Abstract
We derive the hydrodynamic limit of the Kawasaki dynamics for the one-dimensional conservative system of unbounded real-valued spins with arbitrary strong, quadratic and finite-range interactions. This significantly extends prior results for bounded interaction by Rezakhanlou and complements results obtained by H.T. Yau. The result is obtained by adapting two-scale approach of Grunewald, Otto, Villani and Westdickenberg combined with the authors’ recent approach on conservative systems with strong interactions.
Similar content being viewed by others
References
Bakry, D., Émery, M.: Diffusions hypercontractives. In: Séminaire de Probabilités, XIX, 1983/84, volume 1123 of Lecture Notes in Math., pp. 177–206. Springer, Berlin (1985)
Bakry, D., Gentil, I., Ledoux, M.: Analysis and Geometry of Markov Diffusion Operators. Grundlehren der Mathematischen Wissenschaften [Fundamental Principles of Mathematical Sciences], vol. 348. Springer, Cham (2014)
Brascamp, H.J., Lieb, E.H.: Best constants in Young’s inequality, its converse, and its generalization to more than three functions. Adv. Math. 20(2), 151–173 (1976)
Deniz, D., Menz, G., Otto, F., Wu, T.: The quantitative hydrodynamic limit of the Kawasaki dynamics. (2018) arXiv:1807.09850
Dizdar, D., Menz, G., Otto, F., Wu, T.: Toward a quantitative theory of the hydrodynamic limit. (2018) arXiv preprint arXiv:1807.09857
Fathi, M.: A two-scale approach to the hydrodynamic limit part ii: Local Gibbs behavior. Latin Am. J. Prob. Math. Stat. 10, 10 (2012)
Fathi, M., Menz, G.: Hydrodynamic limit for conservative spin systems with super-quadratic, Partially Inhomogeneous Single-Site Potential. (2014) arXiv:1405.3327
Fritz, J.: On the hydrodynamic limit of a one-dimensional Ginzburg–Landau lattice model. The a priori bounds. J. Stat. Phys. 47, 551–572 (1987)
Grunewald, N., Otto, F., Villani, C., Westdickenberg, M.G.: A two-scale approach to logarithmic Sobolev inequalities and the hydrodynamic limit. Ann. Inst. H. Poincaré Probab. Statist. 45(2), 302–351 (2009)
Guo, M.Z., Papanicolaou, G.C., Varadhan, S.R.S.: Nonlinear diffusion limit for a system with nearest neighbor interactions. Comm. Math. Phys. 118(1), 31–59 (1988)
Gross, L.: Logarithmic Sobolev inequalities. Am. J. Math. 97, 1061–1083 (1975)
Henderson, C., Menz, G.: Equivalence of a mixing condition and the LSI in spin systems with infinite range interaction. Stochastic Process. Appl. 126(10), 2877–2912 (2016)
Holley, R., Stroock, D.: Logarithmic Sobolev inequalities and stochastic Ising models. J. Stat. Phys. 46(5–6), 1159–1194 (1987)
Kipnis, C., Landim, C.: Scaling limits of interacting particle systems. Grundlehren der Mathematischen Wissenschaften [Fundamental Principles of Mathematical Sciences], vol. 320. Springer, Berlin (1999)
Kwon, Y., Lee, J., Menz, G.: Equivalence of the grand canonical ensemble and the canonical ensemble on 1d-lattice systems. (2019) arXiv:1910.07702
Kwon, Y., Menz, G.: Strict convexity of the free energy of the canonical ensemble under decay of correlations. J. Stat. Phys., (2018)
Kwon, Y., Menz, G.: Decay of correlations and uniqueness of the infinite-volume Gibbs measure of the canonical ensemble of 1d-lattice systems. J. Stat. Phys. (2019)
Kwon, Y., Menz, G.: Uniform LSI for the canonical ensemble on the 1d-lattice. ESAIM: PS 24, 341–373 (2020)
Kosygina, E.: The behavior of the specific entropy in the hydrodynamic scaling limit. Ann. Probab. 29(3), 1086–1110, 07 (2001)
Ledoux, M.: The Concentration of Measure Phenomenon. Mathematical Surveys and Monographs, vol. 89. American Mathematical Society, Providence, RI (2001)
Ledoux, M.: Logarithmic Sobolev inequalities for unbounded spin systems revisted. Semester Probability XXXV. Lecture Notes in Math, Springer vol. 1755, pp. 167–194 (2001)
