Integral Equations

Integral equations are equations in which an unknown function appears under an integral sign. The central result is the Fredholm alternative: for a linear integral equation

ϕ - λ K ϕ = f

with a compact operator

K

, either the equation has a unique solution for every

f

, or the homogeneous equation has nontrivial solutions. They appear wherever a problem is naturally global — potential theory, scattering, inverse problems, and the reformulation of boundary-value problems. Foundational work by Fredholm (1900), Hilbert, and Schmidt (1900s) created spectral theory and, ultimately, functional analysis.

An integral equation is a functional equation of the form $g (x) = \int_{a}^{b} K (x, t) ϕ (t) d t$ (or a variant where $ϕ$ also appears outside the integral), where the kernel $K$ is given and $ϕ$ is the unknown. They are the integral counterpart of differential equations, but with a crucial structural advantage: the integral operator is typically compact and bounded, whereas differentiation is unbounded. This single fact makes the spectral theory clean and the well-posedness questions tractable, and it is precisely why one converts differential boundary-value problems into integral equations (via Green's functions). The surprising fact is that the entire edifice of Hilbert space theory — eigenfunction expansions, the spectral theorem, Banach/Hilbert spaces themselves — grew historically out of Hilbert's attempt to systematize Fredholm's 1900 solution of these equations.

The Core Insight

Reading an integral equation as an operator equation $(I - λ K) ϕ = f$ converts an analysis problem into a question about the spectrum of a single linear operator $K$ . When $K$ is compact, the solvability theory mirrors finite-dimensional linear algebra almost perfectly: the spectrum is discrete (except possibly at $0$ ), eigenvalues accumulate only at $0$ , and the Fredholm alternative — "uniqueness implies existence" — holds verbatim. The integral sign, far from being an obstacle, is what regularizes the problem.

Classification

Types of Integral Equations

Let $ϕ$ be the unknown function, $K (x, t)$ a given kernel, $f$ a given data function, and $λ$ a (possibly spectral) parameter.

Fredholm equation of the first kind:

f (x) = \int_{a}^{b} K (x, t) ϕ (t) d t

⟳the unknown appears only inside the integral, with fixed limits

a, b

; this is generically ill-posed.

Fredholm equation of the second kind:

ϕ (x) = f (x) + λ \int_{a}^{b} K (x, t) ϕ (t) d t

⟳

ϕ

appears both inside and outside; rewrite as

(I - λ K) ϕ = f

, the well-posed case.

Volterra equation (first / second kind): replace the upper limit

b

by the variable

x

ϕ (x) = f (x) + λ \int_{a}^{x} K (x, t) ϕ (t) d t

⟳variable upper limit makes

K

effectively triangular ("strictly lower-triangular"), forcing zero spectral radius.

An equation is homogeneous if $f \equiv 0$ , linear if $ϕ$ enters linearly (as above), and singular if the interval is infinite or $K$ has a non-integrable-looking singularity (e.g. Cauchy or weakly singular kernels).

The Integral Operator

Associated to a kernel $K \in L^{2} ([a, b]^{2})$ is the integral operator $K : L^{2} [a, b] \to L^{2} [a, b]$ ,

(K ϕ) (x) = \int_{a}^{b} K (x, t) ϕ (t) d t .

⟳by Cauchy–Schwarz and Fubini,

∥ K ϕ ∥_{L^{2}} \leq ∥ K ∥_{L^{2} ([a, b]^{2})} ∥ ϕ ∥_{L^{2}}

, so

K

is bounded with Hilbert–Schmidt norm bounding the operator norm.

A kernel is Hilbert–Schmidt if $\iint ∣ K (x, t) ∣^{2} d x d t < \infty$ ; such $K$ are compact. The kernel is symmetric (Hermitian) if $K (x, t) = \overline{K (t, x)}$ , making $K$ self-adjoint.

The Fredholm Theory

Fredholm Alternative

Let $K$ be a compact operator on a Banach space $X$ and $λ \neq = 0$ . Consider

ϕ - λ K ϕ = f .

⟳

I - λ K

is a compact perturbation of the identity, so it is Fredholm of index

0

; injectivity and surjectivity are equivalent.

Then exactly one of the following holds:

The homogeneous equation $ϕ - λ K ϕ = 0$ has only the trivial solution. Then $(I - λ K)^{- 1}$ exists and is bounded, so $[5]$ has a unique solution for every $f$ .
The homogeneous equation has a nontrivial solution. Then the solution space is finite-dimensional, of the same dimension as that of the adjoint equation $ψ - λ K^{*} ψ = 0$ , and $[5]$ is solvable iff $f$ is orthogonal to all solutions of the adjoint homogeneous equation.

