Dirichlet Spaces: Theory and Applications

• Dirichlet spaces are function spaces defined by the finiteness of the Dirichlet energy integral, emphasizing boundary norm equivalence and reproducing kernel properties.
• They extend to weighted, vector-valued, and fractional order settings, facilitating analysis on fractals, non-smooth domains, and metric measure spaces.
• Their study bridges complex analysis, harmonic analysis, and spectral theory, impacting applications in conformal welding, geometric analysis, and synthetic curvature bounds.

A Dirichlet space is a function space governed by an energy integral, the Dirichlet integral, and plays a central role in complex analysis, harmonic analysis, and potential theory. The classical Dirichlet space comprises holomorphic functions on the unit disk $\mathbb D \subset \mathbb C$ with square-integrable derivatives. Modern theory extends Dirichlet spaces to weighted, vector-valued, higher-dimensional, and non-smooth settings, including on fractals and metric measure spaces. The structure and equivalence of function and boundary norms in various geometric contexts is a principal theme, particularly in the context of curves and domains of complex geometry.

1. Classical Dirichlet Space and Boundary Norms

Let $\mathcal D$ denote the Dirichlet space on the unit disk $\mathbb D$: $$\mathcal D = \left\{ f \in \operatorname{Hol}(\mathbb D) : \int_{\mathbb D} |f'(z)|^2 dA(z) < \infty \right\}.$$ The canonical norm is

$$\|f\|^2_{\mathcal D} = |f(0)|^2 + \int_{\mathbb D} |f'(z)|^2 dA(z).$$

The boundary behavior is encoded by

$\mathcal D = H^{1/2}(S^1) \cap H(\mathbb D),$

where $H^{1/2}(S^1)$ is the Sobolev–Slobodeckij space on the unit circle. The Dirichlet space is a reproducing kernel Hilbert space (RKHS) with the kernel

$$K_z(w) = \sum_{n=0}^\infty \frac{(z \bar w)^n}{n+1},$$

with boundary properties sharply contrasting those of the $H^2$ Hardy space (Arcozzi et al., 2010).

For a $C^\infty$ function $U$ compactly supported in $\mathbb R^2$, its boundary restriction $u$ satisfies

$$\int_{\mathbb D} |\nabla U_i|^2 dxdy = \int_{\mathbb C \setminus \overline{\mathbb D}} |\nabla U_e|^2 dxdy = \frac{1}{2\pi} \iint_{\mathbb S \times \mathbb S} \left|\frac{u(z_1) - u(z_2)}{z_1 - z_2}\right|^2 |dz_1||dz_2|,$$

demonstrating equivalence of the gradient, harmonic extension, and Douglas boundary norms on $\mathbb S$. For a general rectifiable Jordan curve $\Gamma$, there exist analogs, although equality may fail except under additional geometric hypotheses (Wei et al., 2024).

2. Dirichlet Spaces on Curves and Domains (Chord-Arc and Quasicircle Theory)

Given a rectifiable Jordan curve $\Gamma \subset \mathbb C$, the Douglas boundary semi-norm is

$$\|u\|^2_{D(\Gamma)} = \frac{1}{2\pi} \iint_{\Gamma \times \Gamma} \left|\frac{u(\zeta_1) - u(\zeta_2)}{\zeta_1 - \zeta_2}\right|^2\,|d\zeta_1|\,|d\zeta_2|.$$

For $u \in H^{1/2}(\Gamma)$, there are unique harmonic extensions $U_i$ and $U_e$ with Dirichlet energies $D(U_i), D(U_e)$ inside and outside $\Gamma$.

Main theorem: For $\Gamma$,

• $\Gamma$ is chord–arc $\Leftrightarrow$ the Douglas norm, interior, and exterior Dirichlet norms for boundary data are all equivalent (with constants depending only on the chord–arc constant). This equivalence breaks down for quasicircles that are not chord–arc (Wei et al., 2024).
• The proof utilizes the existence of a global bi-Lipschitz map intertwining the domains and the transformation of Dirichlet energies under conformal welding.
• Failure of equivalence is exhibited in non-chord–arc quasicircles (e.g., non-Smirnov domains), where norm comparability only holds in one direction.

