View source for Singular integral operators on closed curves

In [[mathematics]], '''[[singular integral operators]] on closed curves''' arise in problems in [[analysis]], in particular [[complex analysis]] and [[harmonic analysis]]. The two main singular integral operators, the Hilbert transform and the Cauchy transform, can be defined for any smooth Jordan curve in the complex plane and are related by a simple algebraic formula. In the special case of [[Fourier series]] for the unit circle, the operators become the classical [[Szegő kernel|Cauchy transform]], the [[orthogonal projection]] onto [[Hardy space]], and the [[Hilbert transform]] a real orthogonal [[linear complex structure]]. In general the Cauchy transform is a non-self-adjoint [[idempotent]] and the Hilbert transform a non-orthogonal [[linear complex structure|complex structure]]. The range of the Cauchy transform is the Hardy space of the bounded region enclosed by the Jordan curve. The theory for the original curve can be deduced from that of the unit circle, where, because of rotational symmetry, both operators are classical [[singular integral operators of convolution type]]. The Hilbert transform satisfies the [[Sokhotski–Plemelj theorem|jump relations of Plemelj and Sokhotski]], which express the original function as the difference between the boundary values of holomorphic functions on the region and its complement. Singular integral operators have been studied on various classes of functions, including Hölder spaces, L<sup>''p''</sup> spaces and Sobolev spaces. In the case of L<sup>2</sup> spaces—the case treated in detail below—other operators associated with the closed curve, such as the [[Szegő kernel|Szegő projection]] onto Hardy space and the [[Neumann–Poincaré operator]], can be expressed in terms of the Cauchy transform and its adjoint.

==Operators on the unit circle==
{{main|Singular integral operators of convolution type}}
If ''f'' is in L<sup>2</sup>('''T'''), then it has a Fourier series expansion<ref>{{harvnb|Torchinsky|2004|pp=65–66}}</ref><ref>{{harvnb|Bell|1992|pp=14–15}}</ref>

:<math>\displaystyle{f(\theta)=\sum_{n\in {\mathbf Z}} a_n e^{in\theta}.}</math>

[[Hardy space]] H<sup>2</sup>('''T''') consists of the functions for which the negative coefficients vanish, ''a''<sub>''n''</sub> = 0 for ''n'' < 0. These are precisely the square-integrable functions that arise as boundary values of holomorphic functions in the unit disk |''z''| < 1. Indeed, ''f'' is the boundary value of the function

:<math>\displaystyle{F(z)=\sum_{n\ge 0} a_n z^n,}</math>

in the sense that the functions

:<math>\displaystyle{f_r(\theta)=F(re^{i\theta})},</math>

defined by the restriction of ''F'' to the concentric circles |''z''| = ''r'', satisfy

:<math>\displaystyle{\|f_r-f\|_2 \rightarrow 0}</math> as <math>\displaystyle{r \rightarrow 1}</math>.

The orthogonal projection ''P'' of L<sup>2</sup>('''T''') onto H<sup>2</sup>('''T''') is called the '''Szegő projection'''. It is a bounded operator on L<sup>2</sup>('''T''') with [[operator norm]] 1.

By Cauchy's theorem

:<math>\displaystyle{F(z)= {1\over 2\pi i} \int_{|\zeta|=1}  {f(\zeta)\over \zeta -z} \,d\zeta={1\over 2\pi} \int_{-\pi}^{\pi} {f(\theta) \over 1-e^{-i\theta}z} \, d\theta.}</math>

Thus

:<math>\displaystyle{F(re^{i\varphi})={1\over 2\pi} \int_{-\pi}^{\pi} {f(\varphi-\theta) \over 1-re^{i\theta}} \, d\theta.}</math>

When ''r'' equals 1, the integrand on the right hand side has a singularity at θ = 0. The '''truncated Hilbert transform''' is defined by

:<math>\displaystyle{H^\varepsilon f(\varphi) = {i\over \pi} \int_{\varepsilon\le |\theta| \le  \pi} {f(\varphi-\theta) \over 1-e^{i\theta}} \, d\theta={1\over \pi} \int_{|\zeta-e^{i\varphi}|\ge \delta} {f(\zeta)\over \zeta-e^{i\varphi}}\, d\zeta,}</math>

where δ = |1 – ''e''<sup>''i''ε</sup>|. Since it is defined as convolution with a bounded function, it is a bounded operator on L<sup>2</sup>('''T'''). Now

:<math>\displaystyle{H^\varepsilon{1}={i\over\pi}\int_\varepsilon^\pi 2 \Re (1-e^{i\theta})^{-1} \,  d\theta ={i\over\pi}\int_\varepsilon^\pi 1 \, d\theta = i - {i \varepsilon\over \pi}.}</math>