Fredholm Determinant and Resolvent

For a continuous kernel on $[a, b]$ , Fredholm constructed the solution explicitly. Define the Fredholm determinant

D (λ) = n = 0 \sum \infty \frac{( - λ ) ^{n}}{n !} \int_{a}^{b} \dots \int_{a}^{b} det [K (s_{i}, s_{j})]_{i, j = 1}^{n} d s_{1} \dots d s_{n},

⟳discretize the integral as a Riemann sum, form the matrix determinant

det (I - λ K_{h})

, expand it, and pass to the limit using the Hadamard inequality for convergence.

an entire function of $λ$ . The values $λ$ with $D (λ) = 0$ are exactly the characteristic values (reciprocals of eigenvalues). When $D (λ) \neq = 0$ , the unique solution of $[5]$ is

ϕ (x) = f (x) + λ \int_{a}^{b} R (x, t; λ) f (t) d t, R = \frac{D ( x , t ; λ )}{D ( λ )},

⟳

R

is the resolvent kernel of

(I - λ K)^{- 1} = I + λ R

;

D (x, t; λ)

is the first Fredholm minor, defined by the analogous determinantal series.

where

R (x, t; λ)

is the resolvent kernel.

Why It Holds / Derivation

Neumann Series for Small $\lambda$

We construct $(I - λ K)^{- 1}$ directly when $∣ λ ∣$ is small.

Step 1. Bound the operator. With

M = sup_{x, t} ∣ K (x, t) ∣

and interval length

L = b - a

, iterating

K

gives

∥ K^{n} ∥ \leq (M L)^{n}

C [a, b]

(or use

∥ K ∥ < \infty

L^{2}

). Step 2. Form the geometric (Neumann) series

ϕ = n = 0 \sum \infty λ^{n} K^{n} f = f + λ K f + λ^{2} K^{2} f + \dots

⟳substitute

(I - λ K)^{- 1} = \sum λ^{n} K^{n}

and verify termwise that

(I - λ K)

applied to the sum telescopes to

f

Step 3. Convergence. The series converges absolutely in operator norm whenever

∣ λ ∣ ∥ K ∥ < 1

, i.e.

∣ λ ∣ < 1/∥ K ∥

(for Volterra, all

λ

, since

∥ K^{n} ∥ \leq (M L)^{n} / n! \to 0

). Step 4. The iterated kernels

K_{n} (x, t)

, defined by

K_{1} = K

and

K_{n + 1} (x, t) = \int_{a}^{b} K (x, s) K_{n} (s, t) d s

, give the resolvent

R (x, t; λ) = n = 0 \sum \infty λ^{n} K_{n + 1} (x, t),

⟳collect the

K^{n} f

terms of

[8]

for

n \geq 1

and identify each with

\int K_{n} (x, t) f (t) d t

Step 5. Analytic continuation. The Neumann series gives

R

only inside the disk

∣ λ ∣ < 1/∥ K ∥

, but the Fredholm series

[6]

–

[7]

continues it to a meromorphic function on all of

C

, with poles at the characteristic values.

■

Hilbert–Schmidt Expansion

Let $K$ be a symmetric, continuous, non-degenerate Hilbert–Schmidt kernel. Then $K$ has a sequence of real eigenvalues $μ_{n} \to 0$ with orthonormal eigenfunctions ${ψ_{n}}$ , and

K (x, t) = n = 1 \sum \infty μ_{n} ψ_{n} (x) \overline{ψ_{n} (t)} (in L^{2}),

⟳apply the spectral theorem for compact self-adjoint operators; the series converges in

L^{2} ([a, b]^{2})

, and in mean for any

K ϕ

and for any

g

in the range of

K

g = \sum_{n} ⟨ g, ψ_{n} ⟩ ψ_{n}

converges (the Hilbert–Schmidt theorem).

David Hilbert | 1862–1943

Hilbert, building directly on Erik Fredholm's 1900 paper, devoted a celebrated series of six communications (1904–1910) to integral equations. He recognized the kernel's symmetric case as an infinite-dimensional generalization of the principal-axis theorem for quadratic forms, and in doing so introduced the notions of complete orthonormal systems, the $ℓ^{2}$ sequence space, and "spectrum" (a word he coined for this context, before its physical relevance to atoms was known). His student Erhard Schmidt simplified and geometrized the theory using what we now call Hilbert space and the Gram–Schmidt process. The phrase "Hilbert space" was later coined by von Neumann precisely to axiomatize the structure Hilbert had found while studying these equations.