An immediate corollary is that the trace space $H^{1/2}(\Gamma)$ is equivalent to the critical Besov space $B^{1/2}_{2,2}(\Gamma)$, and composition with quasisymmetric weldings preserves the Dirichlet norm if and only if $\Gamma$ is chord–arc (Wei et al., 2024).

On multiply connected domains bounded by quasicircles or Weil–Petersson quasicircles, the Dirichlet space is modeled as a direct sum of disk Dirichlet spaces, with boundary values forming the graph of a bounded Grunsky operator (Radnell et al., 2017, Radnell et al., 2013).

3. Weighted, Vector-Valued, and Abstract Dirichlet Spaces

Weighted Dirichlet Spaces

Weighted Dirichlet spaces $D_\omega$ are defined for holomorphic $f$ on $\mathbb D$: $$D_\omega(f) = \int_{\mathbb D} |f'(z)|^2 \omega(z)\,dA(z),$$ where $\omega$ is a nonnegative integrable and typically superharmonic function (EL-Fallah et al., 2015). Special cases include:

• Power weights $\omega(z) = (1 - |z|^2)^\alpha$, $0 \leq \alpha \leq 1$;
• Harmonic weights, where $\omega$ is the Poisson integral of a positive measure on $\mathbb T$.

Among these, only atomic Poisson kernel weights (e.g., $\omega_{\zeta}(z) = (1 - |z|^2)/|\zeta - z|^2$) produce Dirichlet spaces isometrically isomorphic to de Branges–Rovnyak spaces $\mathcal H(b)$ (EL-Fallah et al., 2015, Dellepiane et al., 22 May 2025). A full characterization for harmonic weights is given for Dirichlet spaces coinciding with $H(b)$ spaces, determined via the spectrum and factorization properties of the outer function $a$ in the so-called Pythagorean pair $(b, a)$ (Dellepiane et al., 22 May 2025).

Dirichlet Spaces Associated to Trees

Vector-valued Dirichlet spaces $\mathscr H_q$ can be constructed as RKHS of $E$-valued holomorphic functions on $\mathbb D$, with structure governed by a locally finite, rooted directed tree $\mathscr T$ (Chavan et al., 2017). The reproducing kernel adapts multiplicities and weights according to the tree's combinatorics, interpolating between classical scalar Dirichlet spaces and genuinely vector-valued models. Classification up to unitary equivalence is governed by the total "branching excess" at each depth.

Higher and Fractional Order Dirichlet-Type Spaces

Generalizations involve Dirichlet integrals of higher-order derivatives with possibly distributional coefficients, yielding Hilbert spaces on which shifts act as $m$-isometries (Rydhe, 2018). Allowability of the weights is analyzed through positivity and boundedness conditions, offering a function-theoretic model for $m$-isometries with concrete examples involving signed measures and unbounded operators.

The Dirichlet space can also be extended to domains in $\mathbb R^n$, such as balls or half-spaces, where identities relating the Dirichlet gradient energy, Fourier, and fractional Sobolev norms hold:

• On balls $B^n$, energies match the $H^{1/2}(S^{n-1})$ semi-norm of boundary traces;
• On half-spaces $\mathbb R^n_+$, energies match $\int_{\mathbb R^n} |\xi| |\widehat{f}(\xi)|^2 d\xi$ and the difference quotient form (Yang et al., 9 Feb 2025, Fardi et al., 2015).

4. Abstract and Geometric Dirichlet Spaces: Metric Measure, Homogeneous Type, and Non-smooth Settings

Within the most general framework, Dirichlet spaces are defined as strongly local, quasi-regular Dirichlet forms on metric measure spaces $(X,d,\mu)$ with full support. Two key analytic structures govern their function spaces:

• Doubling property: Ensures polynomial volume growth.
• Poincaré inequality: Relates mean oscillation to local energy and underlies parabolic Harnack inequalities, guaranteeing joint continuity of the heat kernel (Opadara et al., 22 Dec 2025).

