Abstract
We study membership in Lipschitz classes for a class of α-harmonic functions in the open unit disc in the complex plane. From earlier work by Olofsson and Wittsten we know that such an α-harmonic function u is the α-harmonic Poisson integral of its boundary value function f on the unit circle . We determine when the Poisson integral belongs to a Lipschitz class for the unit disc.
AMS SUBJECT CLASSIFICATIONS:
1. Introduction
Let be the open unit disc in the complex plane, let be the unit circle, and denote by the standard complex partial derivatives, where z=x + iy is the complex coordinate. A chief concern of this paper is the homogeneous second order partial differential equation (1) (1) where (2) (2) is a certain weighted Laplacian with weight parameter . The restriction on the weight parameter is needed for Equation (Equation1(1) (1) ) to have a sufficiently rich supply of solutions with well-behaved boundary behaviour.
A solution u of Equation (Equation1(1) (1) ) is called an α-harmonic function. Such a function u is smooth in but can be rather wild near the boundary . Notice that is one-quarter of the usual Laplacian and that a 0-harmonic function is a harmonic function in in the usual sense. Notice also that the differential operator has a certain singular or degenerate behaviour on the unit circle when . The study of weighted Laplacians of the above form (Equation2(2) (2) ) goes back to Garabedian [Citation1] and has attracted quite some interest recently.
Of particular interest is the Dirichlet boundary value problem (3) (3) for α-harmonic functions. Here is a distribution on and the boundary condition in (Equation3(3) (3) ) is interpreted in a distributional sense that in , where (4) (4) for . From Olofsson and Wittsten [Citation2] it is known that a twice continuously differentiable function in solves the Dirichlet problem (Equation3(3) (3) ) if and only if it has the form of a Poisson integral where (5) (5) is the α-harmonic Poisson kernel, for in accordance with (Equation4(4) (4) ) and is convolution in . For an integrable function on the above Poisson integral has the form (6) (6) where the function is as in (Equation5(5) (5) ). Notice that is the usual Poisson integral representation of a harmonic function in . The Poisson integral representation leads to improved convergence to boundary values (in norm as well as non-tangentially) of solutions of (Equation3(3) (3) ) provided the boundary data has some appropriate regularity, see [Citation2–4] for results.
Let us recall the notion of Lipschitz continuity. For a bounded subset E of the complex plane and parameter we denote by the set of complex-valued functions f on E such that for some finite constant . The class is commonly referred to as the Lipschitz (or Hölder) class for the set E of order β. The space is often equipped with the norm and is then a commutative Banach algebra under pointwise multiplication of functions (see for instance Sherbert [Citation5]). It is well-known that every function f in extends uniquely to a continuous function on the closure of E in such a way that . It is also well-known that if . The family of spaces forms a much natural scale to measure smoothness of continuous functions.
A main purpose of this paper is to study membership in Lipschitz classes for α-harmonic functions. Observe that implies that , where A fundamental question is to determine when the Poisson integral of a function belongs to the space . Notice that this latter question is an extension problem in the sense that we ask when the function extends to an α-harmonic function u in the class . We shall here solve this problem in full generality. The solution we propose depends on the individual parameters and as well as on spectral properties of the function .
Our results contrast the situation in the category of continuous functions where one has that the Poisson integral extends to a continuous function on the closed disc if is a continuous function on and (see Olofsson and Wittsten [Citation2, Section 5]). The study of Lipschitz conditions goes back to classical work of Hardy and Littlewood [Citation6, Section 5] in the case of analytic functions.
An integrable function on is identified with the distribution on defined by where is the distributional pairing and is the space of indefinitely differentiable test functions on . The Fourier coefficients of a function are defined by for integers , and similarly for distributions using the distributional pairing (see for instance [Citation2] for details). The conjugate function of a distribution is the distribution defined by the condition that where is the n-th Fourier coefficient of f, for and . A much influential introduction to Fourier analysis is Katznelson [Citation7].
Let us first recall the situation in the classical case of usual harmonic functions in (). If and , then . Furthermore, a harmonic function u in belongs to the class if and only if it has the form for some function such that (see Theorem 6.7). These results are well-known and follow quite easily from the work of Hardy and Littlewood [Citation6, Section 5] on analytic functions mentioned above. Recall here that the conjugate function is a bounded linear operator on the Lipschitz space for (see Zygmund [Citation8, Theorem III.13.29]). Results of this latter type often go under the name of Privalov's theorem.
Let us next consider the case when and are such that . For such parameters we show that belongs to if (see Theorem 3.2). This generalizes a classical result from the previous paragraph as well as a recent result of Li and Wang [Citation9] for to the bigger parameter range considered; a less precise result can be found inChen [Citation10]. In fact, as we shall see later on, the inequality defines the exact parameter range for the implication to hold true.
