Riesz transforms and the BAUPP and BWGL criteria for uniform rectifiability
Xavier Tolsa · 2025
Read Riesz transforms and the BAUPP and BWGL criteria for uniform rectifiability by Xavier Tolsa (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2509-26547.txt.
Introduction
Xavier Tolsa contributed a short but pointed note in 2025 to shape-based measure theory, the study of which measures behave like surface area on possibly rough sets. A measure is uniformly rectifiable when it is, in a strong average sense, well approximated by planes. Two competing rules catch that estimate trait. And a major problem of David and Semmes asked how they relate through the boundedness of the Riesz transform, the field's central singular integral.
Tolsa shows that for an Ahlfors regular measure in codimension one. Boundedness of the Riesz transform along with the bilateral estimate condition known as BAUPP already forces the stronger BWGL condition, and hence uniform rectifiability. That chain settles the David-Semmes problem in codimension one by a route that avoids the older, heavier BAUP rule entirely, simplifying the known proof form.
Catalog record
| CREATOR | Xavier Tolsa |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 28 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2509-26547
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Riesz transforms and the BAUPP and BWGL criteria for uniform rectifiability Thanks: Supported by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement 101018680). Also partially supported by MICIU (Spain) under the grant PID2024-160507NB-I00. Xavier Tolsa Abstract . In this note it is shown that if μ \mu is an n n -Ahlfors regular measure in ℝ n + 1 {\mathbb{R}}^{n+1} such that the n n -dimensional Riesz transform is bounded in L 2 ( μ ) L^{2}(\mu) and the so-called BAUPP (bilateral approximation by unions of parallel planes) condition holds for μ \mu , then μ \mu satisfies the BWGL (bilateral weak geometric lemma), and so μ \mu is uniformly n n -rectifiable. In this way, one can solve the David-Semmes problem in codimension one without relying on the BAUP (bilateral approximation by unions of planes) criterion of David and Semmes. 1. Introduction A Radon measure \mu in {\mathbb{R}}^{d} is called n -Ahlfors regular if (1.1) C^{-1}r^{n}\leq\mu(B(x,r))\leq Cr^{n}\quad\mbox{ for all $x\in\operatorname{supp}\mu$ and $0<r\leq{\rm diam}(\operatorname{supp}\mu)$.} On the other hand, \mu is called uniformly n -rectifiable if it is n -Ahlfors regular and there exist constants \theta,M>0 such that for all x\in\operatorname{supp}\mu and all 0<r\leq{\rm diam}(\operatorname{supp}\mu) there is a Lipschitz mapping g from the ball B_{n}(0,r) in {\mathbb{R}}^{n} to {\mathbb{R}}^{d} with \text{Lip}(g)\leq M such that \mu(B(x,r)\cap g(B_{n}(0,r)))\geq\theta r^{n}. The notion of uniform rectifiability is a quantitative version of n -rectifiability introduced by David and Semmes (see [ DS1 ] and [ DS2 ] ) which has attracted much attention because its applications to harmonic analysis and PDE’s in rough settings, among other things. An important question in this area is the so-called David-Semmes problem, which consists in proving that if \mu is n -Ahlfors regular and the Riesz transform operator {\mathcal{R}}_{\mu} is bounded in L^{2}(\mu) , then \mu is uniformly n -rectifiable. The ( n -dimensional) Riesz transform of a signed Radon