research32 min studyLesson 817 in Mathematics

Eigenfunction nodal domains and Courant theorem

Count connected sign regions of real eigenfunctions with explicit multiplicity and zero-mode indexing.

On this page

What you’ll understand

  • Define nodal domains as connected components away from a real eigenfunction's zero set.
  • Apply Courant's upper bound with one-based multiplicity indexing.
  • Separate eigenfunctions from arbitrary finite sums of different frequencies.

Eigenvalues describe oscillation scales, but eigenfunctions reveal how that oscillation partitions space. A nodal domain is a connected region on which a real eigenfunction has a fixed nonzero sign. Courant's theorem bounds how many such regions an indexed mode can have. We assume a connected smooth compact manifold, either closed or with a smooth boundary and standard Dirichlet conditions, and use real eigenfunctions. Complex-valued phase patterns are not sign regions and require a different question.

Connected components, not just zero counts

Remove the zero set of a nonzero real eigenfunction and take the connected components of what remains. Continuity makes its sign constant on each component. In one dimension the zero set cuts intervals or circles into pieces, but in higher dimensions counting isolated zeros is not the relevant operation. A nodal set can have intersecting branches or singular points; the component definition still applies. Boundary points of a Dirichlet domain are not themselves extra nodal domains.

The first Dirichlet eigenfunction on a connected smooth bounded domain can be chosen strictly positive in the interior, and its eigenvalue is simple. On a closed connected manifold, the first entry is instead the constant zero mode with one nodal domain. These are different low-index starting points. A mean-zero nonconstant real eigenfunction must take both signs, so it has at least two nodal domains, but the number can depend strongly on geometry and on the selected vector inside a multiple eigenspace.

N(ϕ)=M∖{x:ϕ(x)=0},ν(ϕ)=#{connected components of N(ϕ)}\mathcal N(\phi)=M\setminus\{x:\phi(x)=0\},\qquad \nu(\phi)=\#\{\text{connected components of }\mathcal N(\phi)\}
For a boundary problem, take components in the domain interior; the definition concerns a real eigenfunction.↔ Scroll wide equations horizontally.

Courant counts multiplicity slots

With eigenvalues listed increasingly from index one and repeated by multiplicity, Courant's theorem says that a kth eigenfunction has at most k nodal domains. The list includes zero on a closed manifold. A repeated eigenvalue has several slots and no preferred basis vector within its eigenspace. If only the eigenvalue is specified, using the last occupied index gives an unambiguous valid upper bound for every vector in that eigenspace; a particular ordered-basis assignment can give a sharper bound.

The variational proof restricts an eigenfunction to selected nodal domains and extends those restrictions by zero. They are admissible energy-form functions with disjoint supports, and every nonzero combination has the same Rayleigh quotient as the original eigenfunction. Too many nodal pieces permit orthogonality to enough low modes. Min–max and the equality case, together with unique continuation in the stated smooth setting, yield a contradiction. This sketch relies on admissibility of the zero extensions; drawing disjoint colored regions alone is not the proof.

Δϕk=λkϕk⟹ν(ϕk)≤k,λ1≤λ2≤⋯ with multiplicity\Delta\phi_k=\lambda_k\phi_k\quad\Longrightarrow\quad\nu(\phi_k)\leq k,\qquad \lambda_1\leq\lambda_2\leq\cdots\text{ with multiplicity}
The eigenvalue index is one-based, including the constant mode when present.↔ Scroll wide equations horizontally.

Sharpness is a special property

An eigenfunction attaining the bound is called Courant sharp in the relevant indexing. The interval sine modes attain it at every Dirichlet index. Higher-dimensional domains behave differently; Courant supplies an upper bound, not a rule that every mode must have exactly its index in sign components. Additional theorems can limit how often equality occurs, but their dimension and boundary hypotheses must be stated separately. No such stronger universal classification follows merely from the inequality written here.