Landim, C., Panizo, G., Yau, H.T.: Spectral gap and logarithmic Sobolev inequality for unbounded conservative spin systems. Ann. Inst. H. Poincaré Probab. Statist 38(5), 739–777 (2002)
Shenglin, L.: Hydrodynamic scaling limits with deterministic initial configurations. Ann. Probab. 23(4), 1831–1852, 10 (1995)
Lu, S.L., Yau, H.T.: Spectral gap and logarithmic Sobolev inequality for Kawasaki and Glauber dynamics. Comm. Math. Phys. 156(2), 399–433 (1993)
Menz, G.: LSI for Kawasaki dynamics with weak interaction. Comm. Math. Phys. 307(3), 817–860 (2011)
Menz, G.: The approach of Otto-Reznikoff revisited. Electron. J. Probab. 19(107), 27 (2014)
Menz, G., Nittka, R.: Decay of correlations in 1D lattice systems of continuous spins and long-range interaction. J. Stat. Phys. 156(2), 239–267 (2014)
Otto, F., Reznikoff, M.G.: A new criterion for the logarithmic Sobolev inequality and two applications. J. Funct. Anal. 243(1), 121–157 (2007)
Rezakhanlou, F.: Hydrodynamic limit for a system with finite range interactions. Comm. Math. Phys. 129(3), 445–480 (1990)
Royer, G.: Une initiation aux inégalités de Sobolev logarithmiques. Cours Spéc., Soc. Math. France (1999)
Varadhan, S.R.S., Yau, H.-T.: Diffusive limit of lattice gas with mixing conditions. Asian J. Math. 1(4), 623–678 (1997)
Yau, H.T.: Relative entropy and hydrodynamics of Ginzburg-Landau models. Lett. Math. Phys. 22(1), 63–80 (1991)
Yau, H.T.: Logarithmic Sobolev inequality for lattice gases with mixing conditions. Comm. Math. Phys. 181(2), 367–408 (1996)
Zegarlinski, B.: The strong decay to equilibrium for the stochastic dynamics of unbounded spin systems on a lattice. Comm. Math. Phys. 175(2), 401–432 (1996)
Acknowledgements
This work is dedicated to Thomas M. Liggett (March 29, 1944–May 12, 2020). He was a great friend, mentor, teacher and colleague to us all at UCLA. This research has been partially supported by NSF grant DMS-1407558. KN is supported by the National Research Foundation of Korea (NRF-2019R1A5A1028324). The authors want to thank Felix Otto and H.T. Yau for bringing this problem to their attention. The authors are also thankful to many people discussing the problem and helping to improve the preprint. Among them are Tim Austin, Frank Barthe, Marek Biskup, Pietro Caputo, Jean-Dominique Deuschel, Max Fathi, Andrew Krieger, Michel Ledoux, Sangchul Lee, Felix Otto, Daniel Ueltschi and Tianqi Wu. The authors want to thank Marek Biskup, UCLA and KFAS for financial support.
Author information
Authors and Affiliations
Corresponding author
Additional information
Communicated by Christian Maes.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix A. Basic Properties of the Grand Canonical Ensemble and the Canonical Ensemble
In this section, we provide useful auxiliary results for the GCE and CE, previously obtained in [15,16,17,18]. The results stated in this section hold under the conditions (2), (3), (4) and (5). Recall that a generalized GCE \(\mu ^{\sigma }_N\) with external field \(\sigma \) is defined by
Remark A.1
Again, the CE \(\mu _{N,m}\) can be thought as a conditional probability distribution which emerges from the GCE \(\mu _{N}^{\sigma }\) conditioned on the mean spin
More precisely, we have
The following statement tells that the variance of the mean spin of the modified GCE \(\mu ^{\sigma }_N\) is well behaved.
Lemma A.2
(Lemma 1 in [18]). There exists a constant \(C>0\), uniform in N and \(\sigma \), such that
Define the (normalized) free energy \(A_N : \mathbb {R} \rightarrow \mathbb {R}\) by
The following lemma states that the free energy \(A_N\) is uniformly strictly convex.