If ''f'' is a polynomial in ''z'' then

:<math>\displaystyle{H^\varepsilon f(z) - {i(1-\varepsilon)\over \pi} f(z)={1\over \pi i} \int_{|\zeta -z|\ge \delta} {f(\zeta)-f(z)\over \zeta -z} \, d\zeta.}</math>

By Cauchy's theorem the right hand side tends to 0 uniformly as ε, and hence δ, tends to 0. So

:<math>\displaystyle{H^\varepsilon f \rightarrow if}</math>

uniformly for polynomials. On the other hand, if ''u''(''z'') = ''z'' it is immediate that

:<math>\displaystyle{\overline{H^\varepsilon f} = - u^{-1} H^\varepsilon( u \overline{f}).}</math>

Thus if ''f'' is a polynomial in ''z''<sup>−1</sup> without constant term

:<math> \displaystyle{H^\varepsilon f \rightarrow -i f}</math> uniformly.

Define the '''Hilbert transform''' on the circle by

:<math>\displaystyle{H=i(2P-I).}</math>

Thus if ''f'' is a trigonometric polynomial

:<math>\displaystyle{H^\varepsilon f \rightarrow Hf}</math> uniformly.

It follows that if ''f'' is any L<sup>2</sup> function

:<math>\displaystyle{H^\varepsilon f \rightarrow Hf}</math> in the L<sup>2</sup> norm.

This is a consequence of the result for trigonometric polynomials since the ''H''<sub>ε</sub> are uniformly bounded in [[operator norm]]: indeed their Fourier coefficients are uniformly bounded.

It also follows that, for  a continuous function ''f'' on the circle, ''H''<sub>ε</sub>''f'' converges uniformly to ''Hf'', so in particular pointwise. The pointwise limit is a [[Cauchy principal value]], written

:<math>\displaystyle{Hf= \mathrm{P.V.}\,{1\over \pi} \int {f(\zeta)\over \zeta-e^{i\varphi}}\, d\zeta.}</math>

The Hilbert transform has a natural compatibility with orientation-preserving diffeomorphisms of the circle.<ref>See:
*{{harvnb|Mikhlin|Prössdorf|1986}}
*{{harvnb|Segal|1981}}
*{{harvnb|Pressley|Segal|1986}}</ref> Thus if ''H'' is a diffeomorphism of the circle with

:<math>\displaystyle{H(e^{i\theta})=e^{ih(\theta)},\,\,\, h(\theta+2\pi)=h(\theta)+2\pi,}</math>

then the operators

:<math>\displaystyle{H^\varepsilon_h f(e^{i\varphi})={1\over \pi}\int_{|e^{ih(\theta)} -e^{ih(\varphi)}|\ge \varepsilon}
{f(e^{i\theta})\over e^{i\theta}-e^{i\varphi}} \, e^{i\theta}\, d\theta,}</math>

are uniformly bounded and tend in the strong operator topology to ''H''. Moreover, if ''Vf''(''z'') = ''f''(''H''(''z'')), then ''VHV''<sup>−1</sup> – ''H''  is an operator with smooth kernel, so a [[Hilbert–Schmidt operator]].

==Hardy spaces==
{{See also|Hardy space}}
The Hardy space on the unit circle can be generalized to any multiply connected bounded domain Ω with smooth boundary ∂Ω. The Hardy space H<sup>2</sup>(∂Ω) can be defined in a number of equivalent ways. The simplest way to define it is as the closure in L<sup>2</sup>(∂Ω) of the space of holomorphic functions on Ω which extend continuously to smooth functions on the closure of Ω. As [[Joseph Leonard Walsh|Walsh]] proved, in a result that was a precursor of [[Mergelyan's theorem]], any holomorphic function on Ω that extends continuously to the closure can be approximated in the uniform norm by a rational function with poles in the complementary region Ω<sup>''c''</sup>. If Ω is simply connected, then the rational function can be taken to be a polynomial. There is a counterpart of this theorem on the boundary, the [[Hartogs–Rosenthal theorem]], which states that any continuous function ∂Ω can be approximated in the uniform norm by rational functions with poles in the complement of ∂Ω. It follows that for a simply connected domain when ∂Ω is a simple closed curve, H<sup>2</sup>(∂Ω) is just the closure of the polynomials; in general it is the closure of the space of rational functions with poles lying off ∂Ω.<ref>See:
*{{harvnb|Bell|1992}}
*{{harvnb|Gamelin|2005}}
*{{harvnb|Duren|1970}}
*{{harvnb|Conway|1995}}
*{{harvnb|Conway|2000}}
</ref>

