2(e2t −e2s ) dz
−∞ 2π(e2t − e2s )
Z y q |ρet −xes |
−
= 2πe2t (1 − e−2(t−s) )e 2(e2t (1−e−2(t−s) ) dρ
−∞
Z y |ρ−x|
1 −
= p e 2(1−e−2(t−s) ) dρ.
−∞ 2π(1 − e−2(t−s) )
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−(t−s) |2
= p exp − .
2π(1 − e−2(t−s) ) 2(1 − e−2(t−s) )

Markov stochastic processes appear in a variety of applications in physics, chemistry, biology and fi-
nance. In this and the next chapter we will develop various analytical tools for studying them. In particular,
we will see that we can obtain an equation for the transition probability

P(Xn+1 = in+1 |Xn = in ), P(Xt+h = it+h |Xt = it ), p(Xt + h = y|Xt = x), (4.6)

40
which will enable us to study the evolution of a Markov process. This equation will be called the Chapman-
Kolmogorov equation.
We will be mostly concerned with time-homogeneous Markov processes, i.e. processes for which the
conditional probabilities are invariant under time shifts. For time-homogeneous discrete-time Markov chains
we have
P(Xn+1 = j|Xn = i) = P(X1 = j|X0 = i) =: pij .
We will refer to the matrix P = {pij } as the transition matrix. It is each to check that the transition matrix
P
is a stochastic matrix, i.e. it has nonnegative entries and 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 µi := P(Xn = i). The (possibly infinite dimensional) vector µn 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 µ(n)
µ(n) = µ(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 µ0 and the transition matrix P . Componentwise, the above
equation can be written as X (0)
(n)
µj = µi πij (n).
i
Consider now a continuous time Markov chain with transition probability

pij (s, t) = P(Xt = j|Xs = i), s 6 t.

If the chain is homogeneous, then

pij (s, t) = pij (0, t − s) for all i, j, s, t.

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

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

41
The generator of the Markov chain is defined as
1
G = lim (Ph − I).
h→0 h
Let now µit = P(Xt = i). The vector µt is the distribution of the Markov chain at time t. We can study its
evolution using the equation
µt = µ0 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 contin-
uous paths. As we have seen in Example 4.1.3 the Brownian motion is an example of such a process. It is
a standard result in the theory of partial 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) = δ(y − x). (4.10)
∂t 2 ∂y 2 t→s

Exercise 4.1.5. 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) = δ(y − x).
∂s 2 ∂x2 s→t

Similarly, the conditional distribution of the OU process satisfies the initial value problem
∂p ∂(yp) ∂ 2 p
= + 2, lim p(y, t|x, s) = δ(y − x). (4.11)
∂t ∂y ∂y t→s

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) = δ(y − x). (4.12)
∂t ∂y 2 ∂y 2 t→s

as well as the backward Kolmogorov equation

∂p ∂p 1 ∂2p
− = −a(x, s) + b(x, s) 2 , lim p(y, t|x, s) = δ(y − x). (4.13)
∂s ∂x 2 ∂x t→s

for appropriate functions a(y, t), b(y, t). Hence, a diffusion process is determined uniquely from these two
functions.

Exercise 4.1.6. 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

42
4.2 Definition of a Markov Process
In Section 4.1 we gave the definition of Markov process whose time is either discrete or continuous, and
whose state space is the set of integers. We also gave several examples of Markov chains as well as of
processes whose state space is the real line. In this section we give the precise definition of a Markov
process with t ∈ T , a general index set and S = E, an arbitrary metric space. We will use this formulation
in the next section to derive the Chapman-Kolmogorov equation.
In order to state the definition of a continuous-time Markov process that takes values in a metric space
we need to introduce various new concepts. For the definition of a Markov process we need to use the
conditional expectation of the stochastic process conditioned on all past values. We can encode all past
information about a stochastic process into an appropriate collection of σ-algebras. Our setting will be that
we have a probability space (Ω, F, P) and an ordered set T . Let X = Xt (ω) be a stochastic process from
the sample space (Ω, F) to the state space (E, G), where E is a metric space (we will usually take E to be
either R or Rd ). Remember that the stochastic process is a function of two variables, t ∈ T and ω ∈ Ω.
We start with the definition of a σ–algebra generated by a collection of sets.

Definition 4.2.1. Let K be a collection of subsets of Ω. The smallest σ–algebra on Ω which contains K is
denoted by σ(K) and is called the σ–algebra generated by K.

Definition 4.2.2. Let Xt : Ω 7→ E, t ∈ T . The smallest σ–algebra σ(Xt , t ∈ T ), such that the family of
mappings {Xt , t ∈ T } is a stochastic process with sample space (Ω, σ(Xt , t ∈ T )) and state space (E, G),
is called the σ–algebra generated by {Xt , t ∈ T }.

In other words, the σ–algebra generated by Xt is the smallest σ–algebra such that Xt is a measurable
function (random variable) with respect to it: the set


ω ∈ Ω : Xt (ω) 6 x ∈ σ(Xt , t ∈ T )

for all x ∈ R (we have assumed that E = X).

Definition 4.2.3. A filtration on (Ω, F) is a nondecreasing family {Ft , t ∈ T } of sub–σ–algebras of F:

Fs ⊆ Ft ⊆ F for s 6 t.

We set F∞ = σ(∪t∈T Ft ). The filtration generated by Xt , where Xt is a stochastic process, is

FtX := σ (Xs ; s 6 t) .

Definition 4.2.4. A stochastic process {Xt ; t ∈ T } is adapted to the filtration {Ft } := {Ft , t ∈ T } if for
all t ∈ T , Xt is an Ft –measurable random variable.

Definition 4.2.5. Let {Xt } be a stochastic process defined on a probability space (Ω, F, µ) with values in
E and let FtX be the filtration generated by {Xt }. Then {Xt } is a Markov process if

P(Xt ∈ Γ|FsX ) = P(Xt ∈ Γ|Xs ) (4.14)

for all t, s ∈ T with t > s, and Γ ∈ B(E).

43
Remark 4.2.6. The filtration FtX is generated by events of the form {ω|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

P(Xt ∈ Γ|Xt1 , Xt2 , . . . Xtn ) = P(Xt ∈ Γ|Xtn ) a.s.

for n > 1, 0 6 t1 < t2 < · · · < tn 6 t and Γ ∈ B(E).

Roughly speaking, the statistics of Xt for t > s are completely determined once Xs is known; informa-
tion about Xt for t < s is superfluous. In other words: a Markov process has no memory. More precisely:
when a Markov process is conditioned on the present state, then there is no memory of the past. The past
and future of a Markov process are statistically independent when the present is known.

Remark 4.2.7. A non Markovian process Xt can be described through a Markovian one Yt by enlarging
the state space: the additional variables that we introduce account for the memory in the Xt . This ”Marko-
vianization” trick is very useful since there are many more tools for analyzing Markovian process.

Example 4.2.8. The velocity of a Brownian particle is modeled by the stationary Ornstein-Uhlenbeck pro-
cess Yt = e−t W (e2t ). The particle position is given by the integral of the OU process (we take X0 = 0)
Z t
Xt = Ys ds.
0

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

4.3 The Chapman-Kolmogorov Equation

With a Markov process {Xt } we can associate a function P : T × T × E × B(E) → R+ defined through
the relation
 
P Xt ∈ Γ|FsX = P (s, t, Xs , Γ),
 
for all t, s ∈ T with t > s and all Γ ∈ B(E). Assume that Xs = x. Since P Xt ∈ Γ|FsX = P [Xt ∈ Γ|Xs ]
we can write
P (Γ, t|x, s) = P [Xt ∈ Γ|Xs = x] .
The transition function P (t, Γ|x, s) is (for fixed t, x s) a probability measure on E with P (t, E|x, s) = 1;
it is B(E)–measurable in x (for fixed t, s, Γ) and satisfies the Chapman–Kolmogorov equation
Z
P (Γ, t|x, s) = P (Γ, t|y, u)P (dy, u|x, s). (4.15)
E

for all x ∈ E, Γ ∈ B(E) and s, u, t ∈ T with s 6 u 6 t. The derivation of the Chapman-Kolmogorov

equation is based on the assumption of Markovianity and on properties of the conditional probability. Let

44
(Ω, F, µ) be a probability space, X a random variable from (Ω, F, µ) to (E, G) and let F1 ⊂ F2 ⊂ F.
Then
E(E(X|F2 )|F1 ) = E(E(X|F1 )|F2 ) = E(X|F1 ). (4.16)
Given G ⊂ F we define the function PX (B|G) = P (X ∈ B|G) for B ∈ F. Assume that f is such that
E(f (X)) < ∞. Then Z
E(f (X)|G) = f (x)PX (dx|G). (4.17)
R
Now we use the Markov property, together with equations (4.16) and (4.17) and the fact that s < u ⇒
FsX ⊂ FuX to calculate:

P (Γ, t|x, s) := P(Xt ∈ Γ|Xs = x) = P(Xt ∈ Γ|FsX )

= E(IΓ (Xt )|FsX ) = E(E(IΓ (Xt )|FsX )|FuX )
= E(E(IΓ (Xt )|FuX )|FsX ) = E(P(Xt ∈ Γ|Xu )|FsX )
= E(P(Xt ∈ Γ|Xu = y)|Xs = x)
Z
= P (Γ, t|Xu = y)P (dz, u|Xs = x)
ZR
=: P (Γ, t|y, u)P (dy, u|x, s).
R

IΓ (·) denotes the indicator function of the set Γ. 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 4.3.1. A Markov process is homogeneous if

P (t, Γ|Xs = x) := P (s, t, x, Γ) = P (0, t − s, x, Γ).

We set P (0, t, ·, ·) = P (t, ·, ·). The Chapman–Kolmogorov (CK) equation becomes

Z
P (t + s, x, Γ) = P (s, x, dz)P (t, z, Γ). (4.18)
E

Let Xt be a homogeneous Markov process and assume that the initial distribution of Xt is given
by the probability measure ν(Γ) = P (X0 ∈ Γ) (for deterministic initial conditions–X0 = x– we have
that ν(Γ) = IΓ (x) ). The transition function P (x, t, Γ) and the initial distribution ν determine the finite
dimensional distributions of X by

P(X0 ∈ Γ1 , X(t1 ) ∈ Γ1 , . . . , Xtn ∈ Γn )

Z Z Z
= ... P (tn − tn−1 , yn−1 , Γn )P (tn−1 − tn−2 , yn−2 , dyn−1 )
Γ0 Γ1 Γn−1
· · · × P (t1 , y0 , dy1 )ν(dy0 ). (4.19)

Theorem 4.3.2. ([3, Sec. 4.1]) Let P (t, x, Γ) satisfy (4.18) and assume that (E, ρ) 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).

45
 Let Xt be a homogeneous Markov process with initial distribution ν(Γ) = P (X0 ∈ Γ) and transition
function P (x, t, Γ). We can calculate the probability of finding Xt in a set Γ at time t:
Z
P(Xt ∈ Γ) = P (x, t, Γ)ν(dx).
E

Thus, the initial distribution and the transition function are sufficient to characterize a homogeneous Markov
process. Notice that they do not provide us with any information about the actual paths of the Markov
process. The transition probability P (Γ, t|x, s) is a probability measure. Assume that it has a density for all
t > s: Z
P (Γ, t|x, s) = p(y, t|x, s) dy.
Γ

Clearly, for t = s we have P (Γ, s|x, s) = IΓ (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,
Γ R Γ

and, since Γ ∈ B(R) is arbitrary, we obtain the equation

Z
p(y, t|x, s) = p(y, t|z, u)p(z, u|x, s) dz. (4.20)
R

The transition probability density is a function of 4 arguments: the initial position and time x, s and the final
position and time y, t.
In words, the CK equation tells us that, for a Markov process, the transition from x, s to y, t can be done
in two steps: first the system moves from x to z at some intermediate time u. Then it moves from z to y at
time t. In order to calculate the probability for the transition from (x, s) to (y, t) we need to sum (integrate)
the transitions from all possible intermediary states z. The above description suggests that a Markov process
can be described through a semigroup of operators, i.e. a one-parameter family of linear operators with the
properties
P0 = I, Pt+s = Pt ◦ Ps ∀ t, s > 0.

A semigroup of operators is characterized through its generator.

Indeed, let P (t, x, dy) be the transition function of a homogeneous Markov process. It satisfies the CK
equation (4.18): Z
P (t + s, x, Γ) = P (s, x, dz)P (t, z, Γ).
E

Let X := Cb (E) and define the operator

Z
(Pt f )(x) := E(f (Xt )|X0 = x) = f (y)P (t, x, dy).
E

This is a linear operator with

(P0 f )(x) = E(f (X0 )|X0 = x) = f (x) ⇒ P0 = I.

46
 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 .

4.4 The Generator of a Markov Processes

• Let (E, ρ) be a metric space and let {Xt } be an E-valued homogeneous Markov process. Define the
one parameter family of operators Pt through
Z
Pt f (x) = f (y)P (t, x, dy) = E[f (Xt )|X0 = x]

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 ,
t→0 t
exists.

Definition 4.4.1. The operator L : D(L) → Cb (E) is called the infinitesimal generator of the operator
semigroup Pt .

Definition 4.4.2. The operator L : Cb (E) → Cb (E) defined above is called the generator of the Markov
process {Xt }.

• The study of operator semigroups started in the late 40’s independently by Hille and Yosida. Semi-
group theory was developed in the 50’s and 60’s by Feller, Dynkin and others, mostly in connection
to the theory of Markov processes.

• Necessary and sufficient conditions for an operator L to be the generator of a (contraction) semigroup
are given by the Hille-Yosida theorem (e.g. Evans Partial Differential Equations, AMS 1998, Ch. 7).

47
 • 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.

• Furthermore, u(x, 0) = P0 f (x) = f (x). Consequently, u(x, t) satisfies the initial value problem

∂u
= Lu, u(x, 0) = f (x). (4.21)
∂t

• When the semigroup Pt is the transition semigroup of a Markov process Xt , then equation (4.21) is
called the backward Kolmogorov equation. It governs the evolution of an observable

u(x, t) = E(f (Xt )|X0 = x).

• 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 equation 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 applica-
tions; in particular when there is a measure µ on E, the spaces Lp (E; µ) sometimes arise. We will
quite often use the space L2 (E; µ), where µ will is the invariant measure of our Markov process.

• The generator is frequently 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 E–valued homogeneous Markov process
{Xt } associated with Pt defined through

E[f (X(t)|FsX )] = Pt−s f (X(s))

for all t, s ∈ T with t > s and f ∈ D(Pt ).

Example 4.4.3. The Poisson process is a homogeneous Markov process.

48
Example 4.4.4. The one dimensional Brownian motion is a homogeneous Markov process. The transition
function is the Gaussian defined in the example in Lecture 2:
 
1 |x − y|2
P (t, x, dy) = γt,x (y)dy, γt,x (y) = √ exp − .
2πt 2t
t d2
The semigroup associated to the standard Brownian motion is the heat semigroup Pt = e 2 dx2 . The genera-
d2
tor of this Markov process is 12 dx 2.

• Notice that the transition probability density γt,x of the one dimensional Brownian motion is the
fundamental solution (Green’s function) of the heat (diffusion) PDE
∂u 1 ∂2u
= .
∂t 2 ∂x2

4.5 The Adjoint Semigroup

• The semigroup Pt acts on bounded measurable functions.

• We can also define the adjoint semigroup Pt∗ which acts on probability measures:
Z Z
∗
Pt µ(Γ) = P(Xt ∈ Γ|X0 = x) dµ(x) = p(t, x, Γ) dµ(x).
R R

• The image of a probability measure µ under Pt∗ is again a probability measure. The operators Pt and
Pt∗ are adjoint in the L2 -sense:
Z Z
Pt f (x) dµ(x) = f (x) d(Pt∗ µ)(x). (4.22)
R R

• We can, formally at least, write

Pt∗ = exp(L∗ t),

• where L∗ is the L2 -adjoint of the generator of the process:

Z Z
Lf h dx = f L∗ h dx.

• Let µt := Pt∗ µ. This is the law of the Markov process and µ is the initial distribution. An argument
similar to the one used in the derivation of the backward Kolmogorov equation (4.21) enables us to
obtain an equation for the evolution of µt :
∂µt
= L ∗ µt , µ0 = µ.
∂t

• Assuming that µt = ρ(y, t) dy, µ = ρ0 (y) dy this equation becomes:

∂ρ
= L∗ ρ, ρ(y, 0) = ρ0 (y). (4.23)
∂t

49
 • This is the forward Kolmogorov or Fokker-Planck equation. When the initial conditions are deter-
ministic, X0 = x, the initial condition becomes ρ0 = δ(y − x).

• Given the initial distribution and the generator of the Markov process Xt , we can calculate the tran-
sition 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)ρ(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 (Schrödinger/Frobenious-Perron)
∂(Pt∗ µ)
= L∗ (Pt∗ µ).
∂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.

4.6 Ergodic Markov processes

• A very important concept in the study of limit theorems for stochastic processes is that of ergodicity.

• This concept, in the context of Markov processes, provides us with information on the long–time
behavior of a Markov semigroup.
Definition 4.6.1. A Markov process is called ergodic if the equation

Pt g = g, g ∈ Cb (E) ∀t > 0

has only constant solutions.

• Roughly speaking, ergodicity corresponds to the case where the semigroup Pt is such that Pt − I has
only constants in its null space, or, equivalently, to the case where the generator L has only constants
in its null space. This follows from the definition of the generator of a Markov process.

• Under some additional compactness assumptions, an ergodic Markov process has an invariant mea-
sure µ with the property that, in the case T = R+ ,
Z
1 t
lim g(Xs ) ds = Eg(x),
t→+∞ t 0

50
 • where E denotes the expectation with respect to µ.

• This is a physicist’s definition of an ergodic process: time averages equal phase space averages.

• Using the adjoint semigroup we can define an invariant measure as the solution of the equation

Pt∗ µ = µ.

• 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 µ is invariant if and
only if
L∗ µ = 0

• in some appropriate generalized sense ((L∗ µ, f ) = 0 for every bounded measurable function).

• Assume that µ(dx) = ρ(x) dx. Then the invariant density satisfies the stationary Fokker-Planck
equation
L∗ ρ = 0.

• The invariant measure (distribution) governs the long-time dynamics of the Markov process.

4.7 Stationary Markov Processes

• If X0 is distributed according to µ, then so is Xt for all t > 0. The resulting stochastic process, with
X0 distributed in this way, is stationary .