Lipschitz continuity of a continuously differentiable function in is inferred from a gradient growth estimate of the form (7) (7) (see Theorem 3.1). This observation goes back to classical work of Hardy and Littlewood [Citation6] and has been elaborated on by Gehring and Martio [Citation11] and others. In the context of the previous paragraph we establish (Equation7(7) (7) ) by careful estimations using the Poisson integral representation (see Theorems 2.5 and 2.10). Notably, we show that (8) (8) for with provided (see Theorem 2.13).
It is an important observation that the combination of partial derivatives yields an estimate (Equation8(8) (8) ) which is valid without any extra restriction on β more than what is previously stated. In fact, this more general validity of (Equation8(8) (8) ) compared to (Equation7(7) (7) ) is quite natural from the point of view of convolution structure in (Equation6(6) (6) ) and an interpretation of the differential operator as angular derivative. The growth estimate (Equation8(8) (8) ) is established for with a fairly economical constant using a precise result on the -means of the function from Olofsson and Wittsten [Citation2, Section 2] later refined in Olofsson [Citation3] (see Theorem 2.4). We mention that inequality (Equation8(8) (8) ) generalizes a recent result by Chen [Citation10, Lemma 3.2].
Let us next consider the case when and are such that . For such parameters we show that if u is α-harmonic and , then u is analytic in (see Theorem 5.4). A closely related result is that, in this parameter range, the Poisson integral belongs to if and only if and for n<0 (see Theorem 5.5). This latter result puts earlier examples of Li and Wang [Citation9] and Chen [Citation10] in a precise context.
The results in the previous paragraph are proved using properties of the homogeneous expansion of α-harmonic functions. Let and set for positive integers , where Γ is the standard Gamma function. We now introduce the functions for . From Olofsson and Wittsten [Citation2, Section 5] we know that the Poisson integral has the series expansion (9) (9) for .
We derive a system of two first-order partial differential equations in satisfied by the function (see Proposition 4.2). This system of partial differential equations is shown to characterize the function in the punctured disc up to an additive term of the form (see Theorem 4.3). As an application of this development we show that for (see Theorem 4.7). This latter formula refines an earlier result from Olofsson and Wittsten [Citation2, Theorem 1.9]. As a consequence we have that if and only if (see Lemma 5.1). Combined with some homogeneity type of arguments this leads to the results about Lipschitz continuity of α-harmonic functions in the case of big smoothness discussed above.
Let us now consider the case when and are such that . The analysis of this case is the main technical contribution of this paper. We shall need the Fourier multiplier defined by for , where is the n-th Fourier coefficient of f. Denote also by the space of essentially bounded measurable functions on normed by the essential supremum In this parameter range, we show that the function belongs to the class if and only if is such that (see Theorems 6.3 and 6.5).
The constraint on a Lipschitz function () that belongs to is quite demanding. An interesting example is provided by the Weierstrass function which has the properties that and (see discussion following formula (Equation49(49) (49) )).
It is an important observation that the function above is a constant multiple of a hypergeometric function. Assuming that belongs to the class , we prove necessity of the condition by a differentiation type argument using asymptotic properties of hypergeometric functions (see Theorem 6.3).
Differentiating (Equation9(9) (9) ) using the above mentioned partial differential equations for the 's we show that for , where (see Lemma 6.4). This latter formula suggests an obvious estimation of . Combined with (Equation8(8) (8) ) this yields that satisfies a growth estimate of the form (Equation7(7) (7) ) if is such that (see Theorem 6.5).
In conclusion we have shown that an inequality of the form (Equation7(7) (7) ) is also necessary for Lipschitz continuity of an α-harmonic function u. We point out that such an implication requires interior rigidity of the function u and is not true in general (see Proposition 3.5).
For further results on Lipschitz continuity of analytic functions in we mention Dyakonov [Citation12]. Recent papers dealing with aspects of α-harmonic functions not quoted above are [Citation4, Citation13–17]. We mention also a related interesting paper by Borichev and Hedenmalm [Citation18].
The author thanks Kevin Klein and Jens Wittsten for useful discussions.
2. Bounds for partial derivatives
In this section, we prove some growth bounds for partial derivatives of α-harmonic Poisson integrals. Applications of these bounds appear in later sections.
We first consider the -derivative.
Lemma 2.1
Let . Let and set . Then for with and .
Proof.
The interpretation of the Poisson integral (Equation6(6) (6) ) as solution of the Dirichlet problem (Equation3(3) (3) ) leads to the property that for . Let and notice that for . Differentiating we have that for . The choice gives the conclusion of the lemma.