On the unit circle an L<sup>2</sup> function ''f'' with Fourier series expansion

:<math>\displaystyle{f(e^{i\theta})=\sum a_n e^{in\theta}}</math>

has a unique extension to a harmonic function in the unit disk given by the Poisson integral

:<math>\displaystyle{f(re^{i\theta}) =P_r f(e^{i\theta}) =\sum a_n r^{|n|} e^{in\theta}.}</math>

In particular

:<math>\displaystyle{\|P_rf\|^2=\sum |a_n|^2 r^{2|n|},}</math>

so that the norms increase to the value at ''r'' = 1, the norm of ''f''. A similar in the complement of the unit disk where the harmonic extension is given by

:<math>\displaystyle{F_R(e^{i\theta})=F(Re^{i\theta})=\sum a_n R^{-|n|}a_ne^{in\theta}.}</math>

In this case the norms increase from the value at ''R'' = ∞ to the norm of ''f'', the value at ''R'' = 1.

A similar result holds for a harmonic function ''f'' on a simply connected region with smooth boundary provided the L<sup>2</sup> norms are taken over the level curves in a tubular neighbourhood of the boundary.<ref>{{harvnb|Bell|1992|pp=19–20}}</ref> Using vector notation  '''v'''(''t'') = (''x''(''t''), ''y''(''t'')) to parametrize the boundary curve by arc length, the following classical formulas hold:

:<math>\displaystyle{\dot{\mathbf{v}}\cdot \dot{\mathbf{v}} =1,\,\,\, \ddot{\mathbf{v}}\cdot \dot{\mathbf{v}} = 0.}</math>

Thus the unit tangent vector '''t'''(''t'') at ''t'' and oriented normal vector  '''n'''(''t'') are given by

:<math>\displaystyle{\mathbf{t} =\dot{\mathbf{v}},\,\,\,\,\,\,\mathbf{n}=(-\dot{y},\dot{x}).}</math>

The constant relating the acceleration vector to the normal vector is the [[Curvature#Curvature of plane curves|curvature]] of the curve:

:<math>\displaystyle{\ddot{\mathbf{v}}=\kappa(t)\,\mathbf{n}(t),\,\,\,\,\, \kappa(t) =\ddot{\mathbf{v}}\cdot \mathbf{n}=\ddot{y}\dot {x} - \ddot{x}\dot{y}.}</math>

There are two further formulas of [[Jean Frédéric Frenet|Frenet]]:

:<math>\displaystyle{\dot{\mathbf{n}}=-\kappa \mathbf{t},\,\,\, \ddot{\mathbf{n}}= \dot{\kappa}\mathbf{n} -\kappa^2 \mathbf{t}.}</math>

A tubular neighbourhood of the boundary is given by

:<math>\displaystyle{\mathbf{v}_s(t)=\mathbf{v}(t)+s\mathbf{n}(t),}</math>

so that the level curves ∂Ω<sub>''s''</sub> with ''s'' constant bound domains Ω<sub>''s''</sub>. Moreover<ref>{{harvnb|Bell|1992|pp=19–22}}</ref>

:<math>\displaystyle{\dot{\mathbf{v}}_s(t) =(1-s\kappa)\mathbf{t},\,\,\,\,\partial_s |\dot{\mathbf{v}}_s|=-\kappa.}</math>

Hence differentiating the integral means with respect to ''s'', the derivative in the direction of the ''inward'' pointing normal, gives

:<math>\displaystyle{\partial_s \int_{\partial\Omega_s} |f|^2 =-\int_{\partial\Omega_s} (\partial_n f \overline{f} + f \overline{\partial_n f}) -  \int_{\partial\Omega_s} \kappa(1-\kappa s)^{-1}|f|^2 = -2\iint_{\Omega_s} |\nabla f|^2 -  \int_{\partial\Omega_s} \kappa(1-\kappa s)^{-1}|f|^2,}</math>

using [[Green's theorem]]. Thus for ''s'' small

:<math>\displaystyle{\partial_s \|f|_{\partial \Omega_s}\|^2 \le M  \|f|_{\partial \Omega_s}\|^2,}</math>

for some constant ''M'' independent of ''f''. This implies that

:<math>\displaystyle{\partial_s e^{-Ms}\|f|_{\partial \Omega_s}\|^2 \le 0,}</math>

so that, on integrating this inequality, the norms are bounded near the boundary:

:<math>\displaystyle{\|f|_{\partial \Omega_s}\| \le e^{Ms/2} \|f|_{\partial \Omega}\|.}</math>