• In this case the transition probability density (the solution of the Fokker-Planck equation) is indepen-
dent of time: ρ(x, t) = ρ(x).

• Consequently, the statistics of the Markov process is independent of time.

Example 4.7.1. The one dimensional Brownian motion is not an ergodic process: The null space of the
d2
generator L = 12 dx 2 on R is not one dimensional!

Example 4.7.2. Consider a one-dimensional Brownian motion on [0, 1], with periodic boundary conditions.
d2
The generator of this Markov process L is the differential 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
∂ρ 1 ∂2ρ
=
∂t 2 ∂x2
with periodic boundary conditions in [0, 1]. Fourier analysis shows that the solution converges to a constant
at an exponential rate.

51
Example 4.7.3. • The one dimensional Ornstein-Uhlenbeck (OU) process is a Markov process with
generator
d d2
L = −αx + D 2.
dx dx
• The null space of L comprises constants in x. Hence, it is an ergodic Markov process. In order to
calculate the invariant measure we need to solve the stationary Fokker–Planck equation:

L∗ ρ = 0, ρ > 0, kρkL1 (R) = 1. (4.24)

• Let us calculate the L2 -adjoint of L. Assuming that f, h decay sufficiently fast at infinity, we have:
Z Z
 
Lf h dx = (−αx∂x f )h + (D∂x2 f )h dx
R ZR Z
 
= f ∂x (αxh) + f (D∂x2 h) dx =: f L∗ h dx,
R R

• where
d d2 h
L∗ h := (axh) + D 2 .
dx dx
• We can calculate the invariant distribution by solving equation (4.24).

• The invariant measure of this process is the Gaussian measure

r  α 
α
µ(dx) = exp − x2 dx.
2πD 2D

• If the initial condition of the OU process is distributed according to the invariant measure, then the
OU process is a stationary Gaussian process.

• Let Xt be the 1d OU process and let X0 ∼ N (0, D/α). Then Xt is a mean zero, Gaussian second
order stationary process on [0, ∞) with correlation function

D −α|t|
R(t) = e
α
and spectral density
D 1
f (x) = .
π x + α2
2

Furthermore, the OU process is the only real-valued mean zero Gaussian second-order stationary
Markov process defined on R.

52
Chapter 5

Diffusion Processes

5.1 Definition of a Diffusion Process

• A Markov process consists of three parts: a drift (deterministic), a random process and a jump process.

• A diffusion process is a Markov process that has continuous sample paths (trajectories). Thus, it is a
Markov process with no jumps.

• A diffusion process can be defined by specifying its first two moments:

Definition 5.1.1. A Markov process Xt with transition probability P (Γ, t|x, s) is called a diffusion process
if the following conditions are satisfied.
i. (Continuity). For every x and every ε > 0
Z
P (dy, t|x, s) = o(t − s) (5.1)
|x−y|>ε

uniformly over s < t.

ii. (Definition of drift coefficient). There exists a function a(x, s) such that for every x and every ε > 0
Z
(y − x)P (dy, t|x, s) = a(x, s)(t − s) + o(t − s). (5.2)
|y−x|6ε

uniformly over s < t.

iii. (Definition of diffusion coefficient). There exists a function b(x, s) such that for every x and every
ε>0 Z
(y − x)2 P (dy, t|x, s) = b(x, s)(t − s) + o(t − s). (5.3)
|y−x|6ε
uniformly over s < t.
Remark 5.1.2. In Definition 5.1.1 we had to truncate the domain of integration since we didn’t know
whether the first and second moments exist. If we assume that there exists a δ > 0 such that
Z
1
lim |y − x|2+δ P (s, x, t, dy) = 0, (5.4)
t→s t − s Rd

53
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 (s, x, t, dy)
|y−x|>ε
Z
= |y − x|2+δ |y − x|k−(2+δ) P (s, x, t, dy)
|y−x|>ε
Z
1
6 2+δ−k |y − x|2+δ P (s, x, t, dy)
ε |y−x|>ε
Z
1
6 2+δ−k |y − x|2+δ P (s, x, t, dy).
ε R d

Using this estimate together with (5.4) we conclude that:

Z
1
lim |y − x|k P (s, x, t, dy) = 0, k = 0, 1, 2.
t→s t − s |y−x|>ε

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 (5.2) 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 X
 s = x = a(x, s) (5.5)
t→s t−s
and  
|Xt − Xs |2 
lim E Xs = x = b(x, s) (5.6)
t→s t−s

5.2 The Backward and Forward Kolmogorov Equations

Under mild regularity assumptions we can show that a diffusion process is completely determined from its
first two moments. In particular, we will obtain partial differential equations that govern the evolution of the
conditional expectation of an arbitrary function of a diffusion process Xt , u(x, s) = E(f (Xt )|Xs = x), as
well as of the transition probability density p(y, t|x, s) = P(Xt = y|Xs = s). These are the backward and
forward Kolmogorov equations.

5.2.1 The Backward Kolmogorov Equation

Theorem 5.2.1. (Kolmogorov) Let f (x) ∈ Cb (R) and let
Z
u(x, s) := E(f (Xt )|Xs = x) = f (y)P (dy, t|x, s) ∈ Cb2 (R).

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 s→t

54
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)
|y−x|6ε |y−x|>ε
Z Z
6 f (y)P (dy, t|x, s) + kf kL∞ P (dy, t|x, s)
|y−x|6ε |y−x|>ε
Z
= f (y)P (dy, t|x, s) + o(t − s).
|y−x|6ε

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 R
Z Z
= f (x) + (f (y) − f (x))P (dy, t|x, s) + (f (y) − f (x))P (dy, t|x, s)
|y−x|6ε |y−x|>ε
Z
= f (x) + (f (y) − f (x))P (dy, t|x, s) + o(t − s).
|y−x|6ε

Now the final condition follows from the fact that f (x) ∈ Cb (R) and the arbitrariness of ε.
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, σ) = f (z)P (dz, t|x, σ) (5.8)
ZR Z
= f (z)P (dz, t|y, ρ)P (dy, ρ|x, σ)
ZR R
= u(y, ρ)P (dy, ρ|x, σ). (5.9)
R

The Taylor series expansion of the function u(s, x) gives

∂u(x, ρ) 1 ∂ 2 u(x, ρ)
u(z, ρ) − u(x, ρ) = (z − x) + (z − x)2 (1 + αε ), |z − x| 6 ε, (5.10)
∂x 2 ∂x2

where
 2 
 ∂ u(x, ρ) ∂ 2 u(z, ρ) 
αε = sup   − .
ρ,|z−x|6ε ∂x2 ∂x2 

Notice that, since u(x, s) is twice continuously differentiable in x, limε→0 αε = 0.

55
 We combine now (5.9) with (5.10) to calculate
Z 
u(x, s) − u(x, s + h) 1
= P (dy, s + h|x, s)u(y, s + h) − u(x, s + h)
h h
Z R
1
= P (dy, s + h|x, s)(u(y, s + h) − u(x, s + h))
h R
Z
1
= P (dy, s + h|x, s)(u(y, s + h) − u(x, s)) + o(1)
h |x−y|<ε
Z
∂u 1
= (x, s + h) (y − x)P (dy, s + h|x, s)
∂x h |x−y|<ε
Z
1 ∂2u 1
+ (x, s + h) (y − x)2 P (dy, s + h|x, s)(1 + αε ) + o(1)
2 ∂x2 h |x−y|<ε
∂u 1 ∂2u
= a(x, s) (x, s + h) + b(x, s) 2 (x, s + h)(1 + αε ) + o(1).
∂x 2 ∂x
Equation (5.7) follows by taking the limits ε → 0, h → 0.

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

Substituting this in the backward Kolmogorov equation we obtain

Z  
∂p(y, t|x, s)
f (y) + As,x p(y, t|x, s) = 0 (5.11)
R ∂s

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:

∂p(y, t|x, s) ∂p(y, t|x, s) 1 ∂ 2 p(y, t|x, s)

− = a(x, s) + b(x, s) . (5.12)
∂s ∂x 2 ∂x2
Notice that the variation is with respect to the ”backward” variables x, s. We will obtain an equation with
respect to the ”forward” variables y, t in the next section.

5.2.2 The Forward Kolmogorov Equation

In this section we will obtain the forward Kolmogorov equation. In the physics literature is called the
Fokker-Planck equation. We assume that the transition function has a density with respect to Lebesgue
measure. Z
P (Γ, t|x, s) = p(t, y|x, s) dy.
Γ

56
Theorem 5.2.2. (Kolmogorov) Assume that conditions (5.1), (5.2), (5.3) are satisfied 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) = δ(x − y). (5.13)
∂t ∂y 2 ∂y 2 t→s

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), (5.14)
h→0 h 2
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
h→0 h
Z Z 
1
= lim p(y, t + h|x, s)f (y) dy − p(z, t|s, x)f (z) dz
h→0 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) dz
h→0 h
Z Z 
1
= lim p(z, t|x, s) p(y, t + h|z, t)f (y) dy − f (z) dz
h→0 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.
Since the above equation is valid for every test function f (y), the forward Kolmogorov equation follows.

Exercise 5.2.3. Prove equation (5.14).

Assume now that initial distribution of Xt is ρ0 (x) and set s = 0 (the initial time) in (5.13). Define
Z
p(y, t) := p(y, t|x, 0)ρ0 (x) dx.

We multiply the forward Kolmogorov equation (5.13) by ρ0 (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)) , (5.15)
∂t ∂y 2 ∂y 2
together with the initial condition
p(y, 0) = ρ0 (y). (5.16)

57
The solution of equation (5.15), provides us with the probability that the diffusion process Xt , which initially
was distributed according to the probability density ρ0 (x), is equal to y at time t. Alternatively, we can think
of the solution to (5.13) as the Green’s function for the PDE (5.15).
Quite often we need to calculate joint probability densities. For, example the probability that Xt1 = x1
and Xt2 = x2 . From the properties of conditional expectation we have that

p(x1 , t1 , x2 , t2 ) = PXt1 = x1 , Xt2 = x2 )

= P(Xt1 = x1 |Xt2 = x2 )P(Xt2 = x2 )
= p(x1 , t1 |x2 t2 )p(x2 , t2 ).

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. (5.17)

The autocorrelation function at time t and s is given by

Z Z
E(Xt Xs ) = yxp(y, t|x, s)p(x, s) dxdy.

In particular, Z Z
E(Xt X0 ) = yxp(y, t|x, 0)p(x, 0) dxdy.

Notice also that from (5.17) it follows that

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 (5.15).

5.3 The Fokker-Planck Equation in Arbitrary Dimensions

The drift and diffusion coefficients of a diffusion process in Rd are defined as:
Z
1
lim (y − x)P (dy, t|x, s) = a(x, s)
t→s t − s |y−x|<ε

and Z
1
lim (y − x) ⊗ (y − x)P (dy, t|x, s) = b(x, s).
t→s t − s |y−x|<ε

The drift coefficient a(x, s) is a d-dimensional vector field and the diffusion coefficient b(x, s) is a d × d
symmetric matrix (second order tensor). The generator of a d dimensional diffusion process is
1
L = a(s, x) · ∇ + b(s, x) : ∇∇
2
Xd d
∂ 1 X ∂2
= aj + bij 2 .
∂xj 2 ∂xj
j=1 i,j=1

58
Exercise 5.3.1. Derive rigorously the forward and backward Kolmogorov equations 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 X
 s = x = a(x, s) (5.18)
t→s t−s

and  
(Xt − Xs ) ⊗ (Xt − Xs ) 
lim E Xs = x = b(x, s) (5.19)
t→s t−s
Notice that from the above definition it follows that the diffusion matrix is symmetric and nonnegative
definite.

5.4 Connection with Stochastic Differential Equations

Notice also that the continuity condition can be written in the form

P (|Xt − Xs | > ε|Xs = x) = o(t − s).

Now it becomes clear that this condition implies that the probability of large changes in Xt over short
time intervals is small. Notice, on the other hand, that the above condition implies that the sample paths of a
diffusion process are not differentiable: if they where, then the right hand side of the above equation would
have to be 0 when t − s ≪ 1. The sample paths of a diffusion process have the regularity of Brownian paths.
A Markovian process cannot be differentiable: we can define the derivative of a sample paths only with
processes for which the past and future are not statistically independent when conditioned on the present.
Let us denote the expectation conditioned on Xs = x by Es,x. Notice that the definitions of the drift and
diffusion coefficients (5.5) and (5.6) can be written in the form

Es,x (Xt − Xs ) = a(x, s)(t − s) + o(t − s).

and  
Es,x (Xt − Xs ) ⊗ (Xt − Xs ) = b(x, s)(t − s) + o(t − s).
Consequently, the drift coefficient defines the mean velocity vector for the stochastic process Xt , whereas
the diffusion coefficient (tensor) is a measure of the local magnitude of fluctuations of Xt − Xs about the
mean value. hence, we can write locally:

Xt − Xs ≈ a(s, Xs )(t − s) + σ(s, Xs ) ξt ,

where b = σσ T and ξt is a mean zero Gaussian process with

E s,x (ξt ⊗ ξs ) = (t − s)I.

Since we have that

Wt − Ws ∼ N (0, (t − s)I),

59
we conclude that we can write locally:

∆Xt ≈ a(s, Xs )∆t + σ(s, Xs )∆Wt .

Or, replacing the differences by differentials:

dXt = a(t, Xt )dt + σ(t, Xt )dWt .

Hence, the sample paths of a diffusion process are governed by a stochastic differential equation (SDE).

5.5 Discussion and Bibliography

The argument used in the derivation of the forward and backward Kolmogorov equations goes back to
Kolmogorov’s original work. More material on diffusion processes can be found in [11], [14].

60
Chapter 6

The Fokker-Planck Equation

In the previous chapter we derived the backward and forward (Fokker-Planck) Kolmogorov equations and
we showed that all statistical properties of a diffusion process can be calculated from the solution of the
Fokker-Planck equation. 1 In this chapter we will 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 will also study in some detail various examples. We will restrict attention to
time-homogeneous diffusion processes, for which the drift and diffusion coefficients do not depend on time.

6.1 Basic Properties of the FP Equation

6.1.1 Existence and Uniqueness of Solutions
Consider a homogeneous diffusion process on Rd with drift vector and diffusion matrix a(x) and b(x). The
Fokker-Planck equation is
d
X d
∂p ∂ 1 X ∂2
=− (ai (x)p) + (bij (x)p), t > 0, x ∈ Rd , (6.1a)
∂t ∂xj 2 ∂xi ∂xj
j=1 i,j=1

p(x, 0) = f (x), x ∈ Rd . (6.1b)

Since f (x) is the probability density of the initial condition (which is a random variable), we have that
Z
f (x) > 0, and f (x) dx = 1.
Rd

We can also write the equation in non-divergence form:

d d
∂p X ∂p 1 X ∂2p
= ãj (x) + b̃ij(x) + c̃(x)u, t > 0, x ∈ Rd , (6.2a)
∂t ∂xj 2 ∂xi ∂xj
j=1 i,j=1

p(x, 0) = f (x), x ∈ Rd , (6.2b)

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.

61
where
d
X d d
∂bij 1 X ∂ 2 bij X ∂ai
ãi (x) = −ai (x) + , c̃i (x) = − .
∂xj 2 ∂xi ∂xj ∂xi
j=1 i,j=1 i=1

By definition (see equation (5.19)), 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)ξi ξj > αkξk2 , ∀ ξ ∈ Rd , (6.3)
i,j=1

Furthermore, we will assume that the coefficients ã, b, c̃ are smooth and that they satisfy the growth condi-
tions
kb(x)k 6 M, kã(x)k 6 M (1 + kxk), kc̃(x)k 6 M (1 + kxk2 ). (6.4)

Definition 6.1.1. 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+ ).

ii. ∀T > 0 there exists a c > 0 such that

2
ku(t, x)kL∞ (0,T ) 6 ceαkxk

iii. limt→0 u(t, x) = f (x).

It is a standard result in the theory of parabolic partial differential equations that, under the regularity and
uniform ellipticity assumptions, the Fokker-Planck equation has a unique smooth solution. Furthermore, the
solution can be estimated in terms of an appropriate heat kernel (i.e. the solution of the heat equation on
Rd ).
2
Theorem 6.1.2. Assume that conditions (6.3) and (6.4) are satisfied, and assume that |f | 6 ceαkxk . Then
there exists a unique classical solution to the Cauchy problem for the Fokker–Planck equation. Furthermore,
there exist positive constants K, δ so that
 
2 (−n+2)/2 1 2
|p|, |pt |, k∇pk, kD pk 6 Kt exp − δkxk . (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 6.1.3. 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.

62
6.1.2 The FP equation as a conservation law
The Fokker-Planck equation is in fact a conservation law: it expresses the law of conservation of probability.
To see this we define the probability current to be the vector whose ith component is

d
1X ∂

Ji := ai (x)p − bij (x)p . (6.6)
2 ∂xj
j=1

We use the probability current to write the Fokker–Planck equation as a continuity equation:

∂p
+ ∇ · J = 0.
∂t

Integrating the FP equation over Rd and integrating by parts on the right hand side of the equation we obtain
Z
d
p(x, t) dx = 0.
dt Rd

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.

6.1.3 Boundary conditions for the Fokker–Planck equation

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 (µ) where µ 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 necessary to specify the value of the stochastic process (or equivalently
of the solution to the Fokker-Planck equation) on the boundary.
To understand the type of boundary conditions that we can impose on the Fokker-Planck equation, let
us consider the example of a random walk on the domain {0, 1, . . . N }.2 When the random walker reaches
either the left or the right boundary we can either set

i. X0 = 0 or XN = 0, which means that the particle gets absorbed at the boundary;

ii. X0 = X1 or XN = XN −1 , which means that the particle is reflected at the boundary;

iii. X0 = XN , which means that the particle is moving on a circle (i.e., we identify the left and right
boundaries).
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.

63
Hence, we can have absorbing, reflecting or periodic boundary conditions.
Consider the Fokker-Planck equation posed in Ω ⊂ Rd where Ω is a bounded domain with smooth
boundary. Let J denote the probability current and let n be the unit outward pointing normal vector to the
surface. The above boundary conditions become:

i. The transition probability density vanishes on an absorbing boundary:

p(x, t) = 0, on ∂Ω.

ii. There is no net flow of probability on a reflecting boundary:

n · J(x, t) = 0, on ∂Ω.

iii. The transition probability density is a periodic function in the case of periodic 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.
Consider now a diffusion process in one dimension on the interval [0, L]. The boundary conditions are

p(0, t) = p(L, t) = 0 absorbing,

J(0, t)) = J(L, t) = 0 reflecting,

p(0, t) = p(L, t) periodic,
where the probability current is defined in (6.6). An example of mixed boundary conditions would be
absorbing boundary conditions at the left end and reflecting boundary conditions at the right end:

p(0, t) = J(L, t) = 0.