Lemma A.3
(Lemma 2 in [18]). There is a constant \(C >0\), uniform in N and \(\sigma \), such that
Let us now introduce the definition of local, intensive and extensive functions.
Definition A.4
(Local, intensive and extensive functions/ observables). For a function \(f : \mathbb {R}^{\mathbb {Z}} \rightarrow \mathbb {C}\), denote \({{\,\textrm{supp}\,}}f \) by the minimal subset of \(\mathbb {Z}\) with \(f(x) = f\left( x^{{{\,\textrm{supp}\,}}f} \right) \). We call f a local function if it has a finite support independent of N. A function f is called intensive if there is a positive constant \(\varepsilon \) such that \(|{{\,\textrm{supp}\,}}f | \lesssim N^{1-\varepsilon }\). A function f is called extensive if it is not intensive.
The following proposition claims the exponential decay of correlations for the GCE.
Proposition A.5
(Lemma 6 in [16]). There exists a constant \(C>0\) such that for any intensive functions \(f, g : \mathbb {R}^{N} \rightarrow \mathbb {R}\),
The following moment estimate is a consequence of Proposition A.5.
Lemma A.6
(Lemma 3.2 in [17]). For each \(k \ge 1\), there exists a constant \(C = C(k)>0\) such that for any smooth function \(f : \mathbb {R}^{\Lambda } \rightarrow \mathbb {R}\),
We now introduce the principle of equivalence of observables. We first relate the external field \(\sigma \) of \(\mu _N ^{\sigma }\) and the mean spin m of \(\mu _{N,m}\) as follows:
Definition A.7
For each \(m \in \mathbb {R}\), we choose \(\sigma = \sigma _N(m) \in \mathbb {R}\) such that
Denoting \(m_i : = \int x_i \mu _N ^{\sigma } (\textrm{d}x)\) for each \(i \in [N]\), we equivalently get
The following proposition states the equivalence of ensembles results for GCE \(\mu _N ^{\sigma }\) and CE \(\mu _{N,m}\) with \(\sigma \) and m related by (85).
Proposition A.8
(Theorem 2.7 in [15]). There exist constants \(C,N_0>0\) such that for any intensive function \(f : \mathbb {R}^{N} \rightarrow \mathbb {R}\) and \(N \ge N_0\),
Finally, we state the exponential decay of correlations for the CE.
Proposition A.9
(Theorem 2.10 in [15]). There exist constants \(C,N_0>0\) such that for any intensive functions \(f, g : \mathbb {R}^{N} \rightarrow \mathbb {R}\) and \(N \ge N_0\),
Here, \(\sigma \) and m are related by (85).
Appendix B. Derivatives of Coarse-Grained Hamiltonian
In this section, we provide the first and second derivative formula for the coarse-grained Hamiltonian. First, we state an explicit formula for \(\bar{H}\), obtained in [25].
Lemma B.1
(Lemma 1 in [25]). For \(z \in X\) with \(Pz=0\) and \(y \in Y\), define \(H_{M}(z,y)\) by
where \(M=(M_{ij})_{1\le i,j\le n}\) is an interaction matrix in the Hamiltonian (cf. (1)). Then
Proof of Lemma B.1
For \(x \in X\) with \(Px=y\) and \(y \in Y\), let \(z = x - NP^*y\). Recalling the identity \(PNP^* = Id_Y\), we have \(Pz=0\). We then write the Hamiltonian H as
Because \(Pz = 0\), we have
It also holds by \(PNP^* = \text {Id}_Y\) that
Plugging (88) and (89) into (87) yields
and hence,
\(\square \)
Next, we compute the first derivative of the coarse-grained Hamiltonian \(\bar{H}\) using Lemma B.1.
Lemma B.2
For each \(l \in [M]\), it holds that
Proof of Lemma B.2
Recall that the inner product \(\langle \cdot , \cdot \rangle _Y\) is given by
First of all, noting that
we have
In addition, differentiating (86) yields
As a consequence, we obtain
Combining (90) and (91) with symmetry of \(M_{ij}\), i.e., \(M_{ij} = M_{ji}\), we have the desired equation
\(\square \)
The second derivatives of the coarse-grained Hamiltonian \(\bar{H}\) follow from a similar calculations.
