The Kato square root estimate with Robin boundary conditions
Sebastian Bechtel, Andreas Rosén · 2026
Read The Kato square root estimate with Robin boundary conditions by Sebastian Bechtel, Andreas Rosén (2026) 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-2601-04678.txt.
Introduction
Sebastian Bechtel and Andreas Rosén prove a square root estimate in the one boundary setting that had resisted it, in a 2026 arXiv paper. The Kato square root problem asks whether the square root of a divergence-form elliptic operator behaves like a first derivative. An estimate at the heart of solving boundary value problems with rough, merely bounded complex coefficients. Most boundary conditions have yielded; Robin conditions, where the boundary leaks proportionally to the solution through a possibly unbounded conductivity, had not.
On bounded, locally uniform domains, the authors prove the estimate for accretive coefficients under exactly that Robin condition. The proof is notable for what it cannot use: nearly every earlier Kato estimate ran through a first-order systems approach, and the authors state that no such approach seems possible here. The result closes a gap in the theory of elliptic operators with realistic boundary behaviour.
Catalog record
| CREATOR | Sebastian Bechtel, Andreas Rosén |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 66 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2601-04678
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
The Kato square root estimate with Robin boundary conditions Thanks: Andreas Rosén was supported by the Swedish Research Council (Grant 2022-03996), which also funded the short visit by Sebastian Bechtel to Gothenburg during which this work was initiated. This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 101034255 . Sebastian Bechtel, Andreas Rosén Abstract . We prove the Kato square root estimate for second-order divergence form elliptic operators − div ( A ∇ ) -\Divv(A\nabla) on a bounded, locally uniform domain D ⊆ ℝ n D\subseteq\mathbb{R}^{n} , for accretive coefficients A ∈ L ∞ ( D , ℂ n ) A\in\mathrm{L}^{\infty}(D;\mathbb{C}^{n}) , under the Robin boundary condition ν ⋅ A ∇ u + b u = 0 \nu\cdot A\nabla u+bu=0 for a (possibly unbounded) boundary conductivity b b . In contrast to essentially all previous estimates of Kato square root operators, no first-order approach seems possible for the Robin boundary conditions. 2020 Mathematics Subject Classification Primary: 42B37, 35J25. Secondary: 46E35, 47A60. 1. Introduction This paper concerns estimates for the Kato square root \sqrt{L} of divergence form second-order elliptic operators L=-\Divv(A\nabla) . The estimate (1) \displaystyle\|\sqrt{L}u\|_{\mathrm{L}^{2}(\mathbb{R}^{n})}\approx\|\nabla u\|_{\mathrm{L}^{2}(\mathbb{R}^{n})} conjectured by Kato [ 26 ] , was eventually proved in the seminal work [ 5 ] . Following this breakthrough, there has been many generalizations, for example to domains D\subseteq\mathbb{R}^{n} , imposing Dirichlet, Neumann or mixed such boundary conditions, to manifolds, to degenerate/unbounded coefficients A , to more general elliptic systems, and to parabolic operators. See for example [ 27 , 10 , 11 , 8 , 7 , 4 , 1 , 23 , 19 , 17 , 14 ] . Although the estimate ( 1 ) seems to concern second-order PDEs, it can be argued that it actually concerns first-order elliptic systems. Indeed, ( 1 ) readily follows from the boundedness of a generalized Cauchy singular integral operator \sgn(BD) , where (2) \displaystyle B=\begin{bmatrix}1&0\\ 0&A\end{bmatrix}\quad\text{and}\quad D=\begin{bmatrix}0&-\Divv\\ \nabla&0\end{bmatrix}. In dimension n=1 this observation goes back to [ 6 ] , and after the Kato square root estimate ( 1 ) was established in full generality in [ 5 ] , the framework from [ 6 ] was generalized in [ 9 ] to arbitrary dimension. See [ 3 ] for a simplified presentation of this first-order approach. Although the generalized first-order Cauchy–Riemann system BD looks more complicated than the second-order operator L , with essentially the same techniques one can more generally prove bounds (3) \displaystyle\|f(BD)u\|_{\mathrm{L}^{2}}\lesssim\|f\|_{\infty}\|u\|_{\mathrm{L}^{2}} for all operators f(BD) in the holomorphic functional calculus of BD . This has important applications for example to the study of boundary value problems. See [ 3 , 7 , 2 ] . In all previous works on Kato square root estimates, to the best of our knowledge (and possible with the exception of [ 23 ] ), it has been possible either to do a second-order proof (roughly following the original proof [ 5 ] ), or to do a first-order proof (roughly following [ 9 ] or [ 3 ] ). Also