There is a complete classification of boundary conditions in one dimension, the Feller classification: the
BC can be regular, exit, entrance and natural.

6.2 Examples of Diffusion Processes

6.2.1 Brownian Motion
Set a(y, t) ≡ 0, b(y, t) ≡ 2D > 0. This diffusion process is the Brownian motion with diffusion coefficient
D. The Fokker-Planck equation is:

∂p ∂2p
= D 2, p(y, s|x, s) = δ(x − y). (6.8)
∂t ∂y
As we have already seen, the solution to this equation is
 
1 (y − x)2
pW (y, s|x, t) = p exp − .
4πD(s − t) 4D(s − t)

64
Notice that using the Fokker-Planck equation for the Brownian motion we can immediately show that the
mean squared displacement scales linearly in time:
Z
d d
Ex2 = x2 p(x, t) dx
dt dt R
Z
∂ 2 p(x, t)
= D x2 dx
∂x2
ZR
= D p(x, t) dx = 2D.
R

Consequently,
Ex2 = 2Dt.
Assume now that the initial distribution of the Brownian motion is ρ0 (x). The solution of the Fokker-Planck
equation for the Brownian motion with this initial distribution is
Z
PW (y, t) = p(y, t|x, 0)W (x) dy.

We can also consider Brownian motion in a bounded domain, with either absorbing, reflecting or periodic
boundary conditions. Set D = 12 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. (6.9)
∂t 2 ∂x2
We look for a solution to this equation in a sine Fourier series:
∞
X
p(x, t) = pn (t) sin(nπx). (6.10)
k=1

Notice that the boundary conditions are automatically satisfied. The initial condition it

p(x, 0) = δ(x − x0 ),

where we have assumed that W0 = x0 . The Fourier coefficients of the initial conditions are
Z 1
pn (0) = 2 δ(x − x0 ) sin(nπx) dx = 2 sin(nπx0 ).
0

We substitute the expansion (6.10) into (6.9) and use the orthogonality properties of the Fourier basis to
obtain the equations
n2 π 2
ṗn = − pn n = 1, 2, . . .
2
The solution of this equation is
n2 π 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 π 2
p(x, t|x0 , 0) = 2 e− 2 t sin nπx0 sin(nπx).
n=1

65
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.
1
Exercise 6.2.1. Consider the Brownian motion with D = 2 on the interval [0,1] with reflecting boundary
conditions. The Fokker-Planck equation is

∂p 1 ∂2p
= , ∂x p(0, t) = ∂x p(1, t) = 0.
∂t 2 ∂x2
i. Find the transition probability density for a Brownian particle starting at x0 .

ii. Show that the process is ergodic and find the invariant distribution ps (x).

iii. Calculate the stationary autocorrelation function

Z 1Z 1
E(W (t)W (0)) = xx0 p(x, t|x0 , 0)ps (x0 ), dxdx0 .
0 0

6.2.2 The Onrstein-Uhlenbeck Process

We set now a(y, t) = −αy, b(y, t) ≡ 2D > 0 into the Fokker-Planck equation to obtain

∂p ∂(yp) ∂2p
=α + D 2. (6.11)
∂t ∂x ∂y

This is the Fokker-Planck equation for the Ornstein-Uhlenbeck process. The corresponding stochastic dif-
ferential equation is
√
dXt = −αXt + 2DdWt .

So, in addition to Brownian motion there is a linear force pulling the particle towards the origin. This is the
reason why the OU process is ergodic.
The transition probability density for the the OU process is
r !
α α(y − e−α(t−s) x)2
pOU (y, t|x, s) = exp − . (6.12)
2πD(1 − e−2α(t−s) ) 2D(1 − e−2α(t−s) )

We obtained this formula in Example (4.1.4) (for α = D = 1) by using the fact that the OU process can be
defined through the a time change of the Brownian motion. We can also derive it by solving equation (6.11).

Exercise 6.2.2. Solve equation (6.11) by taking the Fourier transform, using the method of characteristics
for first order PDEs and taking the inverse Fourier transform.

We set x = 0, s = 0 in (6.12):
r  
α αy 2
pOU (y, t|0, 0) = exp − .
2πD(1 − e−2αt ) 2D(1 − e−2αt )

66
We have that
lim pOU (y, t) = pW (y, t).
α→0
Thus, in the limit where the friction coefficient goes to 0, we recover the transition probability of Brownian
motion from the transition probability of the OU processes. Notice also that
r  
α αy 2
lim pOU (y, t) = exp − .
t→+∞ 2πD 2D
As we have already seen, the Ornstein-Uhlenbeck process is an ergodic Markov process, and its invariant
measure is Gaussian.
Using (6.12) we can calculate all moments of the OU process We can calculate all moments of the OU
process
Z Z
E((Xt )n ) = y n p(y, t|x, 0)p0 (x) dxdy,

where p0 (x) is the initial distribution. We will calculate the moments by using the Fokker-Planck equation,
rather than the explicit formula for the transition probability density.
Define the nth moment of the OU process:
Z
Mn = y n p(y, t) dy, n = 0, 1, 2, . . . ,
R
R
where p(y, t) = R p(y, t|x, 0)p0 (x) dx. Let n = 0. We integrate the FP equation over R to obtain:
Z Z Z 2
∂p ∂(yp) ∂ p
=α +D = 0,
∂t ∂y ∂y 2
after an integration by parts and using the fact that p(x, t) decays sufficiently fast at infinity. Consequently:
d
M0 = 0 ⇒ M0 (t) = M0 (0) = 1.
dt
In other words: Z Z
d
kpkL1 (R) = 0 ⇒ p(y, t) dy = p(y, t = 0) dy = 1.
dt R R
Consequently: probability is conserved, as we have already shown. Let n = 1. We multiply the FP
equation for the OU process by x, integrate over R and perform and integration by parts to obtain:
d
M1 = −αM1 .
dt
Consequently, the first moment converges exponentially fast to 0:

M1 (t) = e−αt 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 = −αn y p + Dn(n − 1) y n−2 p.
n n
dt

67
Or, equivalently:
d
Mn = −αnMn + Dn(n − 1)Mn−2 , n > 2.
dt
This is a first order linear inhomogeneous differential equation. We can solve it using the variation of
constants formula:
Z t
−αnt
Mn (t) = e Mn (0) + Dn(n − 1) e−αn(t−s) Mn−2 (s) ds. (6.13)
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
−2αt
M2 (t) = e M2 (0) + D e−2α(t−s) M0 (s) ds
0
D −2αt 2αt
= e−2αt M2 (0) + e (e − 1)
 2α 
D −2αt D
= +e M2 (0) − .
2α 2α

D
Consequently, the second moment converges exponentially fast to its stationary value 2α . The stationary
moments of the OU process are:
r Z
n α αy 2
hy iOU = y n e− 2D dx
2πD R
n

D n/2
= 1.3 . . . (n − 1) α , n even,
0, n odd.

It is not hard to check that

lim Mn (t) = hy n iOU . (6.14)
t→∞

Furthermore, if the initial conditions of the OU process are stationary, then:

Mn (t) = Mn (0) = hxn iOU . (6.15)

Exercise 6.2.3. Use (6.13) to show (6.14) and (6.15).

Exercise 6.2.4. Show that the autocorrelation function of the stationary Ornstein-Uhlenbeck is
Z Z
E(Xt X0 ) = xx0 pOU (x, t|x0 , 0)ps (x0 ) dxdx0
R R
D −α|t|
= e ,
2α

where ps (x) denotes the invariant Gaussian distribution.

68
6.2.3 The Geometric Brownian Motin
We set a(x) = µx, b(x) = 12 σ 2 x2 . This is the geometric Brownian motion. The corresponding stochastic
differential equation is
dXt = µXt dt + σXt dWt .
This equation is one of the basic models of mathematical finance. The coefficient σ is called the volatility.
The generator of this process is
∂ σx2 ∂ 2
L = µx + .
∂x 2 ∂x2
Notice that this operator is not uniformly elliptic. The Fokker-Planck equation of the geometric Brownian
motion is:  
∂p ∂ ∂ 2 σ 2 x2
=− (µx) + 2 p .
∂t ∂x ∂x 2
We can easily obtain an equation for the nth moment of the geometric Brownian motion:

d  σ2 
Mn = µn + n(n − 1) Mn , n > 2.
dt 2
The solution of this equation is
σ2
Mn (t) = e(µ+(n−1) 2
)nt
Mn (0), n>2

and
M1 (t) = eµt M1 (0).
Notice that the nth moment might diverge as t → ∞, depending on the values of µ and σ. Consider for
example the second moment and assume that µ < 0. We have
2 )t
Mn (t) = e(2µ+σ M2 (0),

which diverges when σ 2 + 2µ > 0.

6.3 The OU Process and Hermite Polynomials

The Ornstein-Uhlenbeck process is one of the few stochastic processes for which we can calculate explicitly
the solution of the corresponding SDE, the solution of the Fokker-Planck equation as well as the eigenfunc-
tions of the generator of the process. In this section we will show that the eigenfunctions of the OU process
are the Hermite polynomials. We will also study various properties of the generator of the OU process. In
the next section we will show that many of the properties of the OU process (ergodicity, self-adjointness of
the generator, exponentially fast convergence to equilibrium, real, discrete spectrum) are shared by a large
class of diffusion processes, namely those for which the drift term can be written in terms of the gradient of
a smooth functions. We will refer to these processes as gradient flows.
The generator of the d-dimensional OU process is (we set the drift coefficient equal to 1)

L = −p · ∇p + β −1 ∆p (6.16)

69
We have already seen that the OU process is an ergodic Markov process whose unique invariant measure is
absolutely continuous with respect to the Lebesgue measure on Rd with Gaussian density ρ ∈ C ∞ (Rd )
1 |p|2
ρβ (p) = e−β 2 .
(2πβ −1 )d/2

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
2
kf kρ := f 2 ρβ dp.
Rd

and corresponding inner product Z

(f, h)ρ = f hρβ dp.
R

Similarly, we can define weighted L2 -spaced involving derivatives.

Exercise 6.3.1. Give the definition of the Sobolev space H p (Rd ; ρβ ).

The reason why this is the right function space is that the generator of the OU process becomes a self-
adjoint operator in this space. In fact, L defined in (6.16) has many nice properties that are summarized in
the following proposition.

Proposition 6.3.2. The operator L has the following properties:

• For every f, h ∈ C02 (Rd ) ∩ L2ρ (Rd ),

Z
−1
(Lf, h)ρ = (f, Lh)ρ = −β ∇f · ∇hρβ dp. (6.17)
Rd

• L is a non-positive operator on L2ρ .

• Lf = 0 iff f ≡ const.
R
• For every f ∈ C02 (Rd ) ∩ L2ρ (Rd ) with f ρβ = 0,

(−Lf, f )ρ > kf k2ρ (6.18)

Proof. Equation (6.17) follows from an integration by parts:

Z Z
−1
(Lf, h)ρ = −p · ∇f hρβ dp + β ∆f hρβ dp
Z Z Z
−1
= −p · ∇f hρβ dp − β ∇f · ∇hρβ dp + −p · ∇f hρβ dp

= −β −1 (∇f, ∇h)ρ .

Non-positivity of L follows from (6.17) upon setting h = f :

(Lf, f )ρ = −β −1 k∇f k2ρ 6 0.

70
Similarly, multiplying the equation Lf = 0 by f ρβ , integrating over Rd and using (6.17) gives

kf kρ = 0,

from which we deduce that f ≡ const. The spectral gap follows from (6.17), together with the Poincaré
inequality for Gaussian measures:
Z Z
f 2 ρβ dp 6 β −1 |∇f |2 ρβ dp (6.19)
Rd Rd
R
for every f ∈ H 1 (Rd ; ρβ ) with f ρβ = 0. Indeed, upon combining (6.17) with (6.19) we obtain:

(Lf, f )ρ = −β −1 |∇f |2ρ

6 −kf k2ρ

The spectral gap of the generator of the OU process, which is equivalent to the compactness of its
resolvent, implies that L has discrete spectrum. Furthermore, since it is also a self-adjoint operator, we have
that its eigenfunctions form a countable orthonormal basis for the separable Hilbert space L2ρ . In fact, we
can calculate the eigenvalues and eigenfunctions of the generator of the OU process in one dimension3

Theorem 6.3.3. Consider the eigenvalue problem for the generator of the OU process in one dimension

−Lfn = λn fn . (6.20)

Then the eigenvalues of L are the nonnegative integers:

λn = n, n = 0, 1, 2, . . . .

The corresponding eigenfunctions are the normalized Hermite polynomials:

1 p 
fn (p) = √ Hn βp , (6.21)
n!
where  
n p2 dn − p2
2
Hn (p) = (−1) e 2 e . (6.22)
dpn
Hermite polynomials appear very frequently in applications and they also play a fundamental role in
analysis. It is possible to prove that the Hermite polynomials form an orthonormal basis for L2 (Rd , ρβ )
without using the fact that they are the eigenfunctions of the a symmetric operator with compact resolvent4
A typical result along these lines is [24, Lemma 2.3.4] (we use the notation ρ1 = ρ):
3
The multidimensional problem can be treated similarly by taking tensor products of the eigenfunctions of the one dimensional
problem. See Exercise.
4
In fact, the Poincaré inequality for Gaussian measures can be proved using the fact that that the Hermite polynomials form an
orthonormal basis for L2 (Rd , ρβ ).

71
Proposition 6.3.4. For each λ ∈ C, set

λ2
H(p; λ) = eλp− 2 , p ∈ R.

Then
∞
X λn
H(p; λ) = Hn (p), p ∈ R, (6.23)
n!
n=0

where the convergence is both uniform on compact subsets of R × C, and for λ’s in compact subsets of
√
C, uniform in L2 (C; ρ). In particular, {fn (p) := √1n! Hn ( βp) : n ∈ N} is an orthonormal basis in
L2 (C; ρβ ).

From (6.22) it is clear that Hn is a polynomial of degree n. Furthermore, only odd (even) powers appear
in Hn (p) when n is odd (even). Furthermore, the coefficient multiplying pn in Hn (p) is always 1. The
orthonormality of the modified Hermite polynomials fn (p) defined in (6.21) implies that
Z
fn (p)fm (p)ρβ (p) dp = δnm .
R

The first few Hermite polynomials and the corresponding rescaled/normalized eigenfunctions of the gener-
ator of the OU process are:

H0 (p) = 1, f0 (p) = 1,
p
H1 (p) = p, f1 (p) = βp,
β 1
H2 (p) = p2 − 1, f2 (p) = √ p2 − √ ,
2 2
3/2
√
β 3 β
H3 (p) = p3 − 3p, 3
f3 (p) = √ p − √ p
6 6
1

H4 (p) = p4 − 3p2 + 3, f4 (p) = √ β 2 p4 − 3βp2 + 3
24
1  5/2 5 
H5 (p) = p5 − 10p3 + 15p, f5 (p) = √ β p − 10β 3/2 p3 + 15β 1/2 p .
120

The proof of Theorem 6.3.3 follows essentially from the properties of the Hermite polynomials. First, notice
that by combining (6.21) and (6.23) we obtain

p +∞
X λn
H( βp, λ) = √ fn (p)
n=0 n!

We differentiate this formula with respect to p to obtain

p p +∞
X λn
λ βH( βp, λ) = √ ∂p fn (p),
n=1
n!

72
since f0 = 1. From this equation we obtain

p +∞
X λn−1
H( βp, λ) = √ √ ∂p fn (p)
n=1 β n!
+∞
X λn
= √ p ∂p fn+1 (p)
n=0 β (n + 1)!

from which we deduce that

1 √
√ ∂p fk = kfk−1 . (6.24)
β
Similarly, if we differentiate (6.23) with respect to λ we obtain
+∞ k
X +∞
X +∞
X λk
λ λk
(p − λ)H(p; λ) = pHk (p) − Hk−1 (p) Hk+1 (p)
k! (k − 1)! k!
k=0 k=1 k=0

from which we obtain the recurrence relation

pHk = Hk+1 + kHk−1 .

Upon rescaling, we deduce that

p p
pfk = β −1 (k + 1)fk+1 + β −1 kfk−1 . (6.25)

We combine now equations (6.24) and (6.25) to obtain

 
p 1 √
βp − √ ∂p fk = k + 1fk+1 . (6.26)
β

Now we observe that

 
p 1 1
−Lfn = βp − √ ∂p √ ∂p fn
β β
 
p 1 √
= βp − √ ∂p nfn−1 = nfn .
β
√ 
The operators βp − √1β ∂p and √1β ∂p play the role of creation and annihilation operators. In fact,
we can generate all eigenfunctions of the OU operator from the ground state f0 = 0 through a repeated
application of the creation operator.

Proposition 6.3.5. Set β = 1 and let a− = ∂p . Then the L2ρ -adjoint of a+ is

a+ = −∂p + p.

Then the generator of the OU process can be written in the form

L = −a+ a− .

73
Furthermore, a+ and a− satisfy the following commutation relation

[a+ , a− ] = −1

Define now the creation and annihilation operators on C 1 (R) by

1
S+ = p a+
(n + 1)
and
1
S − = √ a− .
n
Then
S + fn = fn+1 and S − fn = fn−1 . (6.27)
In particular,
1
fn = √ (a+ )n 1 (6.28)
n!
and
1
1 = √ (a− )n fn . (6.29)
n!
Proof. let f, h ∈ C 1 (R) ∩ L2ρ . We calculate
Z Z
∂p f hρ = − f ∂p (hρ) (6.30)
Z


= f − ∂p + p hρ. (6.31)

Now,
−a+ a− = −(−∂p + p)∂p = ∂p − p∂p = L.
Similarly,
a− a+ = −∂p2 + p∂p + 1.
and
[a+ , a− ] = −1
Forumlas (6.27) follow from (6.24) and (6.26). Finally, formulas (6.28) and (6.29) are a consequence
of (6.24) and (6.26), together with a simple induction argument.