We now calculate .
Lemma 2.2
Let and let be as in (Equation5(5) (5) ). Then for .
Proof.
From formula (Equation5(5) (5) ) we have that for . Differentiating we have that for . A simplification of terms now leads to the conclusion of the lemma.
From Lemma 2.2 we have that (10) (10) for .
The following lemma is well-known but included here for the sake of convenience.
Lemma 2.3
Let . Then for .
Proof
Sketch of proof
First show that for x>0. A homogeneity argument then leads to the conclusion of the lemma.
The interest in Lemma 2.3 comes from the fact that the power function operates on metrics: If d is a metric on a set X, then so is the function for .
Recall the standard Gamma function For the sake of easy reference we recall the following result from [Citation2, Section 2].
Theorem 2.4
Let and set . Then for . Furthermore,
For a detailed discussion of Theorem 2.4 and its relation to hypergeometry we refer to [Citation3].
We next estimate the -derivative of the Poisson integral. We denote by the space of continuous functions on .
Theorem 2.5
Let and be such that . Let be such that where C is a finite positive constant. Set . Then for where is as in Theorem 2.4.
Proof.
Let with and . Recall Lemma 2.1. By the triangle inequality we have that By the assumption on f we have that where the last equality follows by a change of variables. We next invoke Lemma 2.2 to the extent that (11) (11) for with .
We shall now estimate the factor in the integrand in (Equation11(11) (11) ). By Lemma 2.3 and the triangle inequality we have that By a geometric consideration we see that for 0<r<1 and .
We now return to the estimation of . By (Equation11(11) (11) ) and the result of the previous paragraph we have that for , where . An application of Theorem 2.4 now yields the conclusion of the theorem.
We now turn our attention to the ∂-derivative of the Poisson integral.
Lemma 2.6
Let . Let and set . Then for .
Proof.
The proof of the lemma is analogous to that of Lemma 2.1 and is therefore omitted.
Recall the partial fraction decomposition (12) (12) of the classical Poisson kernel.
We now calculate .
Lemma 2.7
Let and let be as in (Equation5(5) (5) ). Then for .
Proof.
Recall formulas (Equation5(5) (5) ) and (Equation12(12) (12) ). Differentiating we have that for , where the last equality is straightforward to check.
It is evident that the quotient belongs to for every with .
Lemma 2.8
We have that for every .
Proof.
Let . The line segment between and the point 1 is parametrized by We have that Observe that is the disjoint union of line segments as above. Thus for . This completes the proof of the lemma.
We remark in passing that if and only if .
We now estimate .
Proposition 2.9
Let and let be as in (Equation5(5) (5) ). Then for . The constant is best possible.
Proof.
From Lemma 2.7 we have that for , where Thus (13) (13) for .
It remains to calculate the supremum . By the triangle inequality we have that . From Lemma 2.8 we have that for every . Maximizing over we conclude that This completes the proof of the proposition.
We now return to the α-harmonic Poisson integral.
Theorem 2.10
Let and be such that . Let be such that where C is a finite positive constant. Set . Then where is as in Theorem 2.4.
Proof.
The proof parallels that of Theorem 2.5. Let with and . Recall Lemma 2.6. By the triangle inequality we have that By the assumption on f we have that where the last equality follows by a change of variables. We next invoke Proposition 2.9 to the extent that (14) (14) for with . Formula (Equation14(14) (14) ) should be compared with formula (Equation11(11) (11) ) in the proof of Theorem 2.5. We now proceed as in the proof of Theorem 2.5 to arrive at the conclusion of the theorem.
We next calculate the angular derivative of .
Theorem 2.11
Let and let be as in (Equation5(5) (5) ). Then for .
Proof.
From Lemma 2.7 we have that for . Using also Lemma 2.2 we have that (15) (15) for , where the last equality is straightforward to check.
We now turn our attention to the leftmost factor in (Equation15(15) (15) ). By a partial fraction decomposition we have that for . In view of the partial fraction formula (Equation12(12) (12) ) for the classical Poisson kernel we thus have that (16) (16) for .
We return to the leftmost factor in (Equation15(15) (15) ). We have that (17) (17) for , where the last equality again follows by (Equation12(12) (12) ).
Recall formula (Equation15(15) (15) ). From (Equation16(16) (16) ) and (Equation17(17) (17) ) we have that for . We next rewrite the leftmost factor so that for . This completes the proof of the theorem.
Theorem 2.11 leads to an important inequality for the angular derivative of the Poisson kernel.
Corollary 2.12
Let and let be as in (Equation5(5) (5) ). Then for .
Proof.
Recall Theorem 2.11. The result follows by the triangle inequality.