This inequality shows that a function in the L<sup>2</sup> Hardy space H<sup>2</sup>(Ω) leads, via the Cauchy integral operator ''C'', to a holomorphic function on Ω satisfying the classical condition that the integral means

:<math>\displaystyle{ \int_{\partial\Omega_s} |f|^2}</math>

are bounded. Furthermore, the restrictions ''f''<sub>s</sub> of ''f'' to ∂Ω<sub>''s''</sub>, which can be naturally identified with ∂Ω, tend in L<sup>2</sup> to the original function in Hardy space.<ref>{{harvnb|Bell|1992|pp=16–21}}</ref> In fact H<sup>2</sup>(Ω) has been defined as the closure in L<sup>2</sup>(Ω) of rational functions (which can be taken to be polynomials if Ω is simply connected). Any rational function with poles  only in Ω<sup>''c''</sup> can be recovered inside Ω from its boundary value ''g'' by Cauchy's integral formula

:<math>\displaystyle{Cg(a)={1\over 2\pi i}\int_{\partial\Omega} {g(z)\over z-a}\, dz.}</math>

The estimates above show that the functions ''Cg''|<sub>∂Ω<sub>''s''</sub></sub> depend continuously on ''Cg''|<sub>∂Ω</sub>. Moreover, in this case the functions tend uniformly to the boundary value and hence also in L<sup>2</sup>, using the natural identification of the spaces L<sup>2</sup>(∂Ω<sub>''s''</sub>) with   L<sup>2</sup>(∂Ω). Since ''Ch'' can be defined for any L<sup>2</sup> function as a holomorphic function on Ω since ''h'' is integrable on ∂Ω. Since ''h'' is a limit in L<sup>2</sup> of rational functions ''g'', the same results hold for ''h'' and ''Ch'', with the same inequalities for the integral means. Equally well ''h'' is the limit in L<sup>2</sup>(∂Ω) of the functions ''Ch''|<sub>∂Ω<sub>''s''</sub></sub>.

The estimates above for the integral means near the boundary show that ''Cf'' lies in L<sup>2</sup>(Ω) and that its L<sup>2</sup> norm can be bounded in terms of that of ''f''. Since ''Cf'' is also holomorphic, it lies in the [[Bergman space]] A<sup>2</sup>(Ω) of Ω. Thus the Cauchy integral operator ''C'' defines a natural mapping from the Hardy space of the boundary into the Bergman space of the interior.<ref>{{harvnb|Bell|1992|p=22}}</ref>

The Hardy space H<sup>2</sup>(Ω) has a natural partner, namely the closure in L<sup>2</sup>(∂Ω) of boundary values of rational functions ''vanishing at'' ∞ with poles only in Ω. Denoting this subspace by H<sup>2</sup><sub>+</sub>(∂Ω) to distinguish it from the original Hardy space, which will also denoted by H<sup>2</sup><sub>−</sub>(∂Ω), the same reasoning as above can be applied. When applied to a function ''h'' in H<sup>2</sup><sub>+</sub>(∂Ω), the Cauchy integral operator defines a holomorphic function ''F'' in Ω<sup>''c''</sup> vanishing at ∞ such that near the boundary the restriction of ''F'' to the level curves, each identified with the boundary, tend in L<sup>2</sup> to ''h''. Unlike the case of the circle, H<sup>2</sup><sub>−</sub>(∂Ω) and H<sup>2</sup><sub>+</sub>(∂Ω) are not orthogonal spaces. By the Hartogs−Rosenthal theorem, their sum is dense in L<sup>2</sup>(∂Ω). As shown below, these are the ±i eigenspaces of the Hilbert transform on ∂Ω, so their sum is in fact direct and the whole of L<sup>2</sup>(∂Ω).

==Hilbert transform on a closed curve==
For a bounded simply connected domain Ω in the complex plane with smooth boundary ∂Ω, the theory of the Hilbert transform can be deduced by direct comparison with the Hilbert transform for the unit circle.<ref>See:
*{{harvnb|Gohberg|Krupnik|1992}}
*{{harvnb|Gakhov|1990}}</ref>

To define the Hilbert transform ''H''<sub>∂Ω</sub> on L<sup>2</sup>(∂Ω), take ∂Ω to be parametrized by arclength and thus a function ''z''(''t''). The Hilbert transform is defined to be the limit in the [[strong operator topology]] of the truncated operators ''H''<sub>∂Ω</sub><sup>ε</sup> defined by

