1 Definition and notation

The Dirac delta is an idealized mathematical object used to model a quantity concentrated at a single point. In applied contexts, it is treated as a point impulse, point source, or point mass. In rigorous analysis, it is defined not as an ordinary function but as a distribution: its meaning is given by how it acts on smooth test functions through integration.

1.1 Informal description

Informally, the delta is described as being zero everywhere except at one point, where it is “infinitely large” in such a way that its total integral is 1. This description is useful for intuition, especially in physics and engineering, but it is not a literal pointwise definition. The object is instead designed to isolate the value of another function at a chosen location.

1.2 Distributional definition

In distribution theory, the Dirac delta at a point \(a\) is defined by its action on test functions. Rather than assigning numerical values at each point of its domain, it assigns a number to each admissible test function according to a simple rule involving evaluation at \(a\).

1.2.1 Action on test functions

If \(\varphi\) is a test function, the delta at \(a\) acts by \[ \delta_a(\varphi) = \varphi(a). \] This means that when the delta is paired with a smooth function of compact support, it returns the function’s value at the target point. This property is the basis for its use in modeling pointwise sampling and concentrated sources.

1.2.2 Normalization property

The delta is normalized so that its integral over a domain containing its support is 1. In heuristic notation, \[ \int_{-\infty}^{\infty} \delta(x-a)\,dx = 1. \] This normalization expresses unit total mass or unit total impulse. It also ensures that the delta acts as an identity element for evaluation under integration.

1.3 Notation and common symbols

The most common notation is \(\delta(x)\) for the delta centered at the origin and \(\delta(x-a)\) for a delta centered at \(a\). In several-variable settings, one may write \(\delta(\mathbf{x}-\mathbf{a})\). In physics, the same symbol is often used for the time-domain impulse, spatial point source, or a generalized eigenstate overlap.

2 Mathematical foundations

The Dirac delta belongs to the framework of generalized functions, where objects are defined by their action on auxiliary functions rather than by ordinary pointwise formulas. This approach makes it possible to manipulate singular sources rigorously while preserving the formal rules used in applications.

2.1 Generalized functions and distributions

A distribution is a continuous linear functional on a space of test functions. The delta is one of the simplest examples, and it serves as a basic model for singular behavior in analysis. Its definition does not require a classical graph or values at every point.

2.1.1 Test function spaces

Test functions are typically smooth functions with compact support, chosen so that integration against singular objects remains well behaved. These spaces provide a controlled setting in which derivatives, limits, and localization can be handled cleanly. The delta is defined on such spaces because its action depends only on the value of the test function at one point.

2.1.2 Linear functionals

As a linear functional, the delta satisfies \[ \delta(a\varphi + b\psi) = a\,\delta(\varphi) + b\,\delta(\psi). \] This linearity is essential for compatibility with superposition. It also distinguishes distributions from pointwise formulas, since the object is characterized by how it processes entire functions rather than isolated numerical inputs.

2.2 Comparison with ordinary functions

Although the delta is often written in a function-like form, it does not fit the usual theory of real- or complex-valued functions on a continuum. Its usefulness comes from a formal resemblance to ordinary densities combined with a singular concentration at one point.

2.2.1 Why the delta is not a classical function

No ordinary integrable function can be zero everywhere except at one point and still have integral 1. A single point has measure zero, so a classical function with that behavior would contribute nothing to an integral. The delta avoids this difficulty by living outside the class of standard pointwise-defined functions.

2.2.2 Approximation by sequences

The delta is often approached as the limit of a family of ordinary functions that become narrower and taller while preserving unit area. Such sequences provide intuition and are widely used in computation and modeling. Different approximating families can converge to the same distribution even though their pointwise forms differ.

2.3 Scaling and translation

The delta responds predictably to shifts and changes of scale. These transformations are central in applications, where sources may be located away from the origin or rescaled by coordinate changes.

2.3.1 Shifted delta

A shifted delta, written \(\delta(x-a)\), concentrates at \(x=a\). It selects the value of a function at that point and is the standard representation of a point source located away from the origin. Translation does not alter the unit mass of the distribution.

2.3.2 Scaled delta

Under a change of variables, the delta acquires a reciprocal scaling factor. Formally, \[

\delta(c x) = \frac{1}{c}\delta(x)

\] for nonzero \(c\). This factor preserves total integral and reflects the compression or expansion of coordinates.

3 Core properties

The key features of the delta arise from its role as an evaluation device under integration. These properties make it especially effective in both exact derivations and formal manipulations.