measure \nu is defined by {\mathcal{R}}\nu(x)=\int\frac{x-y}{|x-y|^{n+1}}\,d\nu(y), whenever the integral makes sense. We also write {\mathcal{R}}_{*}\nu(x)=\sup_{{\varepsilon}>0}|{\mathcal{R}}_{{\varepsilon}}\nu(x)|,\qquad\operatorname{pv}{\mathcal{R}}\nu(x)=\lim_{{\varepsilon}>0}{\mathcal{R}}_{{\varepsilon}}\nu(x), in case that the latter limit exists. Remark that, sometimes, abusing notation we will write {\mathcal{R}}\nu instead of \operatorname{pv}{\mathcal{R}}\nu . For f\in L^{1}_{loc}(\mu) and a positive Radon measure \mu , one writes {\mathcal{R}}_{\mu}f(x)={\mathcal{R}}(f\mu)(x) . Given {\varepsilon}>0 , the {\varepsilon} -truncated Riesz transform of \mu equals {\mathcal{R}}_{\varepsilon}\mu(x)=\int_{|x-y|>{\varepsilon}}\frac{x-y}{|x-y|^{n+1}}\,d\mu(y), and the operator {\mathcal{R}}_{\mu,{\varepsilon}} is defined by {\mathcal{R}}_{\mu,{\varepsilon}}f(x)={\mathcal{R}}_{\varepsilon}(f\mu)(x) . We say that {\mathcal{R}}_{\mu} is bounded in L^{2}(\mu) if the operators {\mathcal{R}}_{\mu,{\varepsilon}} are bounded uniformly in L^{2}(\mu) uniformly on {\varepsilon} , and then we denote \|{\mathcal{R}}_{\mu}\|_{L^{2}(\mu)\to L^{2}(\mu)}=\sup_{{\varepsilon}>0}\|{\mathcal{R}}_{\mu,{\varepsilon}}\|_{L^{2}(\mu)\to L^{2}(\mu)}. The David-Semmes problem was solved in the affirmative in the case n=1 in [ MMV ] by Mattila, Melnikov, and Verdera, and in the case n=d-1 in [ NToV ] by Nazarov, the author, and Volberg [ NToV ] . For other values of n , this is still an open problem. One of the essential ingredients of the proof in [ NToV ] is the so called BAUP criterion. To state this, we need need some additional notation. For a given {\varepsilon}>0 and x\in\operatorname{supp}\mu , r>0 , we write (x,r)\in{\mathsf{BAUP}}({\varepsilon}) (BAUP stands for bilateral approximation by a union of planes) if there exists a family \tilde{I}_{x,r} of n -planes such that, setting \tilde{H}_{x,r}=\bigcup_{L\in\tilde{I}_{x,r}}L, it holds \sup_{x\in\operatorname{supp}\mu\cap B(x,r)}\frac{{\rm dist}(x,\tilde{H}_{x,r})}{\ell(Q)}+\sup_{x\in\tilde{H}_{x,r}\cap B(x,r)}\frac{{\rm dist}(x,\operatorname{supp}\mu)}{r}\leq{\varepsilon}. Then we say that \mu satisfies the BAUP condition, if for any {\varepsilon}>0 , the following Carleson type condition holds: \int_{B}\int_{0}^{r(B)}\chi_{(x,t)\not\in{\mathsf{BAUP}}({\varepsilon})}\,\frac{dt}{t}\,d\mu(x)\leq\mu(B), for any ball B centered at \operatorname{supp}\mu , where r(B) denotes the radius of B . A deep result of David and Semmes (see [ DS2 , Chapter II.3] ) asserts that if \mu is n -Ahlfors regular and it satisfies the BAUP condition, then \mu is uniformly n -rectifiable. The proof of this result is difficult and lengthy. The purpose of this note is to give a more straightforward argument that avoids the application of the BAUP criterion for the solution of the David-Semmes problem in codimension 1 and instead relies on the BGWL (bilateral weak geometric lemma). The definition of the BGWL condition is analogous to the one of the BAUP condition, with the family \tilde{I}_{x,r} formed in this case by a single n -plane. Remark that the proof of the fact that the BGWL suffices for the uniform n -rectifiability of \mu is considerably easier than the one for the BAUP condition, specially in codimension one (see [ DS2 , Chapter II.2] ). In [ NToV ] it is shown that if \mu is an n -Ahlfors regular measure in {\mathbb{R}}^{n+1} such