The point of Corollary 2.12 is that the angular derivative of behaves somewhat better than or (see formula (Equation10(10) (10) ) or Proposition 2.9).
Theorem 2.13
Let and . Let be such that where C is a finite positive constant. Set . Then where is as in Theorem 2.4.
Proof.
The proof parallels that of Theorem 2.5. Let with and . Let be the angular derivative. By Lemmas 2.1 and 2.6 we have that By the triangle inequality we have that By the assumption on f we have that where the last equality follows by a change of variables. We next invoke Corollary 2.12 to the extent that (18) (18) for with .
As in the proof of Theorem 2.5 we have that for 0<r<1 and . From (Equation18(18) (18) ) we have that An application of Theorem 2.4 now yields the conclusion of the theorem.
It is important to notice that Theorem 2.13 carries no additional assumption on the Lipschitz parameter (compare Theorems 2.5 and 2.10). The case of Theorem 2.13 is closely related to Chen [Citation10, Lemma 3.2].
3. Lipschitz continuity from gradient growth
It is well-known that Lipschitz continuity of a function is inferred from a gradient growth estimate of the form (Equation7(7) (7) ). This section pertains to a discussion of this aspect of the theory.
Let . We shall consider the quantity where the infimum is taken over all piecewice continuously differentiable curves γ from to in and the integration is with respect to arclength. It is straightforward to check that the function is a metric on .
It is evident that is usual Euclidean distance. Furthermore, conformal invariance leads to the formula for , where (see for instance [Citation19, Theorem 2.2]). The function is known as the hyperbolic metric for , whereas the function is known as the pseudo hyperbolic metric for .
We shall make use of the result that (19) (19) for , where is a finite positive constant. A careful analysis shows that the constant in (Equation19(19) (19) ) can be chosen of the form , where C is an absolute constant. For proofs of (Equation19(19) (19) ) we refer to the original papers Gehring and Martio [Citation11, Section 2] or Hardy and Littlewood [Citation6, Section 5].
Theorem 3.1
Let be such that (20) (20) for some and C>0. Then and for where is as in (Equation19(19) (19) ).
Proof.
Let . By the fundamental theorem of calculus we have that where γ is a piecewise -curve from to (see for instance Gamelin [Citation20, Section III.2]). By the triangle inequality we have that where the last inequality follows by (Equation20(20) (20) ). Passing to an infimum over curves γ we have that where the last equality follows by (Equation19(19) (19) ). This completes the proof of the theorem.
The interest in Theorem 3.1 is when the function u has some interior rigidity.
Theorem 3.2
Let and be such that . Let and set . Then .
Proof.
Recall Theorems 2.5 and 2.10. The result now follows from Theorem 3.1.
The case of analytic functions merits particular mention.
Theorem 3.3
Let and . Let be such that for n<0. Then belongs to .
Proof.
Recall Theorem 2.13. The spectral assumption means that in . The result now follows from Theorem 3.1.
Remark 3.4
We point out that Theorem 3.3 can be regarded as a special case of Theorem 3.2. Indeed, if u is as in Theorem 3.3, then for all and an application of Theorem 3.2 with yields that .
Theorem 3.3 goes back to Hardy and Littlewood [Citation6, Section 5].
Let us revisit a classic function (21) (21) from calculus. Notice that . Repeated differentiation shows that for , where and for . The function is a rational function. From the vanishing of at infinity we see that there is an estimate of the form (22) (22) for , where is a finite positive constant.
We next construct an example where Theorem 3.1 is of less interest.
Proposition 3.5
Let and let w be a positive function on . Then there exists a function such that and
Proof.
We shall make use of the functions for , where f is as in (Equation21(21) (21) ). Notice that for . An application of Lemma 2.3 shows that (23) (23) for . By differentiation we have that (24) (24) In particular, we have that for and . Moreover, an induction using (Equation22(22) (22) ) and (Equation24(24) (24) ) shows that (25) (25) for and , where is a finite positive constant.
Let us now construct a function u with the desired properties. Let be a sequence of points in the interval with accumulation points and 1 only. Let be a sequence of positive numbers such that as . We consider a function u of the form (26) (26) where the 's are as above. From the bound (Equation23(23) (23) ) we see that the Lipschitz norms for are uniformly bounded. By completeness of the space this yields that .
We next show that . Let be a compact set and an integer. Let 0<r<1 be such that . Since the sequence accumulates only at the points and 1 there exists such that for k>N. From the bound (Equation25(25) (25) ) we have that for k>N. This proves that . Varying the compact K and the integer , we see that the series (Equation26(26) (26) ) defining u is convergent in .
By differentiation, we see that . The final conclusion of the proposition is satisfied provided rapidly enough.