Notice that upon using (6.28) and (6.29) and the fact that a+ is the adjoint of a− we can easily check
the orthonormality of the eigenfunctions:
Z Z
1
fn fm ρ = √ fn (a− )m 1 ρ
m! Z
1
= √ (a− )m fn ρ
Z m!
= fn−m ρ = δnm .

From the eigenfunctions and eigenvalues of L we can easily obtain the eigenvalues and eigenfunctions of
L∗ , the Fokker-Planck operator.

74
Lemma 6.3.6. The eigenvalues and eigenfunctions of the Fokker-Planck operator

L∗ · = ∂p2 · +∂p (p·)

are
λ∗n = −n, n = 0, 1, s, . . . and fn∗ = ρfn .

Proof. We have

L∗ (ρfn ) = fn L∗ ρ + ρLfn
= −nρfn.

An immediate corollary of the above calculation is that we can the nth eigenfunction of the Fokker-
Planck operator is given by
1
fn∗ = ρ(p) (a+ )n 1.
n!
Consider now the Fokker-Planck equation with an arbitrary initial condition.
We can relate the generator of the OU process with the Schrödinger operator

6.4 Gradient Flows

Gradient Flows

• Let V (x) = 12 αx2 . The generator of the OU process can be written as:

L = −∂x V ∂x + D∂x2 .

• Consider diffusion processes with a potential V (x), not necessarily quadratic:

L = −∇V (x) · ∇ + D∆. (6.32)

• This is a gradient flow perturbed by noise whose strength is D = kB T where kB is Boltzmann’s

constant and T the absolute temperature.

• The corresponding stochastic differential equation is

√
dXt = −∇V (Xt ) dt + 2D dWt .

• The corresponding FP equation is:

∂p
= ∇ · (∇V p) + D∆p. (6.33)
∂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.

75
Definition 6.4.1. A potential V will be called confining if limx→+∞ V (x) = +∞ and

e−βV (x) ∈ L1 (Rd ). (6.34)

for all β ∈ R+ .

Proposition 6.4.2. Let V (x) be a smooth confining potential. Then the Markov process with genera-
tor (6.32) is ergodic. The unique invariant distribution is the Gibbs distribution
1 −V (x)/D
p(x) = e (6.35)
Z
where the normalization factor Z is the partition function
Z
Z= e−V (x)/D dx.
Rd

• The fact that the Gibbs distribution is an invariant distribution follows by direct substitution. Unique-
ness follows from a PDEs argument (see discussion below).

• It is more convenient to ”normalize” the solution of the Fokker-Planck equation wrt the invariant
distribution.

Theorem 6.4.3. Let p(x, t) be the solution of the Fokker-Planck equation (6.33), assume that (6.34) holds
and let ρ(x) be the Gibbs distribution (6.35). Define h(x, t) through

p(x, t) = h(x, t)ρ(x).

Then the function h satisfies the backward Kolmogorov equation:

∂h
= −∇V · ∇h + D∆h, h(x, 0) = p(x, 0)ρ−1 (x). (6.36)
∂t
Proof. The initial condition follows from the definition of h. We calculate the gradient and Laplacian of p:

∇p = ρ∇h − ρhD −1 ∇V

and
∆p = ρ∆h − 2ρD−1 ∇V · ∇h + hD −1 ∆V ρ + h|∇V |2 D−2 ρ.
We substitute these formulas into the FP equation to obtain
∂h  
ρ = ρ − ∇V · ∇h + D∆h ,
∂t
from which the claim follows.

• Consequently, in order to study properties of solutions to the FP equation, it is sufficient to study the
backward equation (6.36).

• The generator L is self-adjoint, in the right function space.

76
 • We define the weighted L2 space L2ρ :
Z

L2ρ = f| |f |2 ρ(x) dx < ∞ ,
Rd

• where ρ(x) is the Gibbs distribution. This is a Hilbert space with inner product
Z
(f, h)ρ = f hρ(x) dx.
Rd

Theorem 6.4.4. Assume that V (x) is a smooth potential and assume that condition (6.34) holds. Then the
operator
L = −∇V (x) · ∇ + D∆

is self-adjoint in L2ρ . Furthermore, it is non-positive, its kernel consists of constants.

Proof. Let f, ∈ C02 (Rd ). We calculate

Z
(Lf, h)ρ = (−∇V · ∇ + D∆)f hρ dx
Rd
Z Z Z
= (∇V · ∇f )hρ dx − D ∇f ∇hρ dx − D ∇f h∇ρ dx
Rd Z Rd Rd

= −D ∇f · ∇hρ dx,
Rd

from which self-adjointness follows.

If we set f = h in the above equation we get

(Lf, f )ρ = −Dk∇f k2ρ ,

which shows that L is non-positive.

Clearly, constants are in the null space of L. Assume that f ∈ N (L). Then, from the above equation
we get
0 = −Dk∇f k2ρ ,

and, consequently, f is a constant.

Remark

• The expression (−Lf, f )ρ is called the Dirichlet form of the operator L. In the case of a gradient
flow, it takes the form
(−Lf, f )ρ = Dk∇f k2ρ . (6.37)

• 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 the following result.

77
Theorem 6.4.5. Assume that the potential V satisfies the convexity condition

D2 V > λI.

Then the corresponding Gibbs measure satisfies the Poincaré inequality with constant λ:
Z √
f ρ = 0 ⇒ k∇f kρ > λkf kρ . (6.38)
Rd

Theorem 6.4.6. Assume that p(x, 0) ∈ L2 (eV /D ). Then the solution p(x, t) of the Fokker-Planck equa-
tion (6.33) converges to the Gibbs distribution exponentially fast:

kp(·, t) − Z −1 e−V kρ−1 6 e−λDt kp(·, 0) − Z −1 e−V kρ−1 . (6.39)

Proof. We Use (6.36), (6.37) and (6.38) to calculate

 
d 2 ∂h
− k(h − 1)kρ = −2 , h − 1 = −2 (Lh, h − 1)ρ
dt ∂t ρ
= (−L(h − 1), h − 1)ρ = 2Dk∇(h − 1)kρ
> 2Dλkh − 1k2ρ .

Our assumption on p(·, 0) implies that h(·, 0) ∈ L2ρ . Consequently, the above calculation shows that

kh(·, t) − 1kρ 6 e−λDt kH(·, 0) − 1kρ .

This, and the definition of h, p = ρh, lead to (6.39).

• The assumption Z
|p(x, 0)|2 Z −1 eV /D < ∞
Rd

• is very restrictive (think of the case where V = x2 ).

• The function space L2 (ρ−1 ) = L2 (e−V /D ) in which we prove convergence is not the right space to
use. Since p(·, t) ∈ L1 , ideally we would like to prove exponentially fast convergence in L1 .

• We can prove convergence in L1 using the theory of logarithmic Sobolev inequalities. In fact, we
can also prove convergence in relative entropy:
Z  
p
H(p|ρV ) := p ln dx.
R d ρV

• The relative entropy norm controls the L1 norm:

kρ1 − ρ2 kL1 6 CH(ρ1 |ρ2 )

• Using a logarithmic Sobolev inequality, we can prove exponentially fast convergence to equilibrium,
assuming only that the relative entropy of the initial conditions is finite.

78
Theorem 6.4.7. Let p denote the solution of the Fokker–Planck equation (6.33) where the potential is smooth
and uniformly convex. Assume that the the initial conditions satisfy

H(p(·, 0)|ρV ) < ∞.

Then p converges to the Gibbs distribution exponentially fast in relative entropy:

H(p(·, t)|ρV ) 6 e−λDt H(p(·, 0)|ρV ).

• Convergence to equilibrium for kinetic equations, both linear and non-linear (e.g., the Boltzmann
equation) has been studied extensively.

• It has been recognized that the relative entropy plays a very important role.

• For more information see

• On the trend to equilibrium for the Fokker-Planck equation: an interplay between physics and func-
tional analysis by P.A. Markowich and C. Villani, 1999.

6.5 Eigenfunction Expansions

• Consider the generator of a gradient stochastic flow with a uniformly convex potential

L = −∇V · ∇ + D∆. (6.40)

• We know that

i. L is a non-positive self-adjoint operator on L2ρ .

ii. It has a spectral gap:
(Lf, f )ρ 6 −Dλkf k2ρ
where λ is the Poincaré constant of the potential V .

• The above imply that we can study the spectral problem for −L:

−Lfn = λn fn , n = 0, 1, . . .

• The operator −L has real, discrete spectrum with

0 = λ0 < λ1 < λ2 < . . .

• Furthermore, the eigenfunctions {fj }inf ty 2

j=1 form an orthonormal basis in Lρ : we can express every
element of L2ρ in the form of a generalized Fourier series:
∞
X
φ= φn fn , φn = (φ, fn )ρ (6.41)
n=0

79
• with (fn , fm )ρ = δnm .

• This enables us to solve the time dependent Fokker–Planck equation in terms of an eigenfunction
expansion.

• Consider the backward Kolmogorov equation (6.36).

• We assume that the initial conditions h0 (x) = φ(x) ∈ L2ρ and consequently we can expand it in the
form (6.41).

• We look for a solution of (6.36) in the form

∞
X
h(x, t) = hn (t)fn (x).
n=0

• We substitute this expansion into the backward Kolmogorov equation:

∞ ∞
!
∂h X X
= ḣn fn = L hn fn (6.42)
∂t
n=0 n=0
X∞
= −λn hn fn . (6.43)
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
ḣn = −λn hn , n = 0, 1,

• The solution is
h0 (t) = φ0 , hn (t) = e−λn t φn , n = 1, 2, . . .

• Notice that
Z Z
1 = p(x, 0) dx = p(x, t) dx
d Rd
ZR
= h(x, t)Z −1 eβV dx = (h, 1)ρ = (φ, 1)ρ
Rd
= φ0 .

• Consequently, the solution of the backward Kolmogorov equation is

∞
X
h(x, t) = 1 + e−λn t φn fn .
n=1

• 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 exponentially fast.

• The solution of the Fokker–Planck equation is

∞
!
X
p(x, t) = Z −1 e−βV (x) 1+ e−λn t φn fn .
n=1

80
6.6 Self-adjointness
Self-adjointness

• The Fokker–Planck operator of a general diffusion process is not self-adjoint in general.

• In fact, it is self-adjoint if and only if the drift term is the gradient of a potential (Nelson, ≈ 1960).
This is also true in infinite dimensions (Stochastic PDEs).

• Markov Processes whose generator is a self-adjoint operator are called reversible: for all t ∈ [0, T ]
Xt and XT −t have the same transition probability (when Xt is stationary).

• Reversibility is equivalent to the invariant measure being a Gibbs measure.

• See Thermodynamics of the general duffusion process: time-reversibility and entropy production, H.
Qian, M. Qian, X. Tang, J. Stat. Phys., 107, (5/6), 2002, pp. 1129–1141.

6.7 Reduction to a Schrödinger Equation

Reduction to a Schrödinger Equation

Lemma 6.7.1. The Fokker–Planck operator for a gradient flow can be written in the self-adjoint form

∂p   
= D∇ · e−V /D ∇ eV /D p . (6.44)
∂t

Define now ψ(x, t) = eV /2D p(x, t). Then ψ solves the PDE

∂ψ |∇V |2 ∆V
= D∆ψ − U (x)ψ, U (x) := − . (6.45)
∂t 4D 2
Let H := −D∆ + U . Then L∗ and H have the same eigenvalues. The nth eigenfunction φn of L∗ and the
nth eigenfunction ψn of H are associated through the transformation
 
V (x)
ψn (x) = φn (x) exp .
2D

Remarks 6.7.2. i. From equation (6.44) shows that the FP operator can be written in the form
  
L∗ · = D∇ · e−V /D ∇ eV /D · .

ii. The operator that appears on the right hand side of eqn. (6.45) has the form of a Schrödinger oper-
ator:
−H = −D∆ + U (x).

iii. The spectral problem for the FP operator can be transformed into the spectral problem for a Schrödinger
operator. We can thus use all the available results from quantum mechanics to study the FP equation
and the associated SDE.

81
 iv. In particular, the weak noise asymptotics D ≪ 1 is equivalent to the semiclassical approximation
from quantum mechanics.

Proof. We calculate
   

D∇ · e−V /D ∇ eV /D f = D∇ · e−V /D D−1 ∇V f + ∇f eV /D
= ∇ · (∇V f + D∇f ) = L∗ f.

Consider now the eigenvalue problem for the FP operator:

−L∗ φn = λn φn .

1


Set φn = ψn exp − 2D V . We calculate −L∗ φn :
  
−L∗ φn = −D∇ · e−V /D ∇ eV /D ψn e−V /2D
   
−V /D ∇V V /2D
= −D∇ · e ∇ψn + ψn e
2D
   
|∇V |2 ∆V
= −D∆ψn + − + ψn e−V /2D = e−V /2D Hψn .
4D 2D

From this we conclude that e−V /2D Hψn = λn ψn e−V /2D from which the equivalence between the two
eigenvalue problems follows.

Remarks 6.7.3. i. We can rewrite the Schrödinger operator in the form

∇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· = e−U/2D ∇ eU/2D · , A∗ · = eU/2D ∇ e−U/2D ·

iii. The forward the backward Kolmogorov operators have the same eigenvalues. Their eigenfunctions
are related through
φB F
n = φn exp (−V /D) ,

where φB F
n and φn denote the eigenfunctions of the backward and forward operators, respectively.

6.8 The Klein-Kramers-Chandrasekhar Equation

• Consider a diffusion process in two dimensions for the variables q (position) and momentum p. The
generator of this Markov process is

L = p · ∇q − ∇q V ∇p + γ(−p∇p + D∆p ). (6.46)

82
 • The L2 (dpdq)-adjoint is

L∗ ρ = −p · ∇q ρ − ∇q V · ∇p ρ + γ (∇p (pρ) + D∆p ρ) .

• The corresponding FP equation is:

∂p
= L∗ p.
∂t
• The corresponding stochastic differential equations is the Langevin equation
p
Ẍt = −∇V (Xt ) − γ Ẋt + 2γD Ẇt . (6.47)

• This is Newton’s equation perturbed by dissipation and noise.

• The Klein-Kramers-Chandrasekhar equation was first derived by Kramers in 1923 and was studied by
Kramers in his famous paper ”Brownian motion in a field of force and the diffusion model of chemical
reactions”, Physica 7(1940), pp. 284-304.

• Notice that L∗ is not a uniformly elliptic operator: there are second order derivatives only with respect
to p and not q. This is an example of a degenerate elliptic operator. It is, however, hypoelliptic: we
can still prove existence and uniqueness of solutions for the FP equation, and obtain estimates on the
solution.

• It is not possible to obtain the solution of the FP equation for an arbitrary potential.

• We can calculate the (unique normalized) solution of the stationary Fokker-Planck equation.

Theorem 6.8.1. Let V (x) be a smooth confining potential. Then the Markov process with generator (8.6)
is ergodic. The unique invariant distribution is the Maxwell-Boltzmann distribution
1 −βH(p,q)
ρ(p, q) = e (6.48)
Z
where
1
H(p, q) = kpk2 + V (q)
2
is the Hamiltonian, β = (kB T )−1 is the inverse temperature and the normalization factor Z is the
partition function Z
Z= e−βH(p,q) dpdq.
R2d

• It is possible to obtain rates of convergence in either a weighted L2 -norm or the relative entropy norm.

H(p(·, t)|ρ) 6 Ce−αt .

• The proof of this result is very complicated, since the generator L is degenerate and non-selfadjoint.

• See F. Herau and F. Nier, Isotropic hypoellipticity and trend to equilibrium for the Fokker-Planck
equation with a high-degree potential, Arch. Ration. Mech. Anal., 171(2),(2004), 151–218.

83
 Let ρ(q, p, t) be the solution of the Kramers equation and let ρβ (q, p) be the Maxwell-Boltzmann distri-
bution. We can write
ρ(q, p, t) = h(q, p, t)ρβ (q, p),

where h(q, p, t) solves the equation

∂h
= −Ah + γSh (6.49)
∂t
where
A = p · ∇q − ∇q V · ∇p , S = −p · ∇p + β −1 ∆p .

The operator A is antisymmetric in L2ρ := L2 (R2d ; ρβ (q, p)), whereas S is symmetric.

∂
Let Xi := − ∂p i
. The L2ρ -adjoint of Xi is

∂
Xi∗ = −βpi + .
∂pi

We have that
d
X
S = β −1 Xi∗ Xi .
i=1

Consequently, the generator of the Markov process {q(t), p(t)} can be written in Hörmander’s ”sum of
squares” form:
Xd
−1
L = A + γβ Xi∗ Xi . (6.50)
i=1

We calculate the commutators between the vector fields in (6.50):

∂
[A, Xi ] = , [Xi , Xj ] = 0, [Xi , Xj∗ ] = βδij .
∂qi

Consequently,
Lie(X1 , . . . Xd , [A, X1 ], . . . [A, Xd ]) = Lie(∇p , ∇q )

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 L2ρ -adjoint Yi∗ = ∂q∂ i − β ∂q
∂V
i
. We have that
 
∂ ∂V ∂
Xi∗ Yi − Yi∗ Xi = β pi − .
∂qi ∂qi ∂pi

Consequently, the generator can be written in the form

d
X
−1
L=β (Xi∗ Yi − Yi∗ Xi + γXi∗ Xi ) . (6.51)
i=1

Notice also that

d
X
−1 −1
LV := −∇q V ∇q + β ∆q = β Yi∗ Yi .
i=1

84
• The phase-space Fokker-Planck equation can be written in the form
∂ρ
+ p · ∇q ρ − ∇q V · ∇p ρ = Q(ρ, fB )
∂t

• where the collision operator has the form

Q(ρ, fB ) = D∇ · fB ∇ fB−1 ρ .

• 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

• L. Desvillettes and C. Villani: On the trend to global equilibrium for spatially inhomogeneous kinetic
systems: the Boltzmann equation. Invent. Math. 159, 2 (2005), 245-316.

• We can study the backward and forward Kolmogorov equations for (9.7) 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 = p∂q − V ′ (q)∂p + γ −p∂p + ∂p2 .
=: L1 + γL0 ,

• where
L0 := −p∂p + ∂p2 and L1 := p∂q − V ′ (q)∂p .

• The backward Kolmogorov equation is

∂h
= Lh. (6.52)
∂t
• The solution should be an element of the weighted L2 -space
 Z 
2 2 −1 −βH(p,q)
Lρ = f | |f | Z e dpdq < ∞ .
R2

• We notice that the invariant measure of our Markov process is a product measure:
1 2
e−βH(p,q) = e−β 2 |p| e−βV (q) .