Building Intuition

Think of the kernel $K (x, t)$ as an infinite matrix indexed by continuous labels $x, t$ instead of integers $i, j$ . Then $\int K (x, t) ϕ (t) d t$ is "matrix times vector," and a Fredholm equation of the second kind is literally $(I - λ A) ϕ = f$ . Everything you know from linear algebra transfers: eigenvalues, determinants ( $D (λ)$ is " $det (I - λ K)$ "), the alternative "either invertible or has a kernel." The one thing that breaks is that an infinite matrix need not be invertible just because it has no kernel — unless it is compact, meaning it "almost" lives in finite dimensions (its eigenvalues march to zero). Compactness is the price of admission that makes the finite-dimensional intuition exact. Volterra equations are the "strictly triangular" case: a triangular matrix with zero diagonal is nilpotent, so $I - λ K$ is always invertible — which is why Volterra equations never have a homogeneous solution.

Spectrum of a compact integral operator: eigenvalues accumulate only at the origin, leaving the resolvent set open everywhere else.

Worked Examples

A Degenerate (Separable) Kernel

Given. Solve

ϕ (x) = x + λ \int_{0}^{1} x t ϕ (t) d t

[0, 1]

. Step 1. The kernel

K (x, t) = x t

is separable: it factors as

g (x) h (t)

with

g (x) = x, h (t) = t

. So

\int_{0}^{1} x t ϕ (t) d t = x c

where

c = \int_{0}^{1} t ϕ (t) d t

is an unknown constant. Step 2. Then

ϕ (x) = x + λ c x = (1 + λ c) x

. Step 3. Substitute back to determine

c

c = \int_{0}^{1} t (1 + λ c) t d t = (1 + λ c) \int_{0}^{1} t^{2} d t = (1 + λ c) \cdot \frac{1}{3} .

Step 4. Solve:

c = \frac{1}{3} + \frac{λ c}{3} \Rightarrow c (1 - \frac{λ}{3}) = \frac{1}{3} \Rightarrow c = \frac{1}{3 - λ}

. Answer.

ϕ (x) = \frac{3}{3 - λ} x, λ \neq = 3.

The single characteristic value is

λ = 3

(where

D (λ) = 1 - λ /3

vanishes); at

λ = 3

the data

f (x) = x

is not orthogonal to the eigenfunction

x

, so by the Fredholm alternative no solution exists.

A Volterra Equation by Neumann Series

Given. Solve

ϕ (x) = 1 + \int_{0}^{x} ϕ (t) d t

[0, \infty)

(

λ = 1

K \equiv 1

). Step 1. Iterated kernels:

K_{1} (x, t) = 1

for

t \leq x

; then

K_{2} (x, t) = \int_{t}^{x} 1 d s = x - t, K_{3} = \int_{t}^{x} (s - t) d s = \frac{( x - t ) ^{2}}{2 !}, \dots, K_{n + 1} = \frac{( x - t ) ^{n}}{n !} .

Step 2. Resolvent kernel (eqn

[9]

with

λ = 1

R (x, t) = n = 0 \sum \infty \frac{( x - t ) ^{n}}{n !} = e^{x - t} .

Step 3. Solution by

[7]

with

f \equiv 1

ϕ (x) = 1 + \int_{0}^{x} e^{x - t} \cdot 1 d t = 1 + e^{x} [- e^{- t}]_{0}^{x} = 1 + e^{x} (1 - e^{- x}) = e^{x} .

Check.

1 + \int_{0}^{x} e^{t} d t = 1 + (e^{x} - 1) = e^{x} .

✓ Answer.

ϕ (x) = e^{x} .

Note: the Neumann series converges for all

λ

here, confirming Volterra equations have no characteristic values.

Common Misconceptions

First-kind equations are not just "easier second-kind equations." The Fredholm equation of the first kind

[1]

is generically ill-posed: because

K

is compact, its inverse is unbounded, so tiny perturbations in the data

f

produce arbitrarily large changes in

ϕ

. Solving

[1]

numerically requires regularization (Tikhonov, truncated SVD). By contrast the second-kind equation

[2]

involves

I - λ K

, which is boundedly invertible away from characteristic values — it is well-posed. Never apply Fredholm-alternative reasoning blindly to a first-kind problem. Compactness is essential — drop it and the Fredholm alternative fails. Consider the shift operator

S

ℓ^{2}

S (x_{1}, x_{2}, \dots) = (0, x_{1}, x_{2}, \dots)

. It is bounded and injective (no nontrivial kernel), yet it is not surjective (nothing maps to

(1, 0, 0, \dots)

). So "injective

\Rightarrow

surjective" — the heart of the alternative — is false for general bounded operators. The corresponding integral operator with a non-Hilbert–Schmidt kernel (e.g.

K (x, t) = e^{i (x - t)}

R

, which is essentially multiplication after Fourier transform) likewise has continuous spectrum and no discrete eigenvalue structure. Compactness — equivalently

K_{n} \to 0

for the singular values — is exactly what rules this out.