We mention that growth estimates of the form (Equation25(25) (25) ) appear in the definition of symbol classes for pseudo differential operators (see Hörmander [Citation21, Chapter 18]).
4. The homogeneous expansion
In this section, we discuss some generalities related to the homogeneous expansion of α-harmonic functions. Applications of this material appear in later sections.
Let us recall the Beta function defined by for a,b>0. It is well-known that where Γ is the Gamma function (see [Citation22, Section 1.1]).
Let and set for , where B is the Beta function. Clearly, as follows by a passage to the limit.
We now introduce the functions (27) (27) for . From [Citation2, Section 1] we know that a complex-valued function u in is α-harmonic if and only if it has a series expansion of the form (28) (28) for some sequence of complex numbers such that (29) (29) Furthermore, the series expansion (Equation28(28) (28) ) is convergent in provided (Equation29(29) (29) ) holds.
We proceed to study the functions in some more detail. The integral formula (30) (30) extends the function to an analytic function in the slit plane . Here the power is defined in the usual way using a logarithm which is real on the positive real axis.
Lemma 4.1
Let . Let be as in (Equation30(30) (30) ) for some . Then for .
Proof.
Differentiating under the integral sign in (Equation30(30) (30) ) we have that for . An integration by parts now gives that for , where the last equality again follows by (Equation30(30) (30) ). This yields the conclusion of the lemma.
Lemma 4.1 leads to differential equations for the 's.
Proposition 4.2
Let . Let be as in (Equation27(27) (27) ) for some . Then for .
Proof.
Recall formula (Equation27(27) (27) ). Differentiating we have that for . We now use Lemma 4.1 to conclude that for . This yields the first equation in the proposition.
Recall formula (Equation27(27) (27) ). Differentiating we have that for . We now use Lemma 4.1 to conclude that for , where the last equality again follows by (Equation27(27) (27) ). This yields the second equation in the proposition.
Let us digress on the differential equations from Proposition 4.2.
Theorem 4.3
Let and . Then a function satisfies the differential equations (31) (31) for if and only if it has the form (32) (32) for some .
Proof.
Assume first that satisfies (Equation31(31) (31) ). In view of Proposition 4.2, the first equation in (Equation31(31) (31) ) implies that where f is analytic in . Now the second equation in (Equation31(31) (31) ) and Proposition 4.2 implies that Solving for f we see that for some constant .
Conversely, using Proposition 4.2 it is straightforward to check that every function u of the form (Equation32(32) (32) ) satisfies (Equation31(31) (31) ).
Following earlier practice from [Citation2–4], a function u in is called homogeneous of order with respect to rotations if it has the property that for . Notice that the function is homogeneous of order .
Lemma 4.4
Let be homogeneous of order . Then for .
Proof.
Assume first that . Since is homogeneous of order it has the form for some . Differentiating we see that for . By these two differentiation formulas we see that for . This yields the conclusion of the lemma for .
If is homogeneous of order , then its complex conjugate has homogeneity . The case n<0 thus follows by complex conjugation.
Let and consider its homogeneous parts defined by (33) (33) for . Notice that the function is homogeneous of order n with respect to rotations and that for . From a classical result of Fejér we have that (34) (34) with convergence in the space , where the 's are as in (Equation33(33) (33) ) (see Katznelson [Citation7, Section I.2]). We shall refer to the expansion (Equation33(33) (33) )–(Equation34(34) (34) ) as the homogeneous expansion of u.
The differential equations (Equation31(31) (31) ) in Theorem 4.3 exhibit certain structure owing to homogeneity.
Theorem 4.5
Let . Then a function satisfies the differential equation (35) (35) for if and only if it is homogeneous of order n with respect to rotations.
Proof.
By Lemma 4.4 we know that satisfies (Equation35(35) (35) ) if u is homogeneous of order n with respect to rotations. Assume next that is a general solution of (Equation35(35) (35) ). We consider the k-th homogeneous part of u defined as in (Equation33(33) (33) ). Let be the angular derivative. Differentiating under the integral in (Equation33(33) (33) ) we have that for . Since u satisfies (Equation35(35) (35) ) we conclude that for , where the last equality follows by (Equation33(33) (33) ). Also, since is homogeneous of order k, we have by Lemma 4.4 that . We conclude that if . By (Equation34(34) (34) ) we have that is homogeneous of order n.
It is straightforward to check that a function is homogeneous of order m=−n<0 with respect to rotations if and only if it has the form (36) (36) for some function .
Let us return to Equation (Equation31(31) (31) ).
Theorem 4.6
Let and . Then a function satisfies Equation (Equation31(31) (31) ) if and only if it has the form (Equation36(36) (36) ) for some function such that (37) (37) for 0<x<1.
