First Dirichlet eigenvalue of the weighted 1-Laplacian operator

Rosa Barbato, José C. Sabina de Lis and Sergio Segura de León · 2026

Read First Dirichlet eigenvalue of the weighted 1-Laplacian operator by Rosa Barbato, José C. Sabina de Lis and Sergio Segura de León (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-2605-25642.txt.

Introduction

Barbato, Sabina de Lis and Segura de León study the first Dirichlet eigenvalue of the weighted 1-Laplacian operator. They prove it equals the weighted Cheeger constant. The proof follows the weighted p-Laplacian's first Dirichlet eigenvalue as p approaches 1. When the weight is Lipschitz-continuous, the limit of eigenvalues exists and equals the weighted Cheeger constant, and normalized eigenfunctions converge to a bounded limit function. For general weights, they identify the first eigenvalue with the same constant and prove its eigenfunction is bounded.

Catalog record

CREATORRosa Barbato, José C. Sabina de Lis and Sergio Segura de León
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE69169
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.25642

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

First Dirichlet eigenvalue of the weighted 1-Laplacian operator Rosa Barbato, José C. Sabina de Lis and Sergio Segura de León Abstract . In this paper, we study the eigenvalue problem { − div ​ ( a ⁡ ( x ) ​ D ​ u | D ​ u | ) = Λ ​ b ​ ( x ) ​ u | u | in ​ Ω u = 0 on ​ ∂ Ω , \left\{\begin{array}[]{cl}-\hbox{div\,}\left(a(x)\frac{Du}{|Du|}\right)=\Lambda\,b(x)\frac{u}{|u|}&\text{in }\Omega\\ u=0&\text{on }\partial\Omega,\end{array}\right. where a ⁡ ( x ) a(x) and b ⁡ ( x ) b(x) are suitable nonnegative functions. We prove that the first eigenvalue coincides with the weighted Cheeger constant. To see this identity, we analyze the behavior of the first Dirichlet eigenvalue of the weighted p p -Laplacian operator as p p goes to 1 1 . In the case that the weight a ⁡ ( x ) a(x) be Lipschitz-continuous, we show that the limit of eigenvalues of the weighted p p -Laplacian exists, and it is the weighted Cheeger constant. In addition, we check that the sequence of normalized p p -eigenfunctions converges to the normalized eigenfunction of our limiting problem, which turns out to be bounded. For more general weights, we identify the first eigenvalue of the weighted 1-Laplacian operator with the weighted Cheeger constant and prove that the associated eigenfunction is bounded. 2020 Mathematics Subject Classification 35P30, 35J60, 35J92 1. Introduction The study of properties of differential operators with weights is a topic of great interest in mathematical analysis. This interest especially concerns weighted p -Laplacian equations, including eigenvalue problems. In this paper, our objective is to check that the first eigenvalue of the weighted 1-Laplacian coincides with the weighted Cheeger constant. We also consider the problem of the first eigenvalue of the weighted p -Laplacian and let p go to 1. We aim to analyze the convergence of both eigenvalues and eigenfunctions. Moreover, we want to provide conditions to ensure that their limits form an eigenpair of the limit problem, which involves the weighted 1 -Laplacian. 1.1. Our starting point The Dirichlet spectrum of the p -Laplacian operator \Delta_{p}(u)=\dive\left(|\nabla u|^{p-2}\nabla u\right) was exhaustively analyzed at the end of the 20th century, beginning in the mid-80s (see [ 15 , 16 ] ). The main issues concern features of the first eigenvalue (it is the least eigenvalue and the only one whose eigenfunctions do not change of sign) which is proved to be simple on bounded domains and isolated (see [ 3 , 36 ] ). Moreover, the existence of a sequence of eigenvalues that tends to infinity (see [ 26 ] ) is shown. For further information on the p -Laplacian spectrum, we refer to [ 35 ] and [ 38 ] . The analysis of the first eigenvalue has been extended to the following weighted eigenvalue problem (1.1) \begin{cases}-\dive\left(a(x)\left|\nabla u_{p}\right|^{p-2}\nabla u_{p}\right)=\lambda b(x)\left|u_{p}\right|^{p-2}u_{p}&\textrm{in}\;\Omega\\ u_{p}=0&\textrm{on}\;\partial\Omega,\end{cases} where \Omega is an open and bounded domain, while a(x) is a positive function and b(x) is a nonnegative one, possibly unbounded that does not vanish identically. A solution is sought in the weighted Sobolev space W_{0}^{1,p}(\Omega,a) . This problem was originally introduced in [ 24 ] and later studied in [ 14 , 34 , 22 , 23 ] (related problems can be found in [ 27 , 29 ] ). The hypotheses