1 2
• The space L2 (e−β 2 |p| dp) is spanned by the Hermite polynomials. Consequently, we can expand the
solution of (6.52) into the basis of Hermite basis:
∞
X
h(p, q, t) = hn (q, t)fn (p), (6.53)
n=0

85
 √
• where fn (p) = 1/ n!Hn (p).

• Our plan is to substitute (6.53) into (6.52) 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

• Furthermore
L1 h = −∂q V ∂p h + p∂q h.

• We calculate each term on the right hand side of the above equation separately. For this we will need
the formulas
√ √ √
∂p fn = nfn−1 and pfn = nfn−1 + n + 1fn+1 .

∞
X ∞
X
p∂q h = p∂q hn fn = p∂p h0 + ∂q hn pfn
n=0 n=1
∞
X √ √

= ∂q h0 f1 + ∂q hn nfn−1 + n + 1fn+1
n=1
∞
X √ √
= ( n + 1∂q hn+1 + n∂q hn−1 )fn
n=0

• with h−1 ≡ 0.

• Furthermore
∞
X ∞
X √
∂q V ∂p h = ∂q V hn ∂p fn = ∂q V hn nfn−1
n=0 n=0
X∞
√
= ∂q V hn+1 n + 1fn .
n=0

• Consequently:

Lh = L1 + γL1 h
∞
X √
= − γnhn + n + 1∂q hn+1
n=0
√ √

+ n∂q hn−1 + n + 1∂q V hn+1 fn

86
 • Using the orthonormality of the eigenfunctions of L0 we obtain the following set of equations which
determine {hn (q, t)}∞
n=0 .
√
ḣn = −γnhn + n + 1∂q hn+1
√ √
+ n∂q hn−1 + n + 1∂q V hn+1 , n = 0, 1, . . .

• 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 equation.

• For this we need to expand each coefficient hn in an appropriate basis with respect to q.

• Obvious choices are other the Hermite basis (polynomial potentials) or the standard Fourier basis
(periodic potentials).

• We will do this for the case of periodic potentials.

• The resulting method is usually called the continued fraction expansion. See Risken (1989).

• The Hermite expansion of the distribution function wrt to the velocity is used in the study of various
kinetic equations (including the Boltzmann equation). It was initiated by Grad in the late 40’s.

• It quite often used in the approximate calculation of transport coefficients (e.g. diffusion coefficient).

• This expansion can be justified rigorously for the Fokker-Planck equation. See

• J. Meyer and J. Schröter, Comments on the Grad Procedure for the Fokker-Planck Equation, J. Stat.
Phys. 32(1) pp.53-69 (1983).

• This expansion can also be used in order to solve the Poisson equation −Lφ = f (p, q). See G.A.
Pavliotis and T. Vogiannou Diffusive Transport in Periodic Potentials: Underdamped dynamics, Fluct.
Noise Lett., 8(2) L155-L173 (2008).

6.9 Brownian Motion in a Harmonic Potential

There are very few potentials for which we can solve the Langevin equation or to calculate the eigenvalues
and eigenfunctions of the generator of the Markov process {q(t), p(t)}. One case where we can calculate
everything explicitly is that of a Brownian particle in a quadratic (harmonic) potential
1
V (q) = ω02 q 2 . (6.54)
2
The Langevin equation is p
q̈ = −ω02 q − γ q̇ + 2γβ −1 Ẇ (6.55)
or p
q̈ = p, ṗ = −ω02 q − γp + 2γβ −1 Ẇ . (6.56)

87
This is a linear equation that can be solved explicitly. Rather than doing this, we will calculate the eigenval-
ues and eigenfunctions of the generator, which takes the form

L = p∂q − ω02 q∂p + γ(−p∂p + β −1 ∂p2 ). (6.57)

The Fokker-Planck operator is

L = p∂q − ω02 q∂p + γ(−p∂p + β −1 ∂p2 ). (6.58)

The process {q(t), p(t)} is an ergodic Markov process with Gaussian invariant measure
2
βω0 − β2 p2 − βω 0
q2 .
ρβ (q, p) dqdp = e (6.59)
2π
For the calculation of the eigenvalues and eigenfunctions of the operator L it is convenient to introduce
creation and annihilation operator in both the position and momentum variables. We set

a− = β −1/2 ∂p , a+ = −β −1/2 ∂p + β 1/2 p (6.60)

and
b− = ω0−1 β −1/2 ∂q , b+ = −ω0−1 β −1/2 ∂q + ω0 β 1/2 p. (6.61)
We have that
a+ a− = −β −1 ∂p2 + p∂p
and
b+ b− = −β −1 ∂q2 + q∂q
Consequently, the operator
Lb = −a+ a− − b+ b− (6.62)
is the generator of the OU process in two dimensions.
b Show that there exists a transformation
Exercise 6.9.1. Calculate the eigenvalues and eigenfunctions of L.
that transforms Lb into the Schr”odinger operator of the two-dimensional quantum harmonic oscillator.
Exercise 6.9.2. Show that the operators a± , b± satisfy the commutation relations

[a+ , a− ] = −1, (6.63a)

[b+ , b− ] = −1, (6.63b)

[a± , b± ] = 0. (6.63c)
Using now the operators a± and b± we can write the generator L in the form

L = −γa+ a− − ω0 (b+ a− − a+ b− ), (6.64)

which is a particular case of (6.51). In order to calculate the eigenvalues and eigenfunctions of (6.64) we
need to make an appropriate change of variables in order to bring the operator L into the ”decoupled”
form (6.62). Clearly, this is a linear transformation and can be written in the form

Y = AX

88
where X = (q, p) for some 2 × 2 matrix A. It is somewhat easier to make this change of variables at
the level of the creation and annihilation operators. In particular, our goal is to find first order differential
operators c± and d± so that the operator (6.64) becomes

L = −Cc+ c− − Dd+ d− (6.65)

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, (6.66a)

[d+ , d− ] = −1, (6.66b)

[c± , d± ] = 0. (6.66c)
The operators c± and d± should be given as linear combinations of the old operators a± and b± . From
the structure of the generator L (6.64), the decoupled form (6.65) and the commutation relations (6.66)
and (6.63) we conclude that c± and d± should be of the form

c+ = α11 a+ + α12 b+ , (6.67a)

c− = α21 a− + α22 b− , (6.67b)

d+ = β11 a+ + β12 b+ , (6.67c)
d− = β21 a− + β22 b− . (6.67d)
Notice that the c− and d− are not the adjoints of c+ and d+ . If we substitute now these equations into (6.65)
and equate it with (6.64) and into the commutation relations (6.66) we obtain a system of equations for the
coefficients {αij }, {βij }. In order to write down the formulas for these coefficients it is convenient to
introduce the eigenvalues of the deterministic problem

q̈ = −γ q̇ − ω02 q.

The solution of this equation is

q(t) = C1 e−λ1 t + C2 e−λ2 t
with q
γ±δ
λ1,2 = , δ= γ 2 − 4ω02 . (6.68)
2
The eigenvalues satisfy the relations

λ1 + λ2 = γ, λ1 − λ2 = δ, λ1 λ2 = ω02 . (6.69)

Proposition 6.9.3. Let L be the generator (6.64) and let c± , dpm be the operators
1 p p 
c+ = √ λ1 a+ + λ2 b+ , (6.70a)
δ
1 p p 
c− = √ λ1 a− − λ2 b− , (6.70b)
δ

89
 1 p p 
d+ = √ λ2 a+ + λ1 b+ , (6.70c)
δ
1  p p 
d− = √ − λ2 a− + λ1 b− . (6.70d)
δ
± ±
Then c , d satisfy the canonical commutation relations (6.66) as well as

[L, c± ] = −λ1 c± , [L, d± ] = −λ2 d± . (6.71)

Furthermore, the operator L can be written in the form

L = −λ1 c+ c− − λ2 d+ d− . (6.72)

Proof. first we check the commutation relations:

1

[c+ , c− ] = λ1 [a+ , a− ] − λ2 [b+ , b− ]
δ
1
= (−λ1 + λ2 ) = −1.
δ
Similarly,
1

[d+ , d− ] = −λ2 [a+ , a− ] + λ1 [b+ , b− ]
δ
1
= (λ2 − λ1 ) = −1.
δ
Clearly, we have that
[c+ , d+ ] = [c− , d− ] = 0.
Furthermore,
1 p p 
[c+ , d− ] = − λ1 λ2 [a+ , a− ] + λ1 λ2 [b+ , b− ]
δ
1 p p
= (− λ1 λ2 + − λ1 λ2 ) = 0.
δ
Finally:

[L, c+ ] = −λ1 c+ c− c+ + λ1 c+ c+ c−
= −λ1 c+ (1 + c+ c− ) + λ1 c+ c+ c−
= −λ1 c+ (1 + c+ c− ) + λ1 c+ c+ c−
= −λ1 c+ ,

and similarly for the other equations in (6.71). Now we calculate

L = −λ1 c+ c− − λ2 d+ d−
√
λ22 − λ21 + − + − λ1 λ2 1p
= − a a + 0b b + (λ1 − λ2 )a+ b− + λ1 λ2 (−λ1 + λ2 )b+ a−
δ δ δ
= −γa+ a− − ω0 (b+ a− − a+ b− ),

which is precisely (6.64). In the above calculation we used (6.69).

90
 Using now (6.72) we can readily obtain the eigenvalues and eigenfunctions of L. From our experience
with the two-dimensional OU processes (or, the Schrödinger 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 6.9.4. The eigenvalues and eigenfunctions of the generator of the Markov process {q, p} (6.56)
are
1 1
λnm = λ1 n + λ2 m = γ(n + m) + δ(n − m), n, m = 0, 1, . . . (6.73)
2 2
and
1
φnm (q, p) = √ (c+ )n (d+ )m 1, n, m = 0, 1, . . . (6.74)
n!m!
Proof. We have

[L, (c+ )2 ] = L(c+ )2 − (c+ )2 L

= (c+ L − λ1 c+ )c+ − c+ (Lc+ + λ1 c+ )
= −2λ1 (c+ )2

and similarly [L, (d+ )2 ] = −2λ1 (c+ )2 . A simple induction argument now shows that (see Exercise 6.9.5)

[L, (c+ )n ] = −nλ1 (c+ )n and [L, (d+ )m ] = −mλ1 (d+ )m . (6.75)

We use (6.75) to calculate

L(c+ )n (d+ )n 1
= (c+ )n L(d+ )m 1 − nλ1 (c+ )n (d+ m)1
= (c+ )n (d+ )m L1 − mλ2 (c+ )n (d+ m)1 − nλ1 (c+ )n (d+ m)1
= −nλ1 (c+ )n (d+ m)1 − mλ2 (c+ )n (d+ m)1

from which (6.73) and (6.74) follow.

Exercise 6.9.5. Show that

[L, (c± )n ] = −nλ1 (c± )n , [L, (d± )n ] = −nλ1 (d± )n , [c− , (c+ )n ] = n(c+ )n−1 , [d− , (d+ )n ] = n(d+ )n−1 .
(6.76)

Remark 6.9.6. In terms of the operators a± , b± the eigenfunctions of L are

n X
m   k−ℓ
√ − n+m n/2 m/2
X 1 λ1 2
φnm = n!m!δ 2 λ1 λ2 (a+ )n+m−k−ℓ (b+ )ℓ+k 1.
k!(m − k)!ℓ!(n − ℓ)! λ2
ℓ=0 k=0

The first few eigenfunctions are

φ00 = 1.

√ √ √

β λ1 p + λ2 ω0 q
φ10 = √ .
δ

91
 √ √ √

β λ2 p + λ1 ω0 q
φ01 = √
δ

√ √ √ √ √ √
−2 λ1 λ2 + λ1 β p2 λ2 + β pλ1 ω0 q + ω0 β qλ2 p + λ2 ω0 2 β q 2 λ1
φ11 = .
δ
√ √
−λ1 + β p2 λ1 + 2 λ2 β p λ1 ω0 q − λ2 + ω0 2 β q 2 λ2
φ20 = √ .
2δ

√ √
−λ2 + β p2 λ2 + 2 λ2 β p λ1 ω0 q − λ1 + ω0 2 β q 2 λ1
φ02 = √ .
2δ
Notice that the eigenfunctions are not orthonormal.
As we already know, the first eigenvalue, corresponding to the constant eigenfunction, is 0:

λ00 = 0.

Notice that the operator L is not self-adjoint and consequently, we do not expect its eigenvalues to be real.
Indeed, whether the eigenvalues are real or not depends on the sign of the discriminant ∆ = γ 2 − 4ω02 . In
the underdamped regime, γ < 2ω0 the eigenvalues are complex:
q
1 1
λnm = γ(n + m) + i −γ 2 + 4ω02 (n − m), γ < 2ω0 .
2 2
This it to be expected, since the underdamped regime the dynamics is dominated by the deterministic Hamil-
p
tonian dynamics that give rise to the antisymmetric Liouville operator. We set ω = (4ω02 − γ 2 ), i.e.
δ = 2iω. The eigenvalues can be written as
γ
λnm = (n + m) + iω(n − m).
2
In Figure 6.9 we present the first few eigenvalues of L in the underdamped regime. The eigenvalues are
contained in a cone on the right half of the complex plane. The cone is determined by
γ γ
λn0 = n + iωn and λ0m = m − iωm.
2 2
The eigenvalues along the diagonal are real:

λnn = γn.

On the other hand, in the overdamped regime, γ > 2ω0 all eigenvalues are real:
q
1 1
λnm = γ(n + m) + γ 2 − 4ω02 (n − m), γ > 2ω0 .
2 2
In fact, in the overdamped limit γ → +∞ (which we will study in Chapter 8), the eigenvalues of the
generator L converge to the eigenvalues of the generator of the OU process:

ω02
λnm = γn + (n − m) + O(γ −3 ).
γ

92
 3

1
Im (λnm)

−1

−2

−3
0 0.5 1 1.5 2 2.5 3
Re (λ )
nm

Figure 6.1: First few eigenvalues of L for γ = ω = 1.

93
This is consistent with the fact that in this limit the solution of the Langevin equation converges to the
solution of the OU SDE. See Chapter 8 for details.
The eigenfunctions of L do not form an orthonormal basis in L2β := L2 (R2 , Z −1 e−βH ) since L is not
a selfadjoint operator. Using the eigenfunctions/eigenvalues of L we can easily calculate the eigenfunc-
tions/eigenvalues of the L2β adjoint of L. From the calculations presented in Section 6.8 we know that the
adjoint operator is

Lb := −A + γS
= −ω0 (b+ a− − b− a+ ) + γa+ a−
= −λ1 (c− )∗ (c+ )∗ − λ2 (d− ) ∗ (d+ )∗,

where
1 p p 
(c+ )∗ = √ λ1 a− + λ2 b− , (6.77a)
δ
1 p p 
(c− )∗ = √ λ1 a+ − λ2 b+ , (6.77b)
δ
1 p p 
(d+ )∗ = √ λ2 a− + λ1 b− , (6.77c)
δ
1  p p 
(d− )∗ = √ − λ2 a+ + λ1 b+ . (6.77d)
δ
Exercise 6.9.7. i. Show by direct substitution that Lb can be written in the form

Lb = −λ1 (c− )∗ (c+ )∗ − λ2 (d− ) ∗ (d+ ) ∗ .

ii. Calculate the commutators

[(c+ )∗ , (c− )∗ ], [(d+ )∗ , (d− )∗ ], [(c± )∗ , (d± )∗ ], [L,

b (c± )∗ ], [L,
b (d± )∗ ].

Lb has the same eigenvalues as L:

b nm = λnm ψnm ,
−Lψ
where λnm are given by (6.73). The eigenfunctions are
1
ψnm = √ ((c− )∗ )n ((d− )∗ )m 1. (6.78)
n!m!

Proposition 6.9.8. The eigenfunctions of L and Lb satisfy the biorthonormality relation

Z Z
φnm ψℓk ρβ dpdq = δnℓ δmk . (6.79)

Proof. We will use formulas (6.76). 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 > ℓ)

(c− )ℓ (c+ )n 1 = n(n − 1) . . . (n − ℓ + 1)(c+ )n−ℓ . (6.80)

94
We have
Z Z Z Z
1
φnm ψℓk ρβ dpdq = √ ((c+ ))n ((d+ ))m 1((c− )∗ )ℓ ((d− )∗ )k 1ρβ dpdq
n!m!ℓ!k!
Z Z
n(n − 1) . . . (n − ℓ + 1)m(m − 1) . . . (m − k + 1)
= √ ((c+ ))n−ℓ ((d+ ))m−k 1ρβ dpdq
n!m!ℓ!k!
= δnℓ δmk ,

since all eigenfunctions average to 0 with respect to ρβ .

From the eigenfunctions of Lb we can obtain the eigenfunctions of the Fokker-Planck operator. Using
the formula (see equation (6.49))
L∗ (f ρβ ) = ρLf
b
b
we immediately conclude that the the Fokker-Planck operator has the same eigenvalues as those of L and L.
The eigenfunctions are
1
ψ∗nm = ρβ φnm = ρβ √ ((c− )∗ )n ((d− )∗ )m 1. (6.81)
n!m!

6.10 Discussion and Bibliography

The proof of existence and uniqueness of classical solutions for the Fokker-Planck equation of a uniformly
elliptic diffusion process with smooth drift and diffusion coefficients, Theorem 6.1.2, can be found in [7].
A standard textbook on PDEs, with a lot of material on parabolic PDEs is [4]. The Fokker-Planck equation
is studied extensively in Risken’s monograph [20]. See also [10] and [14]. The connection between the
Fokker-Planck equation and stochastic differential equations is presented in Chapter 7. See also [1, 8, 9].
Diffusion processes in one dimension are studied in [16]. The Feller classification for one dimensional
diffusion processes can be also found in [15, 5].

95
96
Chapter 7

Stochastic Differential Equations

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 m–dimensional Brownian motion, h : Z → Rd a smooth vector-valued

function and γ : Z → Rd×m a smooth matrix valued function (in this course we will take Z = Td , Rd
or Rl ⊕ Td−l .

• Consider the SDE

dz dW
= h(z) + γ(z) , z(0) = z0 . (7.1)
dt dt
• We think of the term dWdt as representing Gaussian white noise: a mean-zero Gaussian process with
correlation δ(t − s)I.

• The function h in (7.1) is sometimes referred to as the drift and γ as the diffusion coefficient.