:<math>\displaystyle{H_{\partial\Omega}^\varepsilon f(s) ={1\over \pi i} \int_{|s-t|\ge \varepsilon}\,\,\,\, {f(t)\over z(t)-z(s)}\, \dot{z}(t)\, dt.}</math>

To make the comparison it will be convenient to apply a scaling transformation in '''C''' so that the length of ∂Ω
is 2π. (This only changes the operators above by a fixed positive factor.) There is then a canonical unitary isomorphism of L<sup>2</sup>(∂Ω) onto L<sup>2</sup>('''T'''), so the two spaces can be identified. The truncated operators ''H''<sub>∂Ω</sub><sup>ε</sup>  can be compared  directly with the truncated Hilbert transform
''H''<sup>ε</sup>:

:<math>\displaystyle{H_{\partial\Omega}^\varepsilon  g(s) -H^\varepsilon g(s) = {1\over \pi i} \int_{ |t-s|\ge \varepsilon}  K(s,t)\cdot g(t)\, dt,}</math>

where

:<math>\displaystyle{K(u,v)={\dot{z}(t)\over z(t) - z(s)} - {ie^{it}\over e^{it}-e^{is}}= \partial_t \log\left( {z(t)-z(s)\over e^{it} - e^{is}}\right).}</math>

The kernel ''K'' is thus smooth on '''T''' × '''T''', so the difference above tends in the strong topology to the Hilbert–Schmidt operator defined by the kernel. It follows that the truncated operators ''H''<sub>∂Ω</sub><sup>ε</sup> are uniformly bounded in norm and have a limit in the strong operator topology denoted ''H''<sub>∂Ω</sub> and called the '''Hilbert transform''' on ∂Ω.

:<math>\displaystyle{H_{\partial\Omega}g = \lim_{\varepsilon\rightarrow 0} H^\varepsilon_{\partial\Omega}g.}</math>

Letting ε tend to 0 above yields

:<math>\displaystyle{H_{\partial\Omega}  g(s) -H g(s) = {1\over \pi i} \int_0^{2\pi}  K(s,t)\cdot g(t)\, dt.}</math>

Since ''H'' is skew-adjoint and ''H''<sub>∂Ω</sub> differs from ''H'' by a Hilbert–Schmidt operator with smooth kernel, it follows that ''H''<sub>∂Ω</sub> + ''H''<sub>∂Ω</sub>* is a Hilbert-Schmidt operator with smooth kernel. The kernel can also be computed explicitly using the truncated Hilbert transforms for ∂Ω:

:<math>\displaystyle{C(s,t) ={1\over \pi i}\left({\dot{z}(t)\over z(t) - z(s)} -{\overline{\dot{z}(s)}\over \overline{z(t)} -\overline{z(s)}} \right),}</math>

and it can be verified directly that this is a smooth function on '''T''' × '''T'''.<ref>{{harvnb|Bell|1992|pp=15–16}}</ref>

==Plemelj–Sokhotski relation==
{{See also|Sokhotski–Plemelj theorem}}
Let ''C''<sub>−</sub> and ''C''<sub>+</sub> be the Cauchy integral operators for Ω and Ω<sup>''c''</sup>. Then

:<math>\displaystyle{C_\pm \circ H = \pm i C_\pm.}</math>

Since the operators ''C''<sub>−</sub>, ''C''<sub>+</sub> and ''H'' are bounded, it suffices to check this on rational functions ''F'' with poles off ∂Ω and vanishing at ∞ by the Hartogs–Rosenthal theorem. The rational function can be written as a sum of functions ''F'' = ''F''<sub>−</sub> + ''F''<sub>+</sub> where ''F''<sub>−</sub> has poles only in   Ω<sup>''c''</sup> and ''F''<sub>+</sub> has poles only in Let ''f'', ''f''<sub>±</sub> be the restrictions of ''f'', ''f''<sub>±</sub> to ∂Ω. By [[Cauchy's integral formula]]

:<math>\displaystyle{C_\pm f_\pm = F_\pm,\,\,\, C_\pm f_\mp = 0.}</math>

On the other hand, it is straightforward to check that<ref>See:
*{{harvnb|Gakhov|1990}}
*{{harvnb|Bell|1992}}
*{{harvnb|Goluzin|1969}}
*{{harvnb|Gohberg|Krupnik|1992}}
</ref>

:<math>\displaystyle{Hf_\pm = \pm i f_\pm.}</math>

Indeed, by Cauchy's theorem, since ''F''<sub>−</sub> is holomorphic in  Ω,

:<math>\displaystyle{\int_{\partial\Omega,\,\,|z-w|\ge \varepsilon} {f_-(z)\over z-w}\, dz=-\int_{|z-w|=\varepsilon,\,\,z\in \Omega}  {F_-(z)\over z-w}\, dz.}</math>