3.1 Sifting property

The sifting property is the defining operational feature of the delta. It extracts the value of a function at a single point, provided the function is sufficiently regular near that point.

3.1.1 Evaluation under integration

For a suitable function \(f\), \[ \int f(x)\,\delta(x-a)\,dx = f(a). \] This identity explains why the delta is used to represent pointwise sampling, impulse responses, and localized inputs. In practice, it converts an integral into a direct evaluation.

3.1.2 Higher-dimensional forms

In multiple dimensions, the delta may be written as \[ \delta(\mathbf{x}-\mathbf{a}) = \prod_{i=1}^n \delta(x_i-a_i). \] Its sifting property then gives \[ \int_{\mathbb{R}^n} f(\mathbf{x})\,\delta(\mathbf{x}-\mathbf{a})\,d\mathbf{x} = f(\mathbf{a}). \] This form is widely used for localized sources in vector calculus and field theory.

3.2 Symmetry and support

The delta is highly concentrated and has strong symmetry properties when centered at the origin. Its support is a single point, which makes it one of the most localized distributions in analysis.

3.2.1 Support at a point

The support of \(\delta(x-a)\) is the singleton set \(\{a\}\). Outside that point, the distribution has no effect on test functions whose support avoids \(a\). This extreme localization underlies its use as a mathematical model of point concentration.

3.2.2 Evenness

The centered delta is an even distribution: \[ \delta(-x)=\delta(x). \] This follows from its symmetric placement at the origin. The property is often used in Fourier analysis and in simplifying integrals involving symmetric kernels.

3.3 Derivatives of the delta

Derivatives of the delta are defined in the distributional sense and represent higher-order singular behavior. They are common in advanced analysis, differential equations, and signal theory.

3.3.1 Distributional derivatives

The derivative \(\delta'\) is defined by \[ \delta'(\varphi) = -\delta(\varphi'). \] More generally, higher derivatives are obtained by repeated differentiation in the distributional sense. These objects encode localized rates of change rather than ordinary slopes.

3.3.2 Integration by parts

When paired with a smooth function, \[ \int f(x)\,\delta'(x-a)\,dx = -f'(a). \] This identity is a direct consequence of integration by parts and is frequently used in derivations involving localized forcing terms. It extends naturally to higher derivatives, producing alternating signs and higher derivatives of the test function.

4 Representations and approximations

The delta can be represented through limiting processes, integral formulas, and discrete analogues. These viewpoints are valuable in computation, harmonic analysis, and the connection between continuous and discrete models.

4.1 Approximate identities

An approximate identity is a family of ordinary functions that converges to the delta in the distributional sense. Such families are essential for intuition and for numerical approximation of singular sources.

4.1.1 Gaussian approximation

A common approximation is the normalized Gaussian with shrinking variance: \[ \delta(x) \approx \frac{1}{\sqrt{2\pi}\sigma}e^{-x^2/(2\sigma^2)} \] as \(\sigma \to 0\). The Gaussian is smooth, symmetric, and easy to analyze, making it a standard model for localization.

4.1.2 Rectangular pulse approximation

Another approximation uses a narrow rectangle of width \(\varepsilon\) and height \(1/\varepsilon\). Its integral is always 1, and its support collapses to a point as the width decreases. This construction is intuitive in elementary signal processing.

4.1.3 Lorentzian approximation

The Lorentzian or Cauchy kernel can also approximate the delta: \[ \frac{1}{\pi}\frac{\varepsilon}{x^2+\varepsilon^2}. \] As \(\varepsilon \to 0\), it becomes increasingly concentrated near the origin. This family appears frequently in complex analysis and resonance theory.

4.2 Integral representations

The delta admits several integral expressions that are useful in Fourier analysis and in contour methods. These formulas are often interpreted distributionally.

4.2.1 Fourier representation

A standard representation is \[ \delta(x) = \frac{1}{2\pi}\int_{-\infty}^{\infty} e^{ikx}\,dk. \] This identity expresses the delta as the inverse Fourier transform of a constant function. It is central to the duality between localization in space and spread in frequency.

4.2.2 Complex-analytic forms

In complex analysis, the delta can be represented through boundary values of analytic functions or through contour integrals in suitably interpreted limits. Such formulas are useful in residue computations and in the study of Green’s functions. They are typically understood in a distributional or limiting sense rather than as convergent ordinary integrals.