• Such a process exists only as a distribution. The precise interpretation of (7.1) is as an integral equation
for z(t) ∈ C(R+ , Z):
Z t Z t
z(t) = z0 + h(z(s))ds + γ(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).

7.2 The Itô Stochastic Integral

The Itô Stochastic Integral

• For the rigorous analysis of stochastic differential equations it is necessary to define stochastic inte-
grals of the form Z t
I(t) = f (s) dW (s), (7.3)
0

97
 • where W (t) is a standard one dimensional Brownian motion. This is not straightforward 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
filtration Ft generated by the process W (t), and such that
Z T 
2
E f (s) ds < ∞.
0

• The Itô stochastic integral I(t) is defined as the L2 –limit of the Riemann sum approximation of (7.3):
K−1
X
I(t) := lim f (tk−1 ) (W (tk ) − W (tk−1 )) , (7.4)
K→∞
k=1

• where tk = k∆t and K∆t = t.

• Notice that the function f (t) is evaluated at the left end of each interval [tn−1 , tn ] in (7.4).

• The resulting Itô stochastic integral I(t) is a.s. continuous in t.

• These ideas are readily generalized to the case where W (s) is a standard d dimensional Brownian
motion and f (s) ∈ Rm×d for each s.

• The resulting integral satisfies the Itô isometry

Z t
E|I(t)|2 = E|f (s)|2F ds, (7.5)
0
p
• where | · |F denotes the Frobenius norm |A|F = tr(AT A).

• The Itô stochastic integral is a martingale:

EI(t) = 0

• and
E[I(t)|Fs ] = I(s) ∀ t > s,
where Fs denotes the filtration generated by W (s).

Example 7.2.1. • Consider the Itô stochastic integral

Z t
I(t) = f (s) dW (s),
0

• where f, W are scalar–valued. This is a martingale with quadratic variation

Z t
hIit = (f (s))2 ds.
0

• More generally, for f, W in arbitrary finite dimensions, the integral I(t) is a martingale with quadratic
variation Z t
hIit = (f (s) ⊗ f (s)) ds.
0

98
7.2.1 The Stratonovich Stochastic Integral
The Stratonovich Stochastic Integral

• In addition to the Itô stochastic integral, we can also define the Stratonovich stochastic integral. It is
defined as the L2 –limit of a different Riemann sum approximation of (7.3), namely

1 
K−1
X
Istrat (t) := lim f (tk−1 ) + f (tk ) (W (tk ) − W (tk−1 )) , (7.6)
K→∞ 2
k=1

• where tk = k∆t and K∆t = t. Notice that the function f (t) is evaluated at both endpoints of each
interval [tn−1 , tn ] in (7.6).

• The multidimensional Stratonovich integral is defined in a similar way. The resulting integral is
written as Z t
Istrat (t) = f (s) ◦ dW (s).
0

• The limit in (7.6) gives rise to an integral which differs from the Itô integral.

• The situation is more complex than that arising in the standard theory of Riemann integration for
functions of bounded variation: in that case the points in [tk−1 , 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.

7.3 Existence and Uniqueness of solutions for SDEs

Existence and Uniqueness of solutions for SDEs

Definition 7.3.1. By a solution of (7.1) we mean a Z-valued stochastic process {z(t)} on t ∈ [0, T ] with
the properties:

i. z(t) is continuous and Ft −adapted, where the filtration is generated by the Brownian motion W (t);

ii. h(z(t)) ∈ L1 ((0, T )), γ(z(t)) ∈ L2 ((0, T ));

iii. equation (7.1) holds for every t ∈ [0, T ] with probability 1.

The solution is called unique if any two solutions xi (t), i = 1, 2 satisfy

P(x1 (t) = x2 (t), ∀t ∈ [0.T ]) = 1.

• It is well known that existence and uniqueness of solutions for ODEs (i.e. when γ ≡ 0 in (7.1)) holds
for globally Lipschitz vector fields h(x).

99
 • A very similar theorem holds when γ 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 γ(·) 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

Furthermore, the solution of the SDE is a Markov process.

Remarks 7.3.3. • The Stratonovich analogue of (7.1) is

dz dW
= h(z) + γ(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 + γ(z(s)) ◦ dW (s). (7.8)
0 0

• By using definitions (7.4) and (7.6) it can be shown that z satisfying the Stratonovich SDE (7.7) also
satisfies the Itô SDE
dz 1
1
dW
= h(z) + ∇ · γ(z)γ(z)T − γ(z)∇ · γ(z)T + γ(z) , (7.9a)
dt 2 2 dt
z(0) = z0 , (7.9b)

• provided that γ(z) is differentiable.

• White noise is, in most applications, an idealization of a stationary random process with short correla-
tion 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 Itô integral makes it more important
as a mathematical object.

• It is very useful that we can convert from the Itô 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),

100
 • 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) + √ γ(z) , z(0) = z0 .
ds c c ds

7.4 Derivation of the Stratonovich SDE

The Stratonovich Stochastic Integral: A first application of multiscale methods

• When white noise is approximated by a smooth process this often leads to Stratonovich interpretations
of stochastic integrals, at least in one dimension.

• We use multiscale analysis (singular perturbation theory for Markov processes) to illustrate this phe-
nomenon in a one-dimensional example.

• Consider the equations

dx 1
= h(x) + f (x)y, (7.10a)
dt ε
r
dy αy 2D dV
=− 2 + , (7.10b)
dt ε ε2 dt
• with V being a standard one-dimensional Brownian motion.

• We say that the process x(t) is driven by colored noise: the noise that appears in (7.10a) has non-zero
correlation time. The correlation function of the colored noise η(t) := y(t)/ε is (we take y(0) = 0)
1 D − α2 |t−s|
R(t) = E (η(t)η(s)) = e ε .
ε2 α

• The power spectrum of the colored noise η(t) is:

1 Dε−2 1
f ε (x) = 2
ε π x + (αε−2 )2
2

D 1 D
= 4 2 2
→
π ε x +α πα2

• and, consequently,  
y(t) y(s) 2D
lim E = δ(t − s),
ε→0 ε ε α2
• which implies the heuristic r
y(t) 2D dV
lim = . (7.11)
ε→0 ε α2 dt

101
 • Another way of seeing this is by solving (7.10b) for y/ε:
r
y 2D dV ε dy
= − . (7.12)
ε α2 dt α dt

• If we neglect the O(ε) term on the right hand side then we arrive, again, at the heuristic (7.11).

• Both of these arguments lead us to conjecture the limiting Itô SDE:

r
dX 2D dV
= h(X) + f (X) . (7.13)
dt α 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 inter-
preted in the Stratonovich sense, giving
r
dX 2D dV
= h(X) + f (X) ◦ . (7.14)
dt α 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 7.4.1. Assume that the initial conditions for y(t) are stationary and that the function f is smooth.
Then the solution of eqn (7.10a) converges, in the limit as ε → 0 to the solution of the Stratonovich
SDE (7.14).

Remarks

i. It is possible to prove pathwise convergence under very mild assumptions.

ii. The generator of a Stratonovich SDE has the from

D
Lstrat = h(x)∂x + f (x)∂x (f (x)∂x ) .
α
iii. Consequently, the Fokker-Planck operator of the Stratonovich SDE can be written in divergence form:
D

L∗strat · = −∂x (h(x)·) + ∂x f 2 (x)∂x · .
α
iv. In most applications in physics the white noise is an approximation of a more complicated noise
processes with non-zero correlation time. Hence, the physically correct interpretation of the stochastic
integral is the Stratonovich one.

v. In higher dimensions an additional drift term might appear due to the noncommutativity of the row
vectors of the diffusion matrix. This is related to the Lévy area correction in the theory of rough
paths.

102
 Proof of Proposition 7.4.1 The generator of the process (x(t), y(t)) is

1 2

1
L = −αy∂ y + D∂y + f (x)y∂x + h(x)∂x
ε2 ε
1 1
=: L0 + L1 + L2 .
ε2 ε
The ”fast” process is an stationary Markov process with invariant density
r
α − αy2
ρ(y) = e 2D . (7.15)
2πD
The backward Kolmogorov equation is
 
∂uε 1 1
= L0 + L1 + L2 uε . (7.16)
∂t ε2 ε

We look for a solution to this equation in the form of a power series expansion in ε:

uε (x, y, t) = u0 + εu1 + ε2 u2 + . . .

We substitute this into (7.16) and equate terms of the same power in ε 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).
The second equation in the hierarchy becomes

−L0 u1 = f (x)y∂x 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 + ψ1 (x, t).
α
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 L∗0 : Z  
∂u0
L1 u1 + L2 u0 − ρ(y) dy = 0.
R ∂t
We calculate: Z
∂u0 ∂u
ρ(y) dy = .
R ∂t ∂t

103
Furthermore:
Z
L2 u0 ρ(y) dy = h(x)∂x u.
R
Finally
Z Z  
1
L1 u1 ρ(y) dy = f (x)y∂x f (x)∂x uy + ψ1 (x, t) ρ(y) dy
R R α
1
= f (x)∂x (f (x)∂x u) hy 2 i + f (x)∂x ψ1 (x, t)hyi
α
D
= f (x)∂x (f (x)∂x u)
α2
D D
= 2
f (x)∂x f (x)∂x u + 2 f (x)2 ∂x2 u.
α α
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 α α
from which we read off the limiting Stratonovich SDE
r
dX 2D dV
= h(X) + f (X) ◦ .
dt α dt

7.5 Itô versus Stratonovich

• A Stratonovich SDE
dX(t) = f (X(t)) dt + σ(X(t)) ◦ dW (t) (7.17)

• can be written as an Itô SDE

   
1 dσ
dX(t) = f (X(t)) + σ (X(t)) dt + σ(X(t)) dW (t).
2 dx

• Conversely, and Itô SDE

dX(t) = f (X(t)) dt + σ(X(t))dW (t) (7.18)

• can be written as a Statonovich SDE

   
1 dσ
dX(t) = f (X(t)) − σ (X(t)) dt + σ(X(t)) ◦ dW (t).
2 dx

• The Itô and Stratonovich interpretation of an SDE can lead to equations with very different properties!

• Multiplicative Noise.

– When the diffusion coefficient depends on the solution of the SDE X(t), we will say that we
have an equation with multiplicative noise.

104
 – Multiplicative noise can lead to noise induced phase transitions. See
∗ W. Horsthemke and R. Lefever, Noise-induced transitions, Springer-Verlag, Berlin 1984.
– This is a topic of current interest for SDEs in infinite dimensions (SPDEs).

Colored Noise
• When the noise which drives an SDE has non-zero correlation time we will say that we have colored
noise.

• The properties of the SDE (stability, ergodicity etc.) are quite robust under ”coloring of the noise”.
See

– G. Blankenship and G.C. Papanicolaou, Stability and control of stochastic systems with wide-
band noise disturbances. I, SIAM J. Appl. Math., 34(3), 1978, pp. 437–476.

• Colored noise appears in many applications in physics and chemistry. For a review see

– P. Hanggi and P. Jung Colored noise in dynamical systems. Adv. Chem. Phys. 89 239 (1995).

• In the case where there is an additional small time scale in the problem, in addition to the correlation
time of the colored noise, it is not clear what the right interpretation of the stochastic integral (in the
limit as both small time scales go to 0). This is usually called the Itô versus Stratonovich problem.

• Consider, for example, the SDE

τ Ẍ = −Ẋ + v(X)η ε (t),

• where η ε (t) is colored noise with correlation time ε2 .

• In the limit where both small time scales go to 0 we can get either Itô or Stratonovich or neither. See

– 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), 2005, pp. 1-35.

7.6 The Generator

• Given the function γ(z) in the SDE (7.1) we define
Γ(z) = γ(z)γ(z)T . (7.19)

• The generator L is then defined as

1
Lv = h · ∇v + Γ : ∇∇v. (7.20)
2
• This operator, equipped with a suitable domain of definition, is the generator of the Markov process
given by (7.1).

• The formal L2 −adjoint operator L∗

1
L∗ v = −∇ · (hv) + ∇ · ∇ · (Γv).
2

105
7.7 The Itô Formula

• The Itô formula enables us to calculate the rate of change in time of functions V : Z → Rn evaluated
at the solution of a Z-valued SDE.

• Formally, we can write:

 
d  dW
V (z(t)) = LV (z(t)) + ∇V (z(t)), γ(z(t)) .
dt dt

• 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 Γ, which arises from the lack of smoothness of Brownian
motion.

• The precise interpretation of the expression for the rate of change of V is in integrated form:

Lemma 7.7.1. (Itô’s 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 + h∇V (z(s)), γ(z(s)) dW (s)i .
0 0

• Let φ : Z 7→ R and consider the function

v(z, t) = E φ(z(t))|z(0) = z , (7.21)

• where the expectation is with respect to all Brownian driving paths. By averaging in the Itô formula,
which removes the stochastic integral, and using the Markov property, it is possible to obtain the
Backward Kolmogorov equation.

• For a Stratonovich SDE the rules of standard calculus apply:

– Consider the Stratonovich SDE (7.17) and let V (x) ∈ C 2 (R). Then

dV
dV (X(t)) = (X(t)) (f (X(t)) dt + σ(X(t)) ◦ dW (t)) .
dx

– Consider the Stratonovich SDE (7.17) on Rd (i.e. f ∈ Rd , σ : Rn 7→ Rd , W (t) is standard

Brownian motion on Rn ). The corresponding Fokker-Planck equation is:

∂ρ 1
= −∇ · (f ρ) + ∇ · (σ∇ · (σρ))). (7.22)
∂t 2

106
7.8 Examples of SDEs
i. The SDE for Brownian motion is:
√
dX = 2σdW, X(0) = x.

• The Solution is:

X(t) = x + W (t).

ii. The SDE for the Ornstein-Uhlenbeck process is

√
dX = −αX dt + 2λ dW, X(0) = x.

• We can solve this equation using the variation of constants formula:

√ Z t
−αt
X(t) = e x+ 2λ e−α(t−s) dW (s).
0

• We can use Itô’s formula to obtain equations for the moments of the OU process. The generator
is:
L = −αx∂x + λ∂x2 .

• We apply Itô’s formula to the function f (x) = xn to obtain:

√
dX(t)n = LX(t)n dt + 2λ∂X(t)n dW
√
= −αnX(t)n dt + λn(n − 1)X(t)n−2 dt + n 2λX(t)n−1 dW.

• Consequently:
Z t
n n


X(t) = x + −αnX(t)n + λn(n − 1)X(t)n−2 dt
0
√ Z t
+n 2λ X(t)n−1 dW.
0

• By taking the expectation in the above equation we obtain the equation for the moments of the OU
process that we derived earlier using the Fokker-Planck equation:
Z t
n
Mn (t) = x + (−αnMn (s) + λn(n − 1)Mn−2 (s)) ds.
0

• Consider the geometric Brownian motion

dX(t) = µX(t) dt + σX(t) dW (t), (7.23)

107
 – where we use the Itô interpretation of the stochastic differential. The generator of this process is

σ 2 x2 2
L = µx∂x + ∂ .
2 x
– The solution to this equation is
 
σ2
X(t) = X(0) exp (µ − )t + σW (t) . (7.24)
2

• To derive this formula, we apply Itô’s formula to the function f (x) = log(x):
 
d log(X(t)) = L log(X(t)) dt + σx∂x log(X(t)) dW (t)
  
1 σ 2 x2 1
= µx + − 2 dt + σ dW (t)
x 2 x
 
σ2
= µ− dt + σ dW (t).
2

• Consequently:    
X(t) σ2
log = µ− t + σW (t)
X(0) 2
• from which (7.24) follows.

• Notice that the Stratonovich interpretation of this equation leads to the solution

X(t) = X(0) exp(µt + σW (t))

• Exercise: calculate all moments of the geometric Brownian motion for the Itô and Stratonovich inter-
pretations of the stochastic integral.

7.8.1 The Stochastic Landau Equation

• Consider the Landau equation:
dXt
= Xt (c − Xt2 ), X0 = x. (7.25)
dt

• This is a gradient flow for the potential V (x) = 12 cx2 − 14 x4 .

• When c < 0 all solutions are attracted to the single steady state X∗ = 0.
√ √
• When c > 0 the steady state X∗ = 0 becomes unstable and Xt → c if x > 0 and Xt → − c if
x < 0.

• Consider additive random perturbations to the Landau equation:

dXt √ dWt
= Xt (c − Xt2 ) + 2σ , X0 = x. (7.26)
dt dt

108
• This equation defines an ergodic Markov process on R: There exists a unique invariant distribution:
Z
1 1
ρ(x) = Z −1 e−V (x)/σ , Z = e−V (x)/σ dx, V (x) = cx2 − x4 .
R 2 4

• ρ(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.

• Consider now multiplicative perturbations of the Landau equation.

dXt √ dWt
= Xt (c − Xt2 ) + 2σXt , X0 = x. (7.27)
dt dt

• Where the stochastic differential is interpreted in the Itô sense.

• The generator of this process is

L = x(c − x2 )∂x + σx2 ∂x2 .

• Notice that Xt = 0 is always a solution of (7.27). Thus, if we start with x > 0 (x < 0) the solution
will remain positive (negative).

• We will assume that x > 0.

• Consider the function Yt = log(Xt ). We apply Itô’s formula to this function:

dYt = L log(Xt ) dt + σXt ∂x log(Xt ) dWt

 
2 1 2 1 1
= Xt (c − Xt ) − σXt 2 dt + σXt dWt
Xt Xt Xt
= (c − σ) dt − Xt2 dt + σ dWt .

• Thus, we have been able to transform (7.27) into an SDE with additive noise:
h i
dYt = (c − σ) − e2Yt dt + σ dWt . (7.28)

This is a gradient flow with potential

h 1 i
V (y) = − (c − σ)y − e2y .
2

• The invariant measure, if it exists, is of the form

ρ(y) dy = Z −1 e−V (y)/σ dy.

109
 • Going back to the variable x we obtain:
x2
ρ(x) dx = Z −1 x(c/σ−2) e− 2σ dx.

• We need to make sure that this distribution is integrable:

Z +∞
x2 c
Z= xγ e− 2σ < ∞, γ= − 2.
0 σ

• For this it is necessary that

γ > −1 ⇒ c > σ.

• Not all multiplicative random perturbations lead to ergodic behavior!

• The dependence of the invariant distribution on c is similar to the physical situation of first order phase
transitions.

• Exercise Analyze this problem for the Stratonovich interpretation of the stochastic integral.