Lemma B.3
(Lemma 2 in [25]). For \(l, n \in [M]\), we have
Appendix C. Criteria for the Logarithmic Sobolev Inequality
In this section, we state several standard criteria for deducing a LSI. For proofs we refer to the literature. For a general introduction and more comments on the LSI, we refer the reader to [2, 20, 21, 30].
Theorem C.1
(Tensorization Principle [11]). Let \(\mu _1\) and \(\mu _2\) be probability measures on Euclidean spaces \(X_1\) and \(X_2\), respectively. Suppose that \(\mu _1\) and \(\mu _2\) satisfy LSI\((\rho _1)\) and LSI\((\rho _2)\), respectively. Then the product measure \(\mu _1 \bigotimes \mu _2\) satisfies LSI\((\rho )\), where \(\rho = \min \{\rho _1, \rho _2\}\).
Theorem C.2
(Holley–Stroock Perturbation Principle [13]). Let \(\mu _1\) be a probability measure on Euclidean space X and \( \psi _b : X \rightarrow \mathbb {R}\) be a bounded function. Define a probability measure \(\mu _2\) on X by
Suppose that \(\mu _1\) satisfies LSI\((\rho _1)\). Then \(\mu _2\) also satisfies LSI with constant
where \(osc \psi _b : = \sup \psi _b - \inf \psi _b \).
Theorem C.3
(Bakry–Émery criterion [1]) Let X be a N-dimensional Euclidean space and \(H \in C^2 (X)\). Define a probability measure \(\mu \) on X by
Suppose that there is a constant \(\rho >0\) such that \({{\,\textrm{Hess}\,}}H \ge \rho \), in other words
Then \(\mu \) satisfies LSI\((\rho )\).
Theorem C.4
(Otto-Reznikoff Criterion [28]) Let \(X = X_1 \times \cdots \times X_N\) be a direct product of Euclidean spaces and \(H \in C^2 (X)\). Define a probability measure \(\mu \) on X by
Assume that
-
For each \(i \in [N]\), the conditional measures \(\mu (\textrm{d}x_i | \bar{x}_i )\) satisfy LSI\((\rho _i)\) for any \(\bar{x}_i := (x_1,\ldots ,x_{i-1},x_{i+1},\ldots ,x_N)\in X_1\times \cdots \times X_{i-1} \times X_{i+1} \times \cdots \times X_N\).
-
For each \( i \ne j \in [N]\) there is a constant \(\kappa _{ij}>0\) such that
$$\begin{aligned} \left| \nabla _i \nabla _j H(x) \right| \le \kappa _{ij}, \qquad \forall x \in X. \end{aligned}$$Here, \(| \cdot |\) denotes the operator norm of a bilinear form.
-
Define a symmetric matrix \(A= (A_{ij})_{1 \le i, j \le N}\) by
$$\begin{aligned} A_{ij} = {\left\{ \begin{array}{ll} \rho _i, \qquad &{} i=j, \\ - \kappa _{ij}, \qquad &{} i \ne j. \end{array}\right. } \end{aligned}$$Assume that there is a constant \(\rho >0\) with
$$\begin{aligned} A \ge \rho {{\,\textrm{Id}\,}}, \end{aligned}$$in the sense of quadratic forms.
Then \(\mu \) satisfies LSI\((\rho )\).
Theorem C.5
(Two-Scale Criterion for LSI [9]). Let X and Y be Euclidean spaces. Consider a probability measure \(\mu \) on X defined by
Let \(P : X \rightarrow Y\) be a linear operator such that for some \(N \in \mathbb {N}\),
Define
Assume that
-
\(\kappa < \infty \).
-
There is \(\rho _1>0\) such that the conditional measure \(\mu (\textrm{d}x | Px=y)\) satisfies LSI\((\rho _1)\) for all \(y \in Y\).
-
There is \(\rho _2>0\) such that the marginal measure \(\bar{\mu }= P_{\#}\mu \) satisfies LSI\((\rho _2 N)\).
Then \(\mu \) satisfies LSI\((\rho )\), where
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Kwon, Y., Menz, G. & Nam, K. Hydrodynamic Limit of the Kawasaki Dynamics on the 1d-lattice with Strong, Finite-Range Interaction. Ann. Henri Poincaré 24, 2483–2536 (2023). https://doi.org/10.1007/s00023-023-01271-8
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00023-023-01271-8