4.3 Discrete analogues

Discrete mathematics has its own counterpart to the delta, designed for grids and finite sequences. These analogues preserve the idea of single-point selection in a sampled setting.

4.3.1 Kronecker delta

The Kronecker delta is defined by \[ \delta_{ij}= \begin{cases} 1, & i=j,\\ 0, & i\ne j. \end{cases} \] It plays the same role for sequences and matrices that the Dirac delta plays for continuous variables. It is used extensively in linear algebra, summation formulas, and discrete systems.

4.3.2 Sampling on lattices

On a lattice, a discrete delta localizes mass at a single grid point. This concept appears in finite-difference methods, lattice models, and digital signal processing. It provides a bridge between continuous distributions and numerical approximations.

5 Applications in mathematics and physics

The Dirac delta is a standard tool wherever a model requires a sharply localized input, source, or constraint. Its flexibility makes it common in theoretical derivations and in practical calculations.

5.1 Differential equations

In differential equations, the delta is used to represent impulsive forcing or point-supported sources. It enables the construction of solutions through fundamental solutions and Green’s functions.

5.1.1 Green's functions

A Green’s function is the response of a linear operator to a delta source. Once such a response is known, more general solutions can often be built by superposition. This idea is central in boundary-value problems and operator theory.

5.1.2 Impulse forcing

An equation driven by a delta input models a sudden action applied at a single time or place. The resulting solution often exhibits a jump or a sharp change in derivative. This is common in mechanics, electrical circuits, and wave phenomena.

5.2 Signal processing

In signal processing, the delta represents an ideal impulse and is used to characterize linear time-invariant systems. It serves as a basic probe for system response and as a building block for convolution.

5.2.1 Impulse response

The impulse response of a system is its output when the input is a delta. Because any reasonable signal can be decomposed into shifted impulses, the impulse response determines the system completely within a linear framework. This makes the delta fundamental to filter theory.

5.2.2 Convolution

Convolution with the delta leaves a signal unchanged: \[ f * \delta = f. \] More generally, convolution with a shifted delta produces a translated signal. This property is frequently used to simplify integral equations and to interpret time delays.

5.3 Quantum mechanics

In quantum mechanics, the delta appears in discussions of continuous spectra, position representations, and idealized states. It is a formal but highly useful element of bra-ket notation and Hilbert-space methods.

5.3.1 Position eigenstates

Position eigenstates are represented formally by delta-normalized wavefunctions. Their inner products involve delta functions rather than ordinary numbers. This reflects the continuous nature of position measurements in the idealized formalism.

5.3.2 Completeness relations

Completeness relations often include an integral over delta-normalized states. Such formulas express the fact that a continuous basis spans the relevant state space in a generalized sense. They are essential in deriving expansion formulas and transition amplitudes.

5.4 Electromagnetism and continuum models

The delta is used to describe sharply localized charge and mass distributions. It is also employed in continuum mechanics to encode concentrated sources within field equations.

5.4.1 Point charges

A point charge is modeled by a charge density proportional to the delta. This allows electrostatic potentials and fields to be derived from equations with singular sources. The representation is idealized but highly effective in theoretical work.

5.4.2 Point masses and source terms

In mechanics, point masses are represented by mass densities involving the delta. Similar source terms appear in elasticity, fluid models, and heat conduction when a localized input must be specified. The delta provides a concise way to express such concentrations.

Several other mathematical objects are closely related to the Dirac delta through localization, discrete sampling, or distribution theory. Together they form a network of tools for handling singular and idealized behavior.

6.1 Step functions and Heaviside functions

The Heaviside step function is the distributional antiderivative of the delta. It models a sudden jump from one state to another. In differential equations, step functions and deltas often appear together as complementary descriptions of abrupt change.

6.2 Dirac comb

The Dirac comb is a periodic array of delta functions placed at equally spaced points. It is important in sampling theory, Fourier analysis, and crystallographic models. Its periodic structure links continuous and discrete frequency behavior.

6.3 Green's distribution kernels

Green’s distribution kernels generalize Green’s functions to distributional settings. They encode the response of an operator to localized forcing in a form suitable for singular sources. Such kernels are a natural extension of the delta-based viewpoint.

6.4 Schwartz distributions and tempered distributions

Schwartz distributions provide the broader theoretical setting in which the delta is rigorously defined. Tempered distributions form a subspace adapted to Fourier analysis and polynomial growth conditions. The delta belongs to both frameworks and is one of their most familiar examples.