• Exercise Study additive and multiplicative random perturbations of the ODE

dx
= x(c + 2x2 − x4 ).
dt

• For more information see M.C. Mackey, A. Longtin, A. Lasota Noise-Induced Global Asymptotic
Stability, J. Stat. Phys. 60 (5/6) pp. 735-751.

7.9 The Backward Kolmogorov Equation

Theorem 7.9.1. Assume that φ is chosen sufficiently smooth so that the backward Kolmogorov equation
∂v
= Lv for (z, t) ∈ Z × (0, ∞),
∂t
v = φ for (z, t) ∈ Z × {0} , (7.29)

has a unique classical solution v(x, t) ∈ C 2,1 (Z × (0, ∞), ). Then v is given by (7.21) where z(t) solves
(7.2).

7.9.1 Derivation of the Fokker-Planck Equation

• Now we can derive rigorously the Fokker-Planck equation.

Theorem 7.9.2. Consider equation (7.2) with z(0) a random variable with density ρ0 (z). Assume that the
law of z(t) has a density ρ(z, t) ∈ C 2,1 (Z × (0, ∞)). Then ρ satisfies the Fokker-Planck equation
∂ρ
= L∗ ρ for (z, t) ∈ Z × (0, ∞), (7.30a)
∂t
ρ = ρ0 for z ∈ Z × {0}. (7.30b)

110
Proof. • Let Eµ denote averaging with respect to the product measure induced by the measure µ with
density ρ0 on z(0) and the independent driving Wiener measure on the SDE itself.

• Averaging over random z(0) distributed with density ρ0 (z), we find

Z
µ
E (φ(z(t))) = v(z, t)ρ0 (z) dz
ZZ
= (eLt φ)(z)ρ0 (z) dz
Z Z
∗
= (eL t ρ0 )(z)φ(z) dz.
Z

• But since ρ(z, t) is the density of z(t) we also have

Z
µ
E (φ(z(t))) = ρ(z, t)φ(z)dz.
Z

• Equating these two expressions for the expectation at time t we obtain

Z Z
L∗ t
(e ρ0 )(z)φ(z) dz = ρ(z, t)φ(z) dz.
Z Z

• We use a density argument so that the identity can be extended to all φ ∈ L2 (Z). Hence, from the
above equation we deduce that  ∗ 
ρ(z, t) = eL t ρ0 (z).

• Differentiation of the above equation gives (7.30a).

• Setting t = 0 gives the initial condition (7.30b).

111
112
Chapter 8

The Smoluchowski and Freidlin-Wentzell

Limits

• There are very few SDEs/Fokker-Planck equations that can be solved explicitly.

• In most cases we need to study the problem under investigation either approximately or numerically.

• In this part of the course we will develop approximate methods for studying various stochastic systems
of practical interest.

• There are many problems of physical interest that can be analyzed using techniques from perturbation
theory and asymptotic analysis:

i. Small noise asymptotics at finite time intervals.

ii. Small noise asymptotics/large times (rare events): the theory of large deviations, escape from a po-
tential well, exit time problems.

iii. Small and large friction asymptotics for the Fokker-Planck equation: The Freidlin–Wentzell (under-
damped) and Smoluchowski (overdamped) limits.

iv. Large time asymptotics for the Langevin equation in a periodic potential: homogenization and aver-
aging.

v. Stochastic systems with two characteristic time scales: multiscale problems and methods.

8.1 Asymptotics for the Langevin equation

• We will study various asymptotic limits for the Langevin equation (we have set m = 1)
p
q̈ = −∇V (q) − γ q̇ + 2γβ −1 Ẇ . (8.1)

• There are two parameters in the problem, the friction coefficient γ and the inverse temperature β.

113
• 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 equation
and obtain a simpler equation for fewer variables.

• In the large temperature limit, β ≪ 1, the dynamics of (9.7) is dominated by diffusion: the Langevin
equation (9.7) can be approximated by free Brownian motion:
p
q̇ = 2γβ −1 Ẇ .

• The small temperature asymptotics, β ≫ 1 is much more interesting and more subtle. It leads to
exponential, Arrhenius type asymptotics for the reaction rate (in the case of a particle escaping from
a potential well due to thermal noise) or the diffusion coefficient (in the case of a particle moving in a
periodic potential in the presence of thermal noise)

κ = ν exp (−βEb ) , (8.2)

• where κ 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 γ. 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: γ ≫ 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: γ ≪ 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.

• We rescale the solution to (9.7):

q γ (t) = λγ (t/µγ ).

• This rescaled process satisfies the equation

q
λγ γ γ
q̈ = − 2 ∂q V (q γ /λγ ) −
γ
q̇ + 2γλ2γ µ−3 −1
γ β Ẇ , (8.3)
µγ µγ

• Different choices for these two parameters lead to the overdamped and underdamped limits:

• λγ = 1, µγ = γ −1 , γ ≫ 1.

114
 • In this case equation (8.3) becomes
p
γ −2 q̇ γ = −∂q V (q γ ) − q̇ γ + 2β −1 Ẇ . (8.4)

• Under this scaling, the interesting limit is the overdamped limit, γ ≫ 1.

• We will see later that in the limit as γ → +∞ the solution to (8.4) can be approximated by the solution
to p
q̇ = −∂q V + 2β −1 Ẇ .

• λγ = 1, µγ = γ, γ ≪ 1:
p
q̈ γ = −γ −2 ∇V (q γ ) − q̇ γ + 2γ −2 β −1 Ẇ . (8.5)

• Under this scaling the interesting limit is the underdamped limit, γ ≪ 1.

• We will see later that in the limit as γ → 0 the energy of the solution to (8.5) converges to a stochastic
process on a graph.

8.2 The Kramers to Smoluchowski Limit

• We consider the rescaled Langevin equation (8.4):
p
ε2 q̈ γ (t) = −∇V (q γ (t)) − q̇ γ (t) + 2β −1 Ẇ (t), (8.6)

• where we have set ε−1 = γ, since we are interested in the limit γ → ∞, i.e. ε → 0.

• We will show that, in the limit as ε → 0, q γ (t), the solution of the Langevin equation (8.6), converges
to q(t), the solution of the Smoluchowski equation
p
q̇ = −∇V + 2β −1 Ẇ . (8.7)

• We write (8.6) as a system of SDEs:

1
q̇ = p, (8.8)
ε r
1 1 2
ṗ = − ∇V (q) − 2 p + Ẇ . (8.9)
ε ε βε2

• This systems of SDEs defined a Markov process in phase space. Its generator is
1
1

Lε = 2
− p · ∇p + β −1 ∆ + p · ∇q − ∇q V · ∇p
ε ε
1 1
=: L0 + L1 .
ε2 ε

• This is a singularly perturbed differential operator.

115
• We will derive the Smoluchowski equation (8.7) using a pathwise technique, as well as by analyzing
the corresponding Kolmogorov equations.

• We apply Itô’s formula to p:

1 p −1
dp(t) = Lε p(t) dt + 2β ∂p p(t) dW
ε
1 1 1 p −1
= − 2 p(t) dt − ∇q V (q(t)) dt + 2β dW.
ε ε ε

• Consequently:
Z t Z t p
1
p(s) ds = − ∇q V (q(s)) ds + 2β −1 W (t) + O(ε).
ε 0 0

• From equation (8.8) we have that

Z t
1
q(t) = q(0) + p(s) ds.
ε 0

• Combining the above two equations we deduce

Z t p
q(t) = q(0) − ∇q V (q(s)) ds + 2β −1 W (t) + O(ε)
0

• from which (8.7) follows.

• Notice that in this derivation we assumed that

E|p(t)|2 6 C.

• This estimate is true, under appropriate assumptions on the potential V (q) and on the initial condi-
tions.

• In fact, we can prove a pathwise approximation result:

!1/p
E sup |q γ (t) − q(t)|p 6 Cε2−κ ,
t∈[0,T ]

• where κ > 0, arbitrary small (it accounts for logarithmic corrections).

• For the rigorous proof see

– E. Nelson, Dynamical Theories of Brownian Motion, Princeton University press 1967.

• A similar approximation theorem is also valid in infinite dimensions (i.e. for SPDEs):

116
 – S. Cerrai and M. Freidlin, On the Smoluchowski-Kramers approximation for a system with an
infinite number of degrees of freedom, Probab. Theory Related Fields, 135 (3), 2006, pp. 363–
394.

• The pathwise derivation of the Smoluchowski equation implies that the solution of the Fokker-Planck
equation corresponding to the Langevin equation (8.6) converges (in some appropriate sense to be
explained below) to the solution of the Fokker-Planck equation corresponding to the Smoluchowski
equation (8.7).

• 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 (8.6) using singular perturbation
theory.

• We will consider the problem in one dimension. This mainly to simplify the notation. The multi–
dimensional problem can be treated in a very similar way.

• The Fokker–Planck equation associated to equations (8.8) and (8.9) is

∂ρ
= L∗ ρ
∂t
1 1

= (−p∂q ρ + ∂q V (q)∂p ρ) + 2 ∂p (pρ) + β −1 ∂p2 ρ
ε
  ε
1 ∗ 1 ∗
=: L + L ρ. (8.10)
ε2 0 ε 1

• The invariant distribution of the Markov process {q, p}, if it exists, is

Z
1
ρ0 (p, q) = e−βH(p,q) , Z = e−βH(p,q) dpdq,
Z R 2

• where H(p, q) = 12 p2 + V (q). We define the function f(p,q,t) through

ρ(p, q, t) = f (p, q, t)ρ0 (p, q). (8.11)

Theorem 8.2.1. The function f (p, q, t) defined in (8.11) satisfies the equation
 
∂f 1 −1 2

1
= −p∂ q + β ∂p − (p∂ q − ∂q V (q)∂p ) f
∂t ε2 ε
 
1 1
=: L0 − L1 f. (8.12)
ε2 ε
remark

• This is ”almost” the backward Kolmogorov equation with the difference that we have −L1 instead of
L1 . This is related to the fact that L0 is a symmetric operator in L2 (R2 ; Z −1 e−βH(p,q) ), whereas L1
is antisymmetric.

117
Proof. • We note that L∗0 ρ0 = 0 and L∗1 ρ0 = 0. We use this to calculate:

L∗0 ρ = L0 (f ρ0 ) = ∂p (f ρ0 ) + β −1 ∂p2 (f ρ0 )
= ρ0 p∂p f + ρ0 β −1 ∂p2 f + f L∗0 ρ0 + 2β −1 ∂p f ∂p ρ0


= −p∂p f + β −1 ∂p2 f ρ0 = ρ0 L0 f.

• Similarly,

L∗1 ρ = L∗1 (f ρ0 ) = (−p∂q + ∂q V ∂p ) (f ρ0 )

= ρ0 (−p∂q f + ∂q V ∂p f ) = −ρ0 L1 f.

• Consequently, the Fokker–Planck equation (8.10) becomes

 
∂f 1 1
ρ0 = ρ0 L 0 f − L 1 f ,
∂t ε2 ε

• from which the claim follows.

• We look for a solution to (8.12) in the form of a power series in ε:

∞
X
f (p, q, t) = εn fn (p, q, t). (8.13)
n=0

• We substitute this expansion into eqn. (8.12) to obtain the following system of equations.

L0 f0 = 0, (8.14)
−L0 f1 = −L1 f0 , (8.15)
∂f0
−L0 f2 = −L1 f1 − (8.16)
∂t
−L0 fn+1 = −L1 fn , n = 2, 3 . . . (8.17)

• The null space of L0 consists of constants in p. Consequently, from equation (8.14) we conclude that

f0 = f (q, t).

• Now we can calculate the right hand side of equation (8.15):

L1 f0 = p∂q f.

• Equation (8.15) becomes:

L0 f1 = p∂q f.

118
• The right hand side of this equation is orthogonal to N (L∗0 ) and consequently there exists a unique
solution. We obtain this solution using separation of variables:

f1 = −p∂q f + ψ1 (q, t).

• Now we can calculate the RHS of equation (8.16). We need to calculate L1 f1 :

  
−L1 f1 = p∂q − ∂q V ∂p p∂q f − ψ1 (q, t)
= p2 ∂q2 f − p∂q ψ1 − ∂q V ∂q f.

• The solvability condition for (8.16) is

Z 
∂f0 
− L1 f 1 − ρOU (p) dp = 0,
R ∂t

• from which we obtain the backward Kolmogorov equation corresponding to the Smoluchowski SDE:
∂f
= −∂q V ∂q f + β −1 ∂p2 f. (8.18)
∂t

• Now we solve the equation for f2 . We use (8.18) to write (8.16) in the form
 
L0 f2 = β −1 − p2 ∂q2 f + p∂q ψ1 .

• The solution of this equation is

1 2
f2 (p, q, t) = ∂ f (p, q, t)p2 − ∂q ψ1 (q, t)p + ψ2 (q, t).
2 q

• The right hand side of the equation for f3 is

1 3 3
L1 f 2 = p ∂q f − p2 ∂q2 ψ1 + p∂q ψ2 − ∂q V ∂q2 f p − partialq V ∂q ψ1 .
2
The solvability condition Z
L1 f2 ρOU (p) dp = 0.
R

• This leads to the equation

−∂q V ∂q ψ1 − β −1 ∂q2 ψ1 = 0

• The only solution to this equation which is an element of L2 (e−βV (q) ) is

ψ1 ≡ 0.

• Putting everything together we obtain the first two terms in the ε-expansion of the Fokker–Planck
equation (8.12):  
ρ(p, q, t) = Z −1 e−βH(p,q) f + ε(−p∂q f ) + O(ε2 ) ,

119
 • where f is the solution of (8.18).

• Notice that we can rewrite the leading order term to the expansion in the form
1 2 /2
ρ(p, q, t) = (2πβ −1 )− 2 e−βp ρV (q, t) + O(ε),

• where ρV = Z −1 e−βV (q) f is the solution of the Smoluchowski Fokker-Planck equation

∂ρV
= ∂q (∂q V ρV ) + β −1 ∂q2 ρV .
∂t
• It is possible to expand the n-th term in the expansion (8.13) in terms of Hermite functions (the
eigenfunctions of the generator of the OU process)
n
X
fn (p, q, t) = fnk (q, t)φk (p), (8.19)
k=0

where φk (p) is the k–th eigenfunction of L0 :

−L0 φk = λk φk .

• We can obtain the following system of equations (Lb = β −1 ∂q − ∂q V ):

b n1 = 0,
Lf
s
k+1 b p
Lf n,k+1 + kβ −1 ∂q fn,k−1 = −kfn+1,k , k = 1, 2 . . . , n − 1,
β −1
p
nβ −1 ∂q fn,n−1 = −nfn+1,n,
p
(n + 1)β −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 
−1 2 β 2
ρ(x, y, t) = ρ0 (p, q) f + ε(− β ∂q f φ1 ) + ε √ ∂q f φ2 + f20
2
r !!
β −3  p p 
+ε3 − ∂ 3 f φ3 + − β −1 L∂ b q2 f − β −1 ∂q f20 φ1
3! q
+O(ε4 ),

8.3 The Freidlin–Wentzell Limit

Consider now the rescaling λγ,ε = 1, µγ,ε = γ. The Langevin equation becomes
p
q̈ γ = −γ −2 ∇V (q γ ) − q̇ γ + 2γ −2 β −1 Ẇ . (8.20)

We write equation (8.20) as system of two equations

p
q̇ γ = γ −1 pγ , ṗγ = −γ −1 V ′ (q γ ) − pγ + 2β −1 Ẇ .

120
This is the equation for an O(1/γ) 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 Itô’s formula to
the Hamiltonian of the system to obtain

p
Ḣ = β −1 − p2 + 2β −1 p2 Ẇ

with p2 = p2 (H, q) = 2(H − V (q)).

Thus, in order to study the γ → 0 limit we need to analyze the following fast/slow system of SDEs

p
Ḣ = β −1 − p2 + 2β −1 p2 Ẇ (8.21a)
p
ṗγ = −γ −1 V ′ (q γ ) − pγ + 2β −1 Ẇ . (8.21b)

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
Ḣ = β −1 − hp2 i + 2β −1 hp2 iẆ . (8.22)

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.

• For more information see

– Freidlin and Wentzell, Random Perturbations of Dynamical Systems, Springer 1998.

– Freidlin and Wentzell, Random Perturbations of Hamiltonian Systems, AMS 1994.
– Sowers, A Boundary layer theory for diffusively perturbed transport around a heteroclinic cycle,
CPAM 58 (2005), no. 1, 30–84.

We will study the small γ asymptotics by analyzing the corresponding backward Kolmogorov equation
using singular perturbation theory. The generator of the process {q γ , pγ } is

Lγ = γ −1 (p∂q − ∂q V ∂p ) − p∂p + β −1 ∂p2

= γ −1 L0 + L1 .

121
Let uγ = E(f (pγ (p, q; t), q γ (p, q; t))). It satisfies the backward Kolmogorov equation associated to the
process {q γ , pγ }:  
∂uγ 1
= L0 + L1 uγ . (8.23)
∂t γ
We look for a solution in the form of a power series expansion in ε:

uγ = u0 + γu1 + γ 2 u2 + . . .

We substitute this ansatz into (8.23) and equate equal powers in ε to obtain the following sequence of
equations:
L0 u0 = 0, (8.24a)
∂u0
L0 u1 = −L1 u1 + , (8.24b)
∂t
∂u1
L0 u2 = −L1 u1 + . (8.24c)
∂t
.........

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 . (8.25)

Let us now analyze equations (8.24). We start with (8.24a); eqn. (8.25) implies that u0 depends on q, p
through the Hamiltonian function H:
u0 = u(H(p, q), t) (8.26)

Now we proceed with (8.24b). For this we need to find the solvability condition for equations of the form

L0 u = f (8.27)

My multiply it by an arbitrary smooth function of H(p, q), integrate over R2 and use the skew-symmetry of
the Liouville operator L0 to deduce:1

Z Z
L0 uF (H(p, q)) dpdq = uL∗0 F (H(p, q)) dpdq
R2 Z R2

= u(−L0 F (H(p, q))) dpdq

R2
= 0, ∀F ∈ Cb∞ (R).
1
We assume that both u1 and F decay to 0 as |p| → ∞ to justify the integration by parts that follows.