assumed in [ 24 , Chapter 3] are: (1.2a) a(x)\in L_{loc}^{1}(\Omega) (1.2b) a^{-1}(x)\in L^{s}(\Omega)\text{ for certain }s\in\left(\frac{N}{p},\infty\right)\cap\left[\frac{1}{p-1},\infty\right) (1.2c) b(x)\in L^{r}(\Omega)\text{ for some }r>\frac{Ns}{ps-N}. Under these assumptions, it is proved that (1.3) \lambda_{p}=\inf_{v\in W_{0}^{1,p}(\Omega,a)}\left\{\int_{\Omega}a(x)|\nabla v|^{p}dx\>:\>\int_{\Omega}b(x)|v|^{p}dx=1\right\} is positive, and it is the least eigenvalue of problem ( 1.1 ). This eigenvalue turns out to be simple, and its associated eigenfunction is bounded and does not change its sign. Regarding the existence of a sequence of eigenvalues in ( 1.1 ) that tends to \infty , it can be found in [ 34 ] . It is worth pointing out that the assumption of non-negativeness of function b(x) can be weakened. Indeed, weights b(x) that change sign are handled in [ 14 ] , under the sole hypothesis the positive part b^{+}(x) does not vanish, and the least positive eigenvalue is studied (see also [ 34 ] where the existence of a nondecreasing sequence of positive eigenvalues is established). This extension also applies in the limit case p=1 . Nonetheless, in this paper, we are only dealing with non-negative weights. 1.2. Limit problem As already mentioned, our main concern is to analyze the principal eigenvalue of the limit problem (1.4) \begin{cases}-\dive\left(a(x)\frac{Du}{|Du|}\right)=\Lambda b(x)\frac{u}{|u|}&\textrm{in}\;\Omega\\ u=0&\textrm{on}\;\partial\Omega.\end{cases} As far as we know, this problem has not yet been addressed. The limit operator of the p –Laplacian as p tends to 1 is formally given by \displaystyle\dive\left(\frac{Du}{|Du|}\right) . It is a highly singular operator whose natural energy space is BV(\Omega) , the space of all functions of bounded variation. The notion of solution for the 1 -Laplacian with a Dirichlet boundary condition was introduced in [ 4 , 19 ] through a vector field which plays the rôle of the quotient Du/|Du| (where Du and |Du| are Radon measures). It is also subject to a weak form for the boundary condition. To obtain the desired solution, a Green’s formula is needed. It is stated in [ 4 ] and is based on the pairing theory developed by Anzellotti in [ 5 ] . After these pioneering articles, the study of problems involving 1 -Laplacian has grown hugely, mainly due to its applications to image processing. This interest includes the anisotropic case (see, for example, [ 42 ] for a bounded weight and [ 43 ] for an unbounded one). The first eigenvalue for the 1 -Laplacian was first studied in [ 20 ] and [ 32 ] . In the first paper, the infimum of the Rayleigh quotient \lambda_{1}(\Omega)=\inf\left\{\frac{\int_{\Omega}|\nabla v|\,dx}{\int_{\Omega}|v|\,dx}\>:\>v\in W_{0}^{1,1}(\Omega)\backslash\{0\}\right\} is also identified as (1.5) \inf\left\{\frac{\int_{\Omega}|Dv|+\int_{\partial\Omega}|v|\,d\mathcal{H}^{N-1}}{\int_{\Omega}|v|\,dx}\>:\>v\in BV(\Omega)\backslash\{0\}\right\} and the eigenfunctions are found to have the form of characteristic functions of Cacciopoli sets. The second paper is devoted to checking that the first eigenvalue is exactly the Cheeger constant of the domain while a corresponding eigenfunction is recognized as the characteristic function of a Cheeger set. For further developments, we refer to [ 1 , 21 , 33 ] . The notion of the spectrum of the 1 -Laplacian is due to Chang in [ 11 ] using critical point theory for non-smooth functionals: an eigenvalue is a critical value, in the sense of the weak slope, of the functional given by J(v)=\int_{\Omega}|Dv|+\int_{\partial\Omega}|v|\,d\mathcal{H}^{N-1} . As expected, every eigenvalue is a solution of the associated Dirichlet problem involving the 1 -Laplacian. Nevertheless, there are solutions to this elliptic problem that are not critical values of functional J , and Chang provides some examples of these spurious solutions. In our weighted setting, the variational formulation ( 1.5 ) becomes (1.6) \inf\left\{\frac{\int_{\Omega}a(x)|Dv|+\int_{\partial\Omega}a(x)|v|\,d\mathcal{H}^{N-1}}{\int_{\Omega}b(x)|v|\,dx}\>:\>v\in BV(\Omega,a)\backslash\{0\}\right\}. Our aim is to see that this infimum is attained and its value coincides with both the first eigenvalue \Lambda(\Omega,a,b) and the weighted Cheeger’s constant. 1.3. Asymptotic behavior In order to study equations involving the 1 -Laplacian, a widely used method is to take the limit as p goes to 1 of similar equations in which the p -Laplacian operator appears (see, for instance, [ 4

Read the complete text