Connections

Integral equations sit at a crossroads. Via the Green's Function, any linear boundary-value problem for an ODE or PDE becomes a Fredholm equation of the second kind, trading a singular differential operator for a compact integral one — this is the standard route to existence proofs in Potential Theory (single- and double-layer potentials yield boundary integral equations). The eigenfunction expansion $[10]$ is the Spectral Theorem for compact self-adjoint operators made explicit. The whole theory is the birthplace of Functional Analysis: Hilbert space, compact operators, and the Fredholm Operator index theory all originate here. Volterra equations connect to Semigroup Theory and the variation-of-parameters formula, and singular integral equations (Cauchy kernels) link to Riemann–Hilbert Problems and the theory of Singular Integral Operators.

Real-World Applications

Boundary element method (BEM) — engineering electromagnetics, acoustics, and elasticity reformulate Laplace/Helmholtz problems as boundary integral equations, reducing a 3D PDE to a 2D surface equation (e.g. antenna design, ship hydrodynamics).
Computed tomography / medical imaging — the Radon transform is a first-kind integral equation; inverting it (filtered back-projection) reconstructs CT and PET images.
Scattering theory & quantum mechanics — the Lippmann–Schwinger equation $ψ = ψ_{0} + G_{0} V ψ$ is a Fredholm equation of the second kind for scattering wavefunctions.
Radiative transfer & astrophysics — Chandrasekhar's transfer equation for stellar atmospheres is a Volterra/Fredholm integral equation; the $H$ -functions arise from it.
Renewal theory & actuarial science — the renewal equation $m (t) = F (t) + \int_{0}^{t} m (t - s) d F (s)$ is a Volterra equation governing insurance ruin probabilities and queueing.
Geophysical inversion — recovering subsurface density/conductivity from surface measurements is a first-kind integral equation requiring regularization.
Population dynamics — the Lotka integral equation and age-structured (McKendrick) models are Volterra equations for birth rates.

Historical Development

1782 — Laplace introduces what is essentially the Laplace transform while studying an integral relation; an early "integral equation" structure.
1823 — Abel poses and solves the tautochrone problem, yielding the Abel integral equation $\int_{0}^{x} (x - t)^{- 1/2} ϕ (t) d t = f (x)$ — the first explicitly solved singular integral equation.
1837 — Liouville solves certain equations by an iterative (successive-approximation) method, foreshadowing the Neumann series.
1888–1896 — Volterra systematically studies equations with variable upper limit, proving existence/uniqueness via iteration; these now bear his name.
1900 — Fredholm publishes the determinant theory ( $[6]$ – $[7]$ ), proving the alternative for the second-kind equation by a discretization-and-limit argument — a landmark.
1904–1910 — Hilbert generalizes via quadratic forms in infinitely many variables, introducing spectra, complete orthonormal systems, and $ℓ^{2}$ .
1907 — Schmidt recasts the theory geometrically in (what becomes) Hilbert space, proving the eigenfunction expansion $[10]$ cleanly.
1918 — Fredholm/Riesz — F. Riesz develops the abstract theory of compact operators ("completely continuous"), freeing the alternative from kernels.
1932 — von Neumann axiomatizes Hilbert space and the spectral theorem, completing the abstraction that integral equations had set in motion.

Subtleties & Limits

Nonlinear and singular integral equations remain frontier territory. While the linear compact theory is essentially complete, nonlinear integral equations (Hammerstein, Urysohn type) require degree theory, fixed-point methods, and bifurcation analysis with no universal solvability criterion. Singular integral equations with Cauchy or hypersingular kernels demand the full machinery of the Riemann–Hilbert Problem and Mikhlin–Calderón–Zygmund theory, and their numerical treatment (Nyström, collocation on graded meshes) is delicate near corners and edges. Ill-posed first-kind equations in high dimensions — central to inverse problems and machine-learning-driven imaging — lack optimal, parameter-free regularization strategies; choosing the Tikhonov parameter remains partly heuristic. Finally, the spectral analysis of non-self-adjoint integral operators (where eigenfunctions need not be complete and pseudospectra dominate behavior) is an active area with implications for stability of fluid flows and non-Hermitian quantum systems.

Integral Equations

The Core Insight

Classification

Types of Integral Equations

The Integral Operator

The Fredholm Theory

Fredholm Alternative

Fredholm Determinant and Resolvent

Why It Holds / Derivation

Neumann Series for Small $\lambda$

Hilbert–Schmidt Expansion

David Hilbert | 1862–1943

Building Intuition

Spectrum of a compact integral operator: eigenvalues accumulate only at the origin, leaving the resolvent set open everywhere else.

Worked Examples

A Degenerate (Separable) Kernel

A Volterra Equation by Neumann Series

Common Misconceptions

Connections

Real-World Applications

Historical Development

Subtleties & Limits

See Also

Before this

Where this leads