122
This implies that the solvability condition for equation (8.27) is that
Z
f (p, q)F (H(p, q)) dpdq = 0, ∀F ∈ Cb∞ (R). (8.28)
R2

We use the solvability condition in (8.24b) to obtain that

Z  
∂u0
L1 u1 − F (H(p, q)) dpdq = 0, (8.29)
R2 ∂t

To proceed, we need to understand how L1 acts to functions of H(p, q). Let φ = φ(H(p, q)). We have that

∂φ ∂H ∂φ ∂φ
= =p
∂p ∂p ∂H ∂H

and

 
∂2φ ∂ ∂φ ∂φ ∂2φ
2
= = + p2 .
∂p ∂p ∂H ∂H ∂H 2

The above calculations imply that, when L1 acts on functions φ = φ(H(p, q)), it becomes
h i
L1 = (β −1 − p2 )∂H + β −1 p2 ∂H
2
, (8.30)

where
p2 = p2 (H, q) = 2(H − V (q)).

We want to change variables in the integral (8.29) and go from (p, q) to p, H. The Jacobian of the
transformation is:
∂p ∂p
∂(p, q) ∂p 1
= ∂H∂q
∂q
∂q = = .
∂(H, q) ∂H ∂q
∂H p(H, q)
We use this, together with (8.30), to rewrite eqn. (8.29) as
Z Z 
∂u h −1 i 
+ (β − p2 )∂H + β −1 p2 ∂H 2
u F (H)p−1 (H, q) dHdq = 0.
∂t

We introduce the notation Z

h·i := · dq.

The integration over q can be performed ”explicitly”:

Z h
∂u −1
i
hp i + (β −1 hp−1 i − hpi)∂H + β −1 hpi∂H
2
u F (H) dH = 0.
∂t

This equation should be valid for every smooth function F (H), and this requirement leads to the differ-
ential equation
∂u

hp−1 i = β −1 hp−1 i − hpi ∂H u + hpiβ −1 ∂H2
u,
∂t

123
or,
∂u

= β −1 − hp−1 i−1 hpi ∂H u + γhp−1 i−1 hpiβ −1 ∂H2
u.
∂t
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:

Ḣ = b(H) + σ(H)Ẇ (8.31)

where
b(H) = β −1 − hp−1 i−1 hpi, σ(H) = β −1 hp−1 i−1 hpi.
Notice that the noise that appears in the limiting equation (8.31) is multiplicative, contrary to the additive
noise in the Langevin equation.
As it well known from classical mechanics, the action and frequency are defined as
Z
I(E) = p(q, E) dq

and  −1
dI
ω(E) = 2π ,
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 8.3.1. The limiting Fokker–Planck equation for the energy distribution function ρ(E, t) is
  
∂ρ ∂ −1 ∂ ω(E)ρ
= I(E) + β . (8.32)
∂t ∂E ∂E 2π

Proof. We notice that Z Z

dI ∂p
= dq = p−1 dq
dE ∂E
and consequently
ω(E)
hp−1 i−1 = .
2π
Hence, the limiting Fokker–Planck equation can be written as
    
∂ρ ∂ I(E)ω(E) ∂2 Iω
= − β −1 ρ + β −1
∂t ∂E 2π ∂E 2 2π
     
−1 ∂ρ ∂ Iω −1 ∂ dI ωρ −1 ∂ ∂  ωρ 
= −β + ρ +β +β I
∂E ∂E 2π ∂E dE 2π ∂E ∂E 2π
     
∂ Iω ∂ ∂ ωρ
= ρ + β −1 I
∂E 2π ∂E ∂E 2π
  
∂ −1 ∂ ω(E)ρ
= I(E) + β ,
∂E ∂E 2π

which is precisely equation (8.32).

124
Remarks 8.3.2. i. We emphasize that the above formal procedure does not provide us with the boundary
conditions for the limiting Fokker–Planck equation. We will discuss about this issue in the next section.

ii. If we rescale back to the original time-scale we obtain the equation

  
∂ρ ∂ −1 ∂ ω(E)ρ
=γ I(E) + β . (8.33)
∂t ∂E ∂E 2π

We will use this equation later on to calculate the rate of escape from a potential barrier in the
energy-diffusion-limited regime.

125
126
Chapter 9

Exit Time Problems and Reaction Rate

Theory

9.1 Introduction
There are many systems in physics, chemistry and biology that exist in at least two stable states. Among
the many applications we mention the switching and storage devices in computers. Another example is
biological macromolecules that can exist in many different states. The problems that we would like to solve
are:
• How stable are the various states relative to each other.

• 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 Eijnden etc.

• How does the system relax to an unstable state?

We can separate between the 1d problem, the finite dimensional problem and the infinite dimensional prob-
lem (SPDEs). We we will solve completely the one dimensional problem and discuss in some detail about
the finite dimensional problem. The infinite dimensional situation is an extremely hard problem and we
will 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 calculation of various quantities such as reaction
rates, transition pathways etc.
We will mostly consider the dynamics of a particle moving in a bistable potential, under the influence of
thermal noise in one dimension: p
ẋ = −V ′ (x) + 2kB T β̇. (9.1)
An example of the class of potentials that we will consider is shown in Figure. It has to local minima,
one local maximum and it increases at least quadratically at infinity. This ensures that the state space
is ”compact”, i.e. that the particle cannot escape at infinity. The standard potential that satisfies these
assumptions is
1 1 1
V (x) = x4 − x2 + . (9.2)
4 2 4

127
 3

−1

−2

−3
0 50 100 150 200 250 300 350 400 450 500

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 mathematically!) interesting
case is when the thermal fluctuations are weak when compared to the potential barrier that the particle has
to climb over.
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 Ornstein–Uhlenbeck 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 β := (kB T )−1 :

τ = ν −1 exp(β∆E).

It is more customary to calculate the reaction rate κ := τ −1 which gives the rate with which particles
escape from a local minimum of the potential:

κ = ν exp(−β∆E). (9.3)

128
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
ẋ = −V ′ (x), x(0) = x0 .

In this case the potential becomes a Lyapunov function:

dx dx
= V ′ (x) = −(V ′ (x))2 < 0.
dt dt
Hence, depending on the initial condition the particle will converge either to a or c. The particle cannot
escape from either state of local stability.
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.

9.2 Kramers’ Theory

9.3 The Mean First Passage Time Approach

The Arrhenius-type factor in the formula for the reaction rate, eqn. (9.3) is intuitively and it has been ob-
served experimentally in the late nineteenth century by Arrhenius and others. What is extremely important
both from a theoretical and an applied point of view is the calculation of the prefactor ν, the rate coefficient.
A systematic approach for the calculation of the rate coefficient, as well as the justification of the Arrhenius
kinetics, is that of the mean first passage time method (MFPT). Since this method is of independent interest
and is useful in various other contexts, we will present it in a quite general setting and apply it to the problem
of the escape from a potential barrier in later sections. We will first treat the one dimensional problem and
then extend the theory to arbitrary finite dimensions.
We will restrict ourselves to the case of homogeneous Markov processes. It is not very easy to extend
the method to non-Markovian processes.

9.3.1 The Boundary Value Problem for the MFPT

• Let Xt be a continuous time diffusion process on Rd whose evolution is governed by the SDE

dXtx = b(Xtx ) dt + σ(Xtx ) dWt , X0x = x. (9.4)

• Let D be a bounded subset of Rd with smooth boundary. Given x ∈ D, we want to know how long it
takes for the process Xt to leave the domain D for the first time

x
τD = inf {t > 0 : Xtx ∈
/ D} .

129
 • Clearly, this is a random variable. The average of this random variable is called the mean first passage
time MFPT or the first exit time:
x
τ (x) := EτD .

• We can calculate the MFPT by solving an appropriate boundary value problem.

Theorem 9.3.1. The MFPT is the solution of the boundary value problem

−Lτ = 1, x ∈ D, (9.5a)

τ = 0, x ∈ ∂D, (9.5b)
where L is the generator of the SDE 9.5.

• 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 Itô’s formula.

Derivation

• Let ρ(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.

∂ρ
= L∗ ρ, ρ(X, x, 0) = δ(X − x), ρ|∂D = 0. (9.6)
∂t

• We can write the solution to this equation in the form

∗
ρ(X, x, t) = eL t δ(X − x),

∗
• where the absorbing boundary conditions are included in the definition of the semigroup eL t .

• The homogeneous Dirichlet (absorbing) boundary conditions imply that

lim ρ(X, x, t) = 0.
t→+∞

• 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) = ρ(X, x, t) dx.
D

• Notice that this is a decreasing function of time. We can write

∂S
= −f (x, t),
∂t

130
 • where f (x, t) is the first passage times distribution.

• The MFPT is the first moment of the distribution f (x, t):

Z +∞ Z +∞
dS
τ (x) = f (s, x)s ds = − s ds
0 0 ds
Z +∞ Z +∞ Z
= S(s, x) ds = ρ(X, x, s) dXds
0 0 D
Z +∞ Z
∗
= eL s δ(X − x) dXds
0 D
Z +∞ Z   Z +∞  
Ls
= δ(X − x) e 1 dXds = eLs 1 ds.
0 D 0

• We apply L to the above equation to deduce:

Z +∞   Z t
d  Lt 
Lτ = LeLt 1 dt = Le 1 dt
0 0 dt
= −1.

9.4 Escape from a potential barrier

In this section we use the theory developed in the previous section to study the long time/small temperature
asymptotics of solutions to the Langevin equation for a particle moving in a one–dimensional potential of
the form (9.2): p
ẍ = −V ′ (x) − γ ẋ + 2γkB T Ẇ . (9.7)
In particular, we justify the Arrhenius formula for the reaction rate

κ = ν(γ) exp(−β∆E)

and we calculate the escape rate ν = ν(γ). In particular, we analyze the dependence 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.

9.4.1 Calculation of the Reaction Rate in the Smoluchowski regime

We consider the Langevin equation (9.7) in the limit of large friction. As we saw in Chapter ??, in the
overdamped limit γ ≫ 1, the solution to (9.7) can be approximated by the solution to the Smoluchowski
equation (9.1) p
γ ẋ = −V ′ (x) + 2γkB T Ẇ .
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 minimum.
• Consider the boundary value problem for the MFPT of the one dimensional diffusion process (9.1)
from the interval (a, b):  
−β −1 eβV ∂x e−βV τ = 1 (9.8)

131
• We choose reflecting BC at x = a and absorbing B.C. at x = b. We can solve (9.8) with these
boundary conditions by quadratures:
Z b Z y
−1 βV (y)
τ (x) = β dye dze−βV (z) . (9.9)
x 0

• 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 βV (y)
τ (x) = β dye dze−βV (z) .
x −∞

• When Eb β ≫ 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 +∞  
βω02 2
exp(−βV (z)) dz ≈ exp(−βV (a)) exp − (z − a) dz
−∞ −∞ 2
s
2π
= exp (−βV (a)) ,
βω02

• where we have used the Taylor series expansion around the minimum:
1
V (z) = V (a) + ω02 (z − a)2 + . . .
2
• 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) − ωb2 (y − b)2 + . . .
2
• Assuming that x is close to a, the minimum of the potential, we can replace the lower limit of inte-
gration by −∞. We finally obtain
Z b Z b  
βωb2 2
exp(βV (y)) dy ≈ exp(βV (b)) exp − (y − b) dy
x −∞ 2
s
1 2π
= exp (βV (b)) .
2 βωb2

• Putting everything together we obtain a formula for the MFPT:

π
τ (x) = exp (βEb ) .
ω0 ωb

• The rate of arrival at b is 1/τ . Only have of the particles escape. Consequently, the escape rate (or
1
reaction rate), is given by 2τ :
ω0 ωb
κ= exp (−βEb ) .
2π

132
9.4.2 The Intermediate Regime: γ = O(1)
• Consider now the problem of escape from a potential well for the Langevin equation
p
q̈ = −∂q V (q) − γ q̇ + 2γβ −1 Ẇ . (9.10)

• The reaction rate depends on the fiction coefficient and the temperature. In the overdamped limit
(γ ≫ 1) we retrieve (??), appropriately rescaled with γ:
ω0 ωb
κ= exp (−βEb ) . (9.11)
2πγ

• We can also obtain a formula for the reaction rate for γ = O(1):
q
γ2 γ
4 − ωb2 − 2 ω0
κ= exp (−βEb ) . (9.12)
ωb 2π

• Naturally, in the limit as γ → +∞ (9.12) reduces to (9.11)

9.4.3 Calculation of the Reaction Rate in the energy-diffusion-limited regime

• In order to calculate the reaction rate in the underdamped or energy-diffusion-limited regime γ ≪ 1
we need to study the diffusion process for the energy, (8.31) or (8.32). The result is
ω0 −βEb
κ = γβI(Eb ) e , (9.13)
2π

• where I(Eb ) denotes the action evaluated at b.

• A formula for the escape rate which is valid for all values of friction coefficient was obtained by Mel-
nikov and Meshkov in 1986, J. Chem. Phys 85(2) 1018-1027. This formula requires the calculation of
integrals and it reduced to (9.11) and (9.13) in the overdamped and underdamped limits, respectively.

9.5 Extension to Higher Dimensions

9.6 Discussion and Bibliography

The calculation of reaction rates and the stochastic modeling of chemical reactions has been a very active
area of research since the 30’s. One of the first methods that were developed was that of transition state
theory cite WIGNER. Kramers developed his theory in his celebrated paper CITE KRAMERS. In this
chapter we have based our approach to the calculation of the mean first passage time. Our analysis is based
mostly on [10, Ch. 5, Ch. 9], [26, Ch. 4] and the excellent review article [13]. We highly recommend
this review article for further information on reaction rate theory. See also [12] and the review article of
Melnikov (1991).

133
 There are many applications of interest where it is important to calculate reaction rates for non-Markovian
Langevin equations of the form
Z t
′
ẍ = −V (x) − bγ(t − s)ẋ(s) ds + ξ(t) (9.14a)
0

hξ(t)ξ(0)i = kB T M −1 γ(t) (9.14b)

We will derive generalized non–Markovian equations of the form (9.14a), together with the fluctuation–
dissipation theorem (9.14b), in Chapter ??. The calculation of reaction rates for the generalized Langevin
equation is presented in [12].
The long time/small temperature asymptotics can be studied rigorously by means of the theory of
Freidlin-Wentzell [6]. See also [2]. A related issue is that of the small temperature asymptotics for the
eigenvalues (in particular, the first eigenvalue) of the generator of the Markov process x(t) which is the
solution of p
γ ẋ = −∇V (x) + 2γkB T Ẇ .
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 micromagnetics...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 [21, 17, 18]. 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 [18]. This formula is obtained through a careful analysis of the PDE

p∂q τ − ∂q V ∂p τ + γ(−p∂p + kB T ∂p2 )τ = −1,

for the mean first passage time τ . The PDE is equipped, of course, with the appropriate 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 [22].
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.

134
Bibliography

[1] L. Arnold. Stochastic differential equations: theory and applications. Wiley-Interscience [John Wiley
& Sons], New York, 1974. Translated from the German.

[2] N. Berglund and B. Gentz. Noise-induced phenomena in slow-fast dynamical systems. Probability and
its Applications (New York). Springer-Verlag London Ltd., London, 2006. A sample-paths approach.

[3] 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.

[4] L.C. Evans. Partial Differential Equations. AMS, Providence, Rhode Island, 1998.

[5] W. Feller. An introduction to probability theory and its applications. Vol. II. Second edition. John
Wiley & Sons Inc., New York, 1971.

[6] M.I. Freidlin and A.D. Wentzell. Random Perturbations of dunamical systems. Springer-Verlag, New
York, 1984.

[7] A. Friedman. Partial differential equations of parabolic type. Prentice-Hall Inc., Englewood Cliffs,
N.J., 1964.

[8] A. Friedman. Stochastic differential equations and applications. Vol. 1. Academic Press [Harcourt
Brace Jovanovich Publishers], New York, 1975. Probability and Mathematical Statistics, Vol. 28.

[9] A. Friedman. Stochastic differential equations and applications. Vol. 2. Academic Press [Harcourt
Brace Jovanovich Publishers], New York, 1976. Probability and Mathematical Statistics, Vol. 28.

[10] C. W. Gardiner. Handbook of stochastic methods. Springer-Verlag, Berlin, second edition, 1985. For
physics, chemistry and the natural sciences.

[11] I. I. Gikhman and A. V. Skorokhod. Introduction to the theory of random processes. Dover Publications
Inc., Mineola, NY, 1996.

[12] P. Hanggi. Escape from a metastable state. J. Stat. Phys., 42(1/2):105–140, 1986.

[13] P. Hanggi, P. Talkner, and M. Borkovec. Reaction-rate theory: fifty years after Kramers. Rev. Modern
Phys., 62(2):251–341, 1990.

135
[14] W. Horsthemke and R. Lefever. Noise-induced transitions, volume 15 of Springer Series in Synerget-
ics. Springer-Verlag, Berlin, 1984. Theory and applications in physics, chemistry, and biology.

[15] S. Karlin and H. M. Taylor. A second course in stochastic processes. Academic Press Inc. [Harcourt
Brace Jovanovich Publishers], New York, 1981.

[16] P. Mandl. Analytical treatment of one-dimensional Markov processes. Die Grundlehren der mathe-
matischen Wissenschaften, Band 151. Academia Publishing House of the Czechoslovak Academy of
Sciences, Prague, 1968.

[17] B. J. Matkowsky, Z. Schuss, and E. Ben-Jacob. A singular perturbation approach to Kramers’ diffusion
problem. SIAM J. Appl. Math., 42(4):835–849, 1982.

[18] B. J. Matkowsky, Z. Schuss, and C. Tier. Uniform expansion of the transition rate in Kramers’ problem.
J. Statist. Phys., 35(3-4):443–456, 1984.

[19] Frigyes Riesz and Béla 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.

[20] H. Risken. The Fokker-Planck equation, volume 18 of Springer Series in Synergetics. Springer-Verlag,
Berlin, 1989.

[21] Z. Schuss. Singular perturbation methods in stochastic differential equations of mathematical physics.
SIAM Review, 22(2):119–155, 1980.

[22] Ch. Schütte and W. Huisinga. Biomolecular conformations can be identified as metastable sets of
molecular dynamics. In Handbook of Numerical Analysis (Computational Chemistry), Vol X, 2003.

[23] C. Schwab and R.A. Todor. Karhunen-Loève approximation of random fields by generalized fast
multipole methods. J. Comput. Phys., 217(1):100–122, 2006.

[24] D.W. Stroock. Probability theory, an analytic view. Cambridge University Press, Cambridge, 1993.

[25] G. I. Taylor. Diffusion by continuous movements. London Math. Soc., 20:196, 1921.

[26] R. Zwanzig. Nonequilibrium statistical mechanics. Oxford University Press, New York, 2001.

136