A sum of eigenfunctions with different eigenvalues is generally not an eigenfunction. Its nodal pattern therefore cannot be assigned the index of its largest visible frequency and put into Courant's theorem without an additional result. Mixtures within one eigenspace remain eigenfunctions, but their nodal sets may vary. This distinction is especially important for finite Fourier visualizations, which often display a mixture rather than a single spectral mode even when every term is individually a familiar sine wave.

A Courant-sharp interval mode

On the Dirichlet interval (0,2)(0,2), analyze ϕ(x)=sin⁡(3πx/2)\phi(x)=\sin(3\pi x/2). Find its eigenvalue, index, interior zeros, and nodal domains, then test Courant's bound.

  1. Differentiating twice gives −ϕ′′=(9π2/4)ϕ-\phi''=(9\pi^2/4)\phi. The Dirichlet eigenvalues are (nπ/2)2(n\pi/2)^2 with positive n, so this is the simple third eigenvalue.

  2. Its interior zeros occur at x=2/3x=2/3 and x=4/3x=4/3. The endpoint zeros enforce the boundary condition and do not count as additional interior components.

  3. The three open subintervals between those zeros and the endpoints have alternating signs. Thus the nodal-domain count is three, equal to the eigenvalue index and attaining Courant's upper bound.

The result

This one-dimensional mode is Courant sharp, with its count determined by connected sign intervals rather than by including endpoint zeros.

Explore the model

Interactive lab

Interactive lab · spectral

Contrast a mode with a mixture

Pause/Reset, then set both mode count to 1 and coefficient decay to 0. At zero time sin(x) has two nodal domains on the periodic circle. Reset and set both count to 2 and decay to 0; sin(x)+sin(2x)=sin(x)(1+2cos(x)) has four. Inspect the extra zero crossings without playing the shared phase clock.

Contrast a mode with a mixtureA finite sum of sin(nx − t)/nᵖ. Playback changes t; increasing the cutoff includes finer spatial scales. This is not a proof about a continuum limit or renormalization. Current readings: Sum at x = π/2: 0.867; Smallest wavelength: 2π / 5; Highest-mode coefficient: 0.2. All parameters are adjustable with the labeled controls below.01.573.144.716.28−2.56−1.2801.282.56x · radiansFinite sine sum
Finite sumHighest included mode
Sum at x = π/2
0.867
Smallest wavelength
2π / 5
Highest-mode coefficient
0.2
t = 0 s

A finite sum of sin(nx − t)/nᵖ. Playback changes t; increasing the cutoff includes finer spatial scales. This is not a proof about a continuum limit or renormalization. Parameter changes pause playback and recompute the model at the current elapsed time. Reset restores both parameters and returns time to zero. Playback starts paused. Adjust either parameter, then choose Play or Step.

Look forThe first setting is one eigenfunction; the second mixes different eigenvalues, so its count is not a direct Courant test for one indexed mode.

Check your understanding

Which function is outside the direct scope of Courant's single-eigenfunction nodal bound?

Try it yourself

On the circle of length 2π2\pi, count nodal domains of sin⁡(2x)\sin(2x) and locate its eigenvalue in the full multiplicity list. Give a valid bound using the last slot of its eigenspace and compare with its actual count.

Take a moment to make a prediction or work it out on paper. Open a hint when you need one.

Give me a hint

Begin the list with the constant zero mode and count both sine and cosine modes at each positive frequency.

Compare with the solution

The eigenvalue is four, occupying slots four and five in 0,1,1,4,4,…0,1,1,4,4,\ldots. The function has four nodal domains on the circle, separated by its four periodic zeros. The last-slot bound gives at most five, which is valid but not sharp. Assigning this eigenfunction to slot four gives the sharper bound four. The example shows why multiplicity-aware slot conventions must accompany an index-based statement.

Take it with you

  • Nodal domains are connected sign components of a real eigenfunction.
  • Courant uses an ordered multiplicity list, including zeros when present.
  • Different-eigenvalue mixtures require a different nodal analysis.

Read further

Original sources and deeper explanations to follow at your own pace.

Optional progress, saved in this browser when available. Revisit any lesson at any time.

Lesson not marked complete.