Proof.
A Gauss elimination shows that (Equation31(31) (31) ) is equivalently stated saying that (38) (38) for . By Theorem 4.5, the latter equation in (Equation38(38) (38) ) is equivalent to having the form (Equation36(36) (36) ) for some function . For such u, a differentiation shows that for . In view of this latter differentiation formula, we see that the first equation in (Equation38(38) (38) ) is equivalent to (Equation37(37) (37) ). This yields the conclusion of the theorem.
Notice that the choice in (Equation37(37) (37) ) gives by (Equation36(36) (36) ) rise to the α-harmonic function .
Theorem 4.7
Let . Let be as in (Equation27(27) (27) ) for some . Then for .
Proof.
We shall exhibit a solution of Equation (Equation31(31) (31) ). An interesting solution f of (Equation37(37) (37) ) is the function for 0<x<1. The change of variables leads to the formula (39) (39) for 0<x<1. We observe also that where the last equality follows by a standard formula for the Beta function.
We now consider the function u given by (Equation36(36) (36) ) with f as in (Equation39(39) (39) ). Since f satisfies (Equation37(37) (37) ), we have from Theorem 4.6 that u satisfies Equation (Equation31(31) (31) ). An application of Theorem 4.3 shows that the function u has the form for some . Observe that Thus for . Solving for , we arrive at the conclusion of the theorem.
Theorem 4.7 displays interesting structure in the function . Notice that the factor is a polynomial.
We mention that an earlier version of Theorem 4.7 goes back to Olofsson and Wittsten [Citation2, Theorem 1.9]. The case n=2 of Theorem 4.7 has appeared recently in Chen [Citation10, Example 2.1].
5. The case of big smoothness
In this section, we discuss Lipschitz continuity of an α-harmonic function u in the case when and . We refer to this parameter range as the case of big smoothness.
Recall that the space is a commutative Banach algebra under pointwise multiplication of functions. We shall use that the algebra is inverse closed in the sense that if and for (see Sherbert [Citation5, Proposition 1.7]).
Lemma 5.1
Let and . Let be an integer. Then if and only if .
Proof.
From Theorem 4.7 we have that (40) (40) for . We set and (41) (41) Observe that for .
Assume now that . By the algebra property of we have from (Equation40(40) (40) ) that . Since is inverse closed we can divide by p to conclude that . This yields that .
Assume next that . From Proposition 4.2 we have that the function satisfies an estimate of the form (Equation20(20) (20) ). An application of Theorem 3.1 yields that .
Remark 5.2
Let p be as in (Equation41(41) (41) ). The fact that p is a polynomial makes division by p a simpler operation not dependent on the full strength of commutative Banach algebras (see end of Section 4).
Remark 5.3
As a byproduct from the proof of Lemma 5.1 we have the norm bound that for if and .
We now return to the study of α-harmonic functions.
Theorem 5.4
Let and be such that . Let u be an α-harmonic function such that . Then u is analytic in .
Proof.
We consider the series expansion (Equation28(28) (28) ) for u. We shall show that for n<0. Consider the n-th homogeneous part of u for . The assumption that gives that for . Indeed, since for by assumption, we have by the triangle inequality that for .
We shall next evaluate the constraint . Let . By (Equation28(28) (28) ) and homogeneity we have that for , where the last equality follows by an orthogonality property of exponential monomials. Observe that the change of order of integration and summation is permitted by uniform convergence. Thus for .
Since by assumption, we have from Lemma 5.1 that for . Now, since belongs to by the results of the previous two paragraphs, we conclude that for n<0. From (Equation28(28) (28) ) we then have that is analytic in .
We point out that Theorem 5.4 generalizes an earlier result by Olofsson and Wittsten [Citation2, Theorem 1.11] to the scale of spaces .
A variant of Theorem 5.4 goes as follows.
Theorem 5.5
Let and be such that . Let . Then if and only if for n<0.
Proof.
Assume first that belongs to . By Theorem 5.4 we conclude that u is analytic in . By passage to boundary values we see that for n<0.
Conversely, assume that for n<0. Then is analytic in and Theorem 3.3 yields that .
6. The case of critical smoothness
In this section, we discuss Lipschitz continuity of an α-harmonic function u in the case when and . We refer to this parameter range as the case of critical smoothness.
We shall need some more properties of the functions defined in (Equation30(30) (30) ) for and . By differentiation we see that (42) (42) for .
Lemma 6.1
Let . Let be as in (Equation30(30) (30) ) for some . Then is totally monotone on the interval in the sense that for x<1 and .
Proof.
The conclusion of the lemma is evident from formula (Equation42(42) (42) ).