These prerequisites enable Gaussian-type heat kernel bounds, Hölder continuity, and the spectral decomposition necessary for constructing band-limited frames. Coupled with the spectral calculus for the generator $L$, one obtains nearly exponentially localized frames and corresponding atomic decompositions for Besov and Triebel-Lizorkin spaces (Opadara et al., 22 Dec 2025).

Applications encompass Euclidean spaces, compact manifolds, Lie groups of polynomial growth, and fractals with sub-Gaussian, nontrivial heat kernels.

5. Synthetic Curvature, Distributional and Metric Extensions

Dirichlet spaces arise as the analytic framework for synthetic Ricci lower bounds via distribution-valued perturbations of the energy form ("tamed spaces") (Erbar et al., 2020). Given a moderate distribution $\kappa$, the Feynman–Kac semigroup $P_t^{\kappa}$ associated to $\kappa$ yields equivalence between distributional Bochner inequalities (lower Ricci curvature bounds) and gradient estimates for the heat flow. This machinery enables the study of:

• Riemannian or sub-Riemannian manifolds with singular Ricci curvature;
• Manifolds with boundary under Neumann or Dirichlet conditions;
• Spaces with highly oscillatory or singular potentials.

6. Spectral Theory, Sample Path Properties, and Non-smooth Examples

In the setting of metric measure Dirichlet spaces with ultracontractive and irreducible semigroups, the heat kernel admits uniform bounds, implying that restrictions to open sets of finite measure yield a pure-point spectrum without regularity assumptions on the boundary (Carfagnini et al., 2024). The spectral gap is positive for connected $U$, and the first eigenfunction is strictly positive.

Functional consequences include:

• Log-Sobolev, Poincaré, and small deviation principles for the associated Hunt process, with explicit asymptotics for both small-time and large-time regimes.
• Applicability in Riemannian, sub-Riemannian, fractal, and jump-process contexts, with all main results obtained solely from heat kernel control.

7. Connections, Generalizations, and Open Problems

Dirichlet spaces are foundational in several branches:

• Complex analysis: Theory of holomorphic and harmonic functions with controlled boundary regularity.
• Operator theory: Model Hilbert spaces for $m$-isometries and connections to de Branges–Rovnyak theory (Rydhe, 2018, Dellepiane et al., 22 May 2025).
• Harmonic analysis: Atomic decomposition and function space theory via frames in homogeneous-type spaces (Opadara et al., 22 Dec 2025).
• Geometric analysis: Analysis of PDEs and heat flow on non-smooth and metric measure spaces, extension of curvature-dimension theories (Erbar et al., 2020).

Current frontiers include the geometric characterization of Carleson measures for Dirichlet spaces in all settings, interpolation sequences, operator-theoretic dualities for weighted spaces, and the role of Dirichlet spaces in Teichmüller theory (e.g., Weil–Petersson class boundaries) (Wei et al., 2024).

---

Key references:

• "Dirichlet spaces over chord-arc domains" (Wei et al., 2024)
• "The Dirichlet Space: A Survey" (Arcozzi et al., 2010)
• "Dirichlet spaces with superharmonic weights and de Branges–Rovnyak spaces" (EL-Fallah et al., 2015)
• "On the equality of De Branges-Rovnyak and Dirichlet spaces" (Dellepiane et al., 22 May 2025)
• "Dirichlet space of domains bounded by quasicircles" (Radnell et al., 2017)
• "Dirichlet Spaces of Homogeneous Type Via Heat Kernel" (Opadara et al., 22 Dec 2025)
• "Tamed spaces — Dirichlet spaces with distribution-valued Ricci bounds" (Erbar et al., 2020)