As ε tends to 0, the latter integral tends to π''i'' ''f''<sub>−</sub>(''w'') by the [[residue calculus]]. A similar argument applies to ''f''<sub>+</sub>, taking the circular contour on the right inside Ω<sup>''c''</sup>.<ref>{{harvnb|Titchmarsh|1939}}</ref>

By continuity it follows that ''H'' acts as multiplication by ''i'' on H<sup>2</sup><sub>−</sub> and as multiplication by −''i'' on H<sup>2</sup><sub>+</sub>. Since these spaces are closed and their sum dense, it follows that

:<math>\displaystyle{H^2=-I.}</math>

Moreover,  H<sup>2</sup><sub>−</sub> and H<sup>2</sup><sub>+</sub> must be the ±''i'' eigenspaces of ''H'', so their sum is the whole of L<sup>2</sup>(∂Ω). The '''Plemelj–Sokhotski relation''' for ''f'' in  L<sup>2</sup>(∂Ω) is the relation

:<math> \displaystyle{-iHf= C_-f|_{\partial\Omega} - C_+f|_{\partial\Omega}.}</math>

It has been verified for ''f'' in the Hardy spaces H<sup>2</sup><sub>±</sub>(∂Ω), so is true also for their sum. The '''Cauchy idempotent''' ''E'' is defined by

:<math>\displaystyle{H=i(2E-I).}</math>

The range of ''E'' is thus  H<sup>2</sup><sub>−</sub>(∂Ω) and that of ''I'' − ''E'' is H<sup>2</sup><sub>+</sub>(∂Ω). From the above<ref>{{harvnb|Bell|1992}}</ref>

:<math>\displaystyle{Ef=C_-f|_{\partial\Omega},\,\,\,\, (I-E)f=C_+f|_{\partial\Omega}.}</math>

==Operators on a closed curve==
{{See also|Szegő kernel|Neumann–Poincaré operator}}
Two other operators defined on a closed curve ∂Ω can be expressed in terms of the Hilbert and Cauchy transforms ''H'' and ''E''.
<ref>See:
*{{harvnb|Kerzman|Stein|1978}}
*{{harvnb|Bell|1992}}
*{{harvnb|Shapiro|1992}}</ref>

The '''Szegő projection''' ''P'' is defined to be the orthogonal projection onto Hardy space H<sup>2</sup>(∂Ω). Since ''E'' is an idempotent with range H<sup>2</sup>(∂Ω), ''P'' is given by the '''Kerzman–Stein formula''':

:<math>\displaystyle{P=E(I+E-E^*)^{-1}.}</math>

Indeed, since ''E'' − ''E''* is skew-adjoint its spectrum is purely imaginary, so the operator ''I'' + ''E'' − ''E''* is invertible.<ref>{{harvnb|Shapiro|1992|p=65}}</ref> It is immediate that

:<math>\displaystyle{EP=P,\,\,\,PE=E.}</math>

Hence ''PE''* = ''P''. So

:<math>\displaystyle{P(I+E-E^*)= P + E -P = E.}</math>

Since the operator ''H'' + ''H''* is a [[Hilbert–Schmidt operator]] with smooth kernel, the same is true for ''E'' − ''E''*.<ref>{{harvnb|Bell|1992}}</ref>

Moreover, if ''J'' is the conjugate-linear operator of complex conjugation and ''U'' the operator of multiplication by the unit tangent vector:

:<math>\displaystyle{Jf(t)=\overline{f(t)},\,\,\, Uf(t)=\dot{z}(t)\cdot f(t).}</math>

then the formula for the truncated Hilbert transform on ∂Ω immediately yields the following identity for adjoints

:<math>\displaystyle{(H^\varepsilon)^*=JUH^\varepsilon U^*J.}</math>

Letting ε tend to 0, it follows that

:<math>\displaystyle{H^*=JUH U^*J}</math>

and hence

:<math>\displaystyle{I-E^*=JUEU^*J.}</math>

The comparison with the Hilbert transform for the circle shows that commutators of ''H'' and ''E'' with diffeomorphisms of the circle are Hilbert–Schmidt operators. Similar their commutators with the multiplication operator corresponding to a smooth function ''f'' on the circle is also Hilbert–Schmidt operators. Up to a constant the kernel of the commutator with ''H'' is given by the smooth function