The study of the functions belongs to the subject of hypergeometry. We observe that (43) (43) where is a standard hypergeometric function. In fact, formula (Equation42(42) (42) ) is a specialization of the Euler integral formula valid when c>b>0 (see [Citation22, Theorem 2.2.1]).
The study of asymptotic behaviour of hypergeometric functions is a classical topic which goes back to Gauss. We shall use the result that (44) (44) when c<a+b (see [Citation22, Theorem 2.1.3]). The asymptotic result (Equation44(44) (44) ) belongs to the realm of the famous Gauss summation formula.
Lemma 6.2
Let . Let be as in (Equation30(30) (30) ) for some . Then
Proof.
We consider the derivative . From formula (Equation43(43) (43) ) above we have that By (Equation44(44) (44) ) we have that The conclusion of the lemma now follows from L'Hospital's rule.
We shall consider the Fourier multiplier defined by for . From Stirling's formula we have that (45) (45) asymptotically in the sense that the quotient tends to 1 as (see [Citation22, Section 1.4]).
Recall that the space is the dual of and has as such a natural weak topology. A sequence in converges to a function f in in the weak topology of if and only if and for every . It follows by a well-known diagonal procedure that every bounded sequence in has a weak convergent subsequence. An interesting discussion of Fourier coefficients for linear functionals is found in Katznelson [Citation7, Section I.7]; see [Citation2, Section 4] for information on a related notion of relative completeness.
Theorem 6.3
Let and set . Let be such that . Then .
Proof.
Since and we have from Theorem 3.2 that belongs to . Set . By assumption we have that . Since also v=0 on by construction, there exists a finite positive constant C such that (46) (46) for .
We consider now the functions for 0<r<1. Clearly for 0<r<1. Furthermore, from (Equation46(46) (46) ) we have that for 0<r<1. The functions thus form a bounded sequence in .
We now turn our attention to Fourier coefficients. From (Equation9(9) (9) ) we have that (47) (47) for . This latter series expansion (Equation47(47) (47) ) makes evident that for . Let . Again from (Equation47(47) (47) ) we have that for 0<r<1. By Lemma 6.2 we have that for .
From the result of the previous paragraph we know that the limit exists for every . Since the sequence is bounded in , it follows that the limit exists in the weak topology of . We conclude that belongs to and , where C is as in (Equation46(46) (46) ).
The Fourier multiplier is naturally derived from the binomial series: (48) (48) Let and consider the anti-analytic function From the binomial series (Equation48(48) (48) ) we have that for . From here we see that where in is the distributional boundary limit of the function h, is as in (Equation4(4) (4) ) and is convolution in .
The Fourier coefficients of a Lipschitz function decay as as (see Zygmund [Citation8, Theorem II.4.7]). An interesting example is provided by the Weierstrass function (49) (49) where b>1 is an integer. It is known that for (see Zygmund [Citation8, Theorem II.4.9]).
Let and set . Using (Equation45(45) (45) ) and (Equation49(49) (49) ) it is straightforward to check that the Fourier coefficients of the distribution are not vanishing at infinity. By Riemann–Lebesgue lemma we conclude that for such parameters. The conclusion in Theorem 6.3 is thus quite demanding of a Lipschitz function .
Lemma 6.4
Let . Let and set . Then for .
Proof.
Recall the homogeneous expansion of the Poisson integral (Equation9(9) (9) ). Differentiating we have that for . We now use the first of the differential equations in Proposition 4.2 to conclude that for , where the last equality is straightforward to check. Since for , this yields the conclusion of the lemma.
We now return to the study of Lipschitz classes.
Theorem 6.5
Let and set . Let be such that . Then the function belongs to .
Proof.
We shall apply Theorem 3.1. Recall Lemma 6.4. A standard estimation of gives that for . Recall Theorem 2.13. An application of Theorem 3.1 now yields the conclusion of the theorem.
We denote by the distributional derivative of . Recall that for .
Lemma 6.4 is accompanied with the following lemma.
Lemma 6.6
Let . Let and set . Then for .
Proof.
Recall the homogeneous expansion of the Poisson integral (Equation9(9) (9) ). Differentiating we have that for . We now use the second of the differential equations in Proposition 4.2 to conclude that for , where the last equality is straightforward to check. In view of the series expansion (Equation9(9) (9) ), this yields the conclusion of the lemma.
Let . The conjugate function of f is the distribution defined by the condition that (50) (50) where for and . From (Equation50(50) (50) ) we have that the α-harmonic functions and are such that the function g=u+iv is analytic in and .
The following result is well-known but included here for the sake of completeness.
Theorem 6.7
Let and set . Then if and only if .
Proof.