that {\mathcal{R}}_{\mu} is bounded in L^{2}(\mu) , then the BAUP condition holds for \mu and so \mu is uniformly n -rectifiable, by the aforementioned criterion of David and Semmes. In fact, a quick inspection of the arguments in [ NToV ] shows that \mu satisfies something (a priori) stronger than the BAUP condition, namely what we call the BAUPP (bilateral approximation by union of parallel planes. This is defined as the BAUP, but only allowing parallel n -planes in family \tilde{I}_{x,r} . Since this notion is important for our purposes we state it more precisely. For a given {\varepsilon}>0 and x\in\operatorname{supp}\mu , r>0 , we write (x,r)\in{\mathsf{BAUPP}}({\varepsilon}) if the exists a family I_{x,r} of parallel n -planes such that, setting H_{x,r}=\bigcup_{L\in I_{x,r}}L, it holds \sup_{x\in\operatorname{supp}\mu\cap B(x,r)}\frac{{\rm dist}(x,H_{x,r})}{\ell(Q)}+\sup_{x\in H_{x,r}\cap B(x,r)}\frac{{\rm dist}(x,\operatorname{supp}\mu)}{r}\leq{\varepsilon}. Then we say that \mu satisfies the BAUPP condition, if for any {\varepsilon}>0 , the following holds: \int_{B}\int_{0}^{r(B)}\chi_{(x,t)\not\in{\mathsf{BAUP}}\c{C}P({\varepsilon})}\,\frac{dt}{t}\,d\mu(x)\leq\mu(B), for any ball B centered at \operatorname{supp}\mu . Remark that, trivially, we have (1.2) BWGL\;\Rightarrow\;BAUPP\;\Rightarrow\;BAUP. The implication BAUP\;\Rightarrow\;BWGL also holds, but it is much more complicated. Hence, the three conditions in ( 1.2 ) are equivalent to the uniform rectifiability of \mu . In this note we will prove the following. Theorem 1.1 . Let \mu be an n -Ahlfors regular measure in {\mathbb{R}}^{n+1} satisfying the BAUPP condition such that the n -dimensional Riesz transform {\mathcal{R}}_{\mu} is bounded in L^{2}(\mu) . Then \mu satisfies the BWGL. By the discussion above, it is clear that, combining this result with the arguments in [ NToV ] , one deduces that the L^{2}(\mu) boundedness of the n -dimensional Riesz transform {\mathcal{R}}_{\mu} implies the BWGL for \mu , and so its uniform n -rectifiability. In this way, we can skip the BAUP criterion for the solution of the David-Semmes problem. 2. Proof of Theorem 1.1 For an n -Ahlfors regular measure \mu such as the one in the theorem, we let {\mathcal{D}}_{\mu} be an associated David-Semmes dyadic lattice, such as the one introduced in [ DS2 , Chapter I.3] . For Q\in{\mathcal{D}}_{\mu,k} , we denote \ell(Q)=2^{-k} , and we assume that {\rm diam}(Q)\leq\ell(Q) . We let B_{Q} be a ball concentric with Q with radius 2\ell(Q)=2^{-k+1} . For a given {\varepsilon}>0 , we write Q\in{\mathsf{BAUPP}}({\varepsilon}) if the exists a family I_{Q} of parallel n -planes intersecting B_{Q} such that, setting H_{Q}=\bigcup_{L\in I_{Q}}L, it holds \sup_{x\in\operatorname{supp}\mu\cap B_{Q}}\frac{{\rm dist}(x,H_{Q})}{\ell(Q)}+\sup_{x\in H_{Q}\cap B_{Q}}\frac{{\rm dist}(x,\operatorname{supp}\mu)}{\ell(Q)}\leq{\varepsilon}. It is easy to check that \mu satisfies the BAUPP condition if and only if, for any {\varepsilon}>0 , the family {\mathcal{D}}_{\mu}\setminus{\mathsf{BAUPP}}({\varepsilon}) is a Carleson family. That is, for any R\in{\mathcal{D}}_{\mu} , (2.1) \sum_{Q\in{\mathcal{D}}_{\mu}(R)\setminus{\mathsf{BAUPP}}({\varepsilon})}\mu(Q)\leq C({\varepsilon})\,\mu(R). Our goal is to show that the assumptions of Theorem 1.1 imply that ( 2.1 ) holds for any R\in{\mathcal{D}}_{\mu} . To this end, for a