Sobolev regularity of the Beurling transform on planar domains

Abstract

Consider a Lipschitz domain $\Omega$ and the Beurling transform of its characteristic function $\mathcal{B} \chi_\Omega(z)= - {\rm p.v.}\frac1{\pi z^2}*\chi_\Omega (z) $. It is shown that if the outward unit normal vector $N$ of the boundary of the domain is in the trace space of $W^{n,p}(\Omega)$ (i.e., the Besov space $B^{n-1/p}_{p,p}(\partial\Omega)$) then $\mathcal{B} \chi_\Omega \in W^{n,p}(\Omega)$. Moreover, when $p>2$ the boundedness of the Beurling transform on $W^{n,p}(\Omega)$ follows. This fact has far-reaching consequences in the study of the regularity of quasiconformal solutions of the Beltrami equation.Comment: 33 pages, 8 figures. arXiv admin note: text overlap with arXiv:1507.0433

Sufficient conditions for C^1,α parametrization and rectifiability

“…By appropriate generalization we mean to use polynomials instead of d-planes to approximate the set. This idea is not new, see for instance [Dor85a,Dor85b] and, more recently, [Pra17], Section 2.2.…”

Sobolev regularity of quasiconformal mappings on domains

“…and, in particular, h m3 is infinitely many times differentiable in Ω. Therefore, by Green's formula (2.3) and the cancellation of the integrand (see [Pra15,(3.2)]), for j > 0 we have…”

“…In [Pra15] the author proved that the Beurling transform is bounded in W n,p (Ω), reaching the following result:…”

“…The reader may expect to find a bound with a polynomial behavior with respect to m, but the fact is that the author has not been able to get such an estimate. Instead, we will use an upper bound for the norm with exponential growth on m but the base will be chosen as close to 1 as desired, as shown in [Pra15,Theorem 3.15]. This will suffice to prove Theorem 1.1.…”

Dorronsoro's theorem in Heisenberg groups

“…Rearranging the terms gives (2.3). We learned this quick argument from a paper of Prats [17,Remark 2.4].…”