:<math>\displaystyle{A(s,t)={f(s) - f(t)\over z(s) - z(t)}.}</math>

The '''Neumann–Poincaré operator''' ''T'' is defined on real functions ''f'' as

:<math>\displaystyle{Tf(w)={1\over 2\pi}\int_{\partial\Omega}\partial_n (\log|z-w|) f(z)={1\over 2}\Re (Hf)(w).}</math>

Writing ''h'' = ''f'' + ''ig'',<ref>{{harvnb|Shapiro|1992|pp=66–67}}</ref>

:<math>\displaystyle{2Th =\Re(Hf) + i\Re(Hg)={1\over 2} (Hf +JHf +iHg +iJHg)={1\over 2}(H+JHJ)h}</math>

so that

:<math>\displaystyle{T={1\over 4} (H +JHJ)={1\over 4}(H+UH^*U^*),}</math>

a Hilbert–Schmidt operator.

==Classical definition of Hardy space==
The classical definition of Hardy space is as the space of holomorphic functions ''F'' on Ω for which the functions ''F''<sub>''s''</sub> = ''F''|<sub>∂Ω<sub>''s''</sub></sub> have bounded norm in L<sup>2</sup>(∂Ω). An argument based on the [[Carathéodory kernel theorem]] shows that this condition is satisfied whenever there is a family of Jordan curves in Ω, eventually containing any compact subset in their interior, on which the integral means of ''F'' are bounded.<ref>{{harvnb|Duren|1970|p=168}}</ref>

To prove that the classical definition of Hardy space gives the space H<sup>2</sup>(∂Ω), take ''F'' as above. Some subsequence ''h''<sub>''n''</sub> = ''F''<sub>''s''<sub>''n''</sub></sub> converges weakly in L<sup>2</sup>(∂Ω) to ''h'' say. It follows that ''Ch'' = ''F'' in Ω. In fact, if ''C''<sub>n</sub> is the Cauchy integral operator corresponding to Ω<sub>''s''<sub>''n''</sub></sub>, then<ref>{{harvnb|Bell|1992|pp=17–18}}</ref>

:<math>\displaystyle{Ch(a)-F(a)= Ch(a) - C_nh_n(a)=C(h-h_n)(a) + [(C-C_n)h_n](a).}</math>

Since the first term on the right hand side is defined by pairing ''h'' − ''h''<sub>''n''</sub> with a fixed L<sup>2</sup> function, it tends to zero. If ''z''<sub>''n''</sub>(''t'') is the complex number corresponding to '''v'''<sub>''s''<sub>''n''</sub></sub>, then

:<math>\displaystyle{[(C-C_n)h_n](a)={1\over 2\pi i} \int h_n(t) \left({\dot{z}(t) \over z(t)-a} - {\dot{z}_n(t) \over z_n(t)-a}\right)\, dt.}</math>

This integral tends to zero because the L<sup>2</sup> norms of ''h''<sub>''n''</sub> are uniformly bounded while the bracketed expression in the integrand tends to 0 uniformly and hence in L<sup>2</sup>.

Thus ''F'' = ''Ch''. On the other hand, if ''E'' is the Cauchy idempotent with range H<sup>2</sup>(∂Ω), then ''C'' ∘ ''E'' = ''C''. Hence ''F'' =''Ch'' = ''C'' (''Eh''). As already shown ''F''<sub>''s''</sub> tends to ''Ch'' in L<sup>2</sup>(∂Ω). But a subsequence tends weakly to ''h''. Hence ''Ch'' = ''h'' and therefore the two definitions are equivalent.<ref>{{harvnb|Bell|1992|pp=19–20}}</ref>

==Generalizations==
The theory for multiply connected bounded domains with smooth boundary follows easily from the simply connected case.<ref>See:
*{{harvnb|Gakhov|1990}}
*{{harvnb|Mikhlin|Prössdorf|1986}}
</ref> There are analogues of the operators ''H'', ''E'' and ''P''. On a given component of the boundary, the singular contributions to ''H'' and ''E'' come from the singular integral on that boundary component, so the technical parts of the theory are direct consequences of the simply connected case.

Singular integral operators on spaces of [[Hölder continuous]] functions are discussed in {{harvtxt|Gakhov|1990}}. Their action on L<sup>''p''</sup> and Sobolev spaces is discussed in {{harvtxt|Mikhlin|Prössdorf|1986}}.