Assume first that . We set . Since we have that for , where C is a finite positive constant. This latter Lipschitz condition gives that the partial derivatives of u are uniformly bounded in . Since v is a harmonic conjugate for u we conclude that v also has uniformly bounded partial derivatives in . By Theorem 3.1 we have that . Passing to the boundary limit of v we see that .
Assume next that . Now the function is analytic in and . By Theorem 3.3 we have that . Similarly, the function is anti-analytic in and . An application of Theorem 3.3 gives that . We now conclude that .
The Fourier multiplier is naturally expressed in terms of the conjugate function. It is straightforward to check that (51) (51) where the prime indicates distributional derivative.
A classical result is that the kernel of the distributional derivative is the space of constant functions (see Hörmander [Citation23, Theorem 3.1.4]). It is also well-known that the distributional derivative maps the Lipschitz space onto .
Proposition 6.8
Let . Then if and only if .
Proof.
Recall formula (Equation51(51) (51) ). If , then and a differentiation gives that . Conversely, if , then which yields that .
Disclosure statement
No potential conflict of interest was reported by the author.
References
- Garabedian PR. A partial differential equation arising in conformal mapping. Pacific J Math. 1951;1:485–524. doi: 10.2140/pjm.1951.1.485
- Olofsson A, Wittsten J. Poisson integrals for standard weighted Laplacians in the unit disc. J Math Soc Jpn. 2013;65:447–486. doi: 10.2969/jmsj/06520447
- Olofsson A. Differential operators for a scale of Poisson type kernels in the unit disc. J Anal Math. 2014;123:227–249. doi: 10.1007/s11854-014-0019-4
- Olofsson A. On a weighted harmonic Green function and a theorem of Littlewood. Bull Sci Math. Forthcoming.
- Sherbert DR. The structure of ideals and point derivations in Banach algebras of Lipschitz functions. Trans Amer Math Soc. 1964;111:240–272. doi: 10.1090/S0002-9947-1964-0161177-1
- Hardy GH, Littlewood JE. Some properties of fractional integrals. II. Math Z. 1932;34:403–439. doi: 10.1007/BF01180596
- Katznelson Y. An introduction to harmonic analysis. New York: Dover; 1976.
- Zygmund A. Trigonometric series. Vols. I & II. London: Cambridge University Press; 1968.
- Li P, Wang X. Lipschitz continuity of α-harmonic functions. Hokkaido Math J. 2019;48:85–97. doi: 10.14492/hokmj/1550480645
- Chen X. Lipschitz continuity for solutions of the α¯-Poisson equation. Sci China Math. 2019;62. DOI:10.1007/s11425-016-9300-5.
- Gehring FW, Martio O. Lipschitz classes and quasiconformal mappings. Ann Acad Sci Fenn Ser A I Math. 1985;10:203–219. doi: 10.5186/aasfm.1985.1022
- Dyakonov KM. Equivalent norms on Lipschitz-type spaces of holomorphic functions. Acta Math. 1997;178:143–167. doi: 10.1007/BF02392692
- Carlsson M, Wittsten J. The Dirichlet problem for standard weighted Laplacians in the upper half plane. J Math Anal Appl. 2016;436:868–889. doi: 10.1016/j.jmaa.2015.12.026
- Chen X, Kalaj D. A representation theorem for standard weighted harmonic mappings with an integer exponent and its applications. J Math Anal Appl. 2016;444:1233–1241. doi: 10.1016/j.jmaa.2016.07.035
- Courteaut K, Olofsson A. Asymptotics for a Beta function arising in potential theory. Submitted for publication.
- Krantz S, Wójcicki P. The weighted Bergman kernel and the Green's function. Complex Anal Oper Theory. 2017;11:217–225. doi: 10.1007/s11785-016-0593-9
- Li P, Wang X, Xiao Q. Several properties of α-harmonic functions in the unit disk. Monatsh Math. 2017;184:627–640. doi: 10.1007/s00605-017-1065-7
- Borichev A, Hedenmalm H. Weighted integrability of polyharmonic functions. Adv Math. 2014;264:464–505. doi: 10.1016/j.aim.2014.07.020
- Beardon AF, Minda D. The hyperbolic metric and geometric function theory. New Delhi: Narosa; 2007. (Quasiconformal mappings and their applications; 956).
- Gamelin TW. Complex analysis. New York: Springer; 2001.
- Hörmander L. The analysis of linear partial differential operators III. Berlin: Springer; 1994.
- Andrews GE, Askey R, Roy R. Special functions. Cambridge: Cambridge University Press; 1999. (Encyclopedia Math. Appl.; 71).
- Hörmander L. The analysis of linear partial differential operators I. Berlin: Springer; 1990.