for Kato square root estimates on domains, both choices have been available [ 10 , 15 , 21 ] . For Dirichlet boundary conditions one uses (4) \displaystyle D=\begin{bmatrix}0&-\Divv\\ \nabla_{0}&0\end{bmatrix}, where \dom(\nabla_{0})=\mathrm{W}^{1,2}_{0}(D) and -\Divv=\nabla_{0}^{*} . For Neumann boundary conditions, one uses (5) \displaystyle D=\begin{bmatrix}0&-\Divv_{0}\\ \nabla&0\end{bmatrix}, where \dom(\nabla)=\mathrm{W}^{1,2}(D) and -\Divv_{0}=\nabla^{*} . If the domain D is sufficiently regular, one finds \dom(\Divv_{0})=\{f\in\mathrm{L}^{2}(D)\colon\Divv(f)\in\mathrm{L}^{2}(D),\nu\cdot f|_{\partial D}=0\} , where \nu denotes the outwards normal to D . Even for mixed Dirichlet/Neumann boundary conditions one can define a self-adjoint operator D by putting the Dirichlet boundary conditions on \nabla and the Neumann boundary conditions on \Divv . In the present paper, we prove the Kato square root estimate on domains D\subseteq\mathbb{R}^{n} under Robin boundary conditions (formally) given by (6) \displaystyle\nu\cdot A\nabla u+bu=0\quad\text{on }\partial D. We recall that when Fourier’s law is used to model heat equilibrium by \Divv(A\nabla u)=0 , then A is the conductivity in D , while b\geq 0 is the conductivity across the boundary \partial D . The extreme case b=0 means that D is an insulated body and corresponds to Neumann boundary conditions, while the limit b\to\infty means that u|_{\partial D} must equal the outside temperature 0 . Unlike for previous works, for the Robin boundary conditions we are forced to do a second-order proof of the Kato square root estimate. Indeed, it is not clear to us even how to define a first-order self-adjoint operator D which is relevant for the Robin boundary conditions. Also our second-order proof here presents new challenges since the Robin form (7) \displaystyle a(u,v)=\iint_{D}A\nabla u\cdot\overline{\nabla v}\,\mathrm{d}x+\int_{\partial D}b\gamma(u)\overline{\gamma(v)}\,\mathrm{d}\sigma contains a boundary term that needs to be converted to a boundary condition for the Robin operator L . At the technical level, a main novelty in the proof is Lemma 5.6 , which allows us to reduce the Kato square root estimate to a square function estimate for a family of operators acting only on functions in the interior D . Our main result is Theorem 3.8 , where the Robin boundary conditions \nu\cdot A\nabla u+bu=0 are imposed in a weak sense on a subset \Gamma\subseteq\partial D of dimension d , n-2<d<n , of the boundary, and Neumann boundary conditions on the remainder of \partial D . 2. Geometry and function spaces Fix the dimension n\geq 2 . We work in an open, bounded and possibly non-connected set D\subseteq\mathbb{R}^{n} . Fix a (relatively) open, non-empty subset \Gamma\subseteq\partial D . Throughout, we distinguish integrals in the interior and on the boundary by writing \iint_{D}\,\mathrm{d}x and \int_{\Gamma}\,\mathrm{d}\sigma . To save space, we occasionally drop \,\mathrm{d}x and \,\mathrm{d}\sigma from the notation. We introduce some notions from geometric measure theory in order to ensure some crucial properties for D and \Gamma as well as for function spaces on them. Definition 2.1 (Locally uniform) . Call D locally uniform, if there exist \varepsilon\in(0,1] , \delta\in(0,\infty] and c\in(0,\infty] such that the following hold: (i) All points x,y\in D with |x-y|<\delta can be joined in D by an \varepsilon -cigar , that is to say, a rectifiable curve c\subseteq D of length \ell(c)\leq\varepsilon^{-1}|x-y| such that (8) \displaystyle\dist(z,\partial D)\geq\frac{\varepsilon|z-x||z-y|}{|x-y|},\quad z\in c. (ii) All connected components D^{\prime} of D satisfy \diam(D^{\prime})\geq c . Remark 2.2 . When D is connected, the definition coincides with the one of Jones [ 25 ] . He has studied properties of Sobolev spaces on D , including extension and density results. The non-connected case is a by-product of [ 12 ] . If a set D\subseteq\mathbb{R}^{n} is locally uniform, then it is in particular interior thick in the sense of the following definition, see [ 14 , Prop. 2.9] . Interior thickness will be important to ensure standard properties of function spaces on D on the one hand, and to ensure the existence of a dyadic structure on D on the other hand, see Section 5.4 for the latter. Definition 2.3 (Interior thickness) . Call O interior thick if the following holds true: (ITC) \displaystyle\exists c\in(0,1]\colon\;\forall x\in D,r\in(0,1]\colon\quad|D\cap\mathrm{B}(x,r)|\geq c|\mathrm{B}(x,r)|. For the boundary part \Gamma , we introduce the following. Definition 2.4 ( d -set) . Call \Gamma a d -set for d\in(0,n] if the following holds true: (9)