==Notes==
{{reflist|3}}

==References==
*{{citation|last=Bell|first= S. R.|title=The Cauchy transform, potential theory, and conformal mapping|series=Studies in Advanced Mathematics|publisher= CRC Press|year= 1992|isbn= 0-8493-8270-X}}
*{{citation|last=Bell|first= S. R.|title=The Cauchy transform, potential theory, and conformal mapping|series=Studies in Advanced Mathematics|publisher= CRC Press|year= 2016|edition=2nd|isbn=  9781498727211}}
*{{citation|title=Functions of one complex variable II|series=Graduate texts in mathematics|volume=159|first=John B.|last= Conway|publisher=Springer|year= 1995|isbn=0387944605|page=197}}
*{{citation|title=A course in operator theory|volume=21|series=[[Graduate Studies in Mathematics]]|first=John B.|last= Conway|publisher=[[American Mathematical Society]]|year= 2000|isbn=0821820656|pages=175–176}}
*{{citation|last=David|first= Guy|title= Opérateurs intégraux singuliers sur certaines courbes du plan complexe|journal=Ann. Sci. Éc. Norm. Supér. |volume= 17|year=1984|pages= 157–189|doi= 10.24033/asens.1469|doi-access= free}}
*{{citation|last=Duren|first= Peter L.|title= Theory of H<sup>p</sup> spaces|series= Pure and Applied Mathematics|volume=38 |publisher=Academic Press|year= 1970}}
*{{citation|last=Gakhov|first= F. D.|title=Boundary value problems. Reprint of the 1966 translation|publisher= Dover Publications|year=1990|isbn=0-486-66275-6}}
*{{citation|first=Theodore W.|last=Gamelin|title=Uniform algebras|edition=2nd|publisher=[[American Mathematical Society]]|year= 2005|
isbn=0821840495|pages=46–47}}
*{{citation|last=Garnett|first=J. B.|title=Bounded analytic functions|series=Graduate Texts in Mathematics|volume= 236|publisher= Springer|year=2007|isbn=978-0-387-33621-3}}
*{{citation|last1=Gohberg|first1=Israel|last2= Krupnik|first2= Naum|title= One-dimensional linear singular integral equations. I. Introduction|series= Operator Theory: Advances and Applications|volume= 53|publisher= Birkhäuser|year= 1992|isbn=3-7643-2584-4}}
*{{citation|last=Goluzin|first= G. M. |title=Geometric theory of functions of a complex variable|series= Translations of Mathematical Monographs|volume= 26 |publisher=American Mathematical Society|year=1969}}
* {{citation|first=Yitzhak|last= Katznelson|authorlink=Yitzhak Katznelson|year=2004|title=An Introduction to Harmonic Analysis|  publisher=Cambridge University Press|isbn= 978-0-521-54359-0}}
*{{citation|last1=Kerzman|first1= N.|last2= Stein|first2= E. M.|title=The Cauchy kernel, the Szegö kernel, and the Riemann mapping function|journal= Math. Ann.|volume= 236|year=1978|pages= 85–93|doi=10.1007/bf01420257|s2cid= 121336615}}
*{{citation|last=Muskhelishvili|first= N. I. |title=Singular integral equations. Boundary problems of function theory and their application to mathematical physics|publisher= Dover|year= 1992|isbn= 0-486-66893-2}}
*{{citation|last1=Mikhlin|first1= Solomon G.|authorlink=Solomon Mikhlin|last2= Prössdorf|first2=Siegfried|title=Singular integral operators|publisher= Springer-Verlag|year= 1986|isbn= 3-540-15967-3}}
*{{citation|last1=Pressley|first1= Andrew|last2= Segal|first2= Graeme|title=Loop groups|publisher=Oxford University Press|year=1986
|isbn=0-19-853535-X}}
*{{citation|last=Segal|first= Graeme|title= Unitary representations of some infinite-dimensional groups|journal= Comm. Math. Phys. |volume=80|year= 1981|issue= 3|pages= 301–342|doi=10.1007/bf01208274|bibcode= 1981CMaPh..80..301S|s2cid= 121367853|url= http://projecteuclid.org/euclid.cmp/1103919978}}
*{{citation|last=Shapiro|first=H. S.|authorlink=Harold S. Shapiro|title=The Schwarz function and its generalization to higher dimensions|series=University of Arkansas Lecture Notes in the Mathematical Sciences|volume= 9|publisher= Wiley-Interscience|year= 1992|isbn=0-471-57127-X}}
*{{Citation|author-link=E.C. Titchmarsh|first=E. C.|last=Titchmarsh|title= The Theory of Functions|publisher=Oxford University Press| year=1939|edition=2nd|
isbn=0-19-853349-7}}
* {{citation|first=Alberto|last= Torchinsky|year=2004|title=Real-Variable Methods in Harmonic Analysis| publisher=Dover|isbn= 0-486-43508-3}}

[[Category:Operator theory]]
[[Category:Harmonic analysis]]
[[Category:Complex analysis]]