Abstract
This is a detailed study of damped quintic wave equations with non-regular and non-autonomous external forces which are measures in time. In the 3D case with periodic boundary conditions, uniform energy-to- Strichartz estimates are established for the solutions, the existence of uniform attractors in a weak or strong topology in the energy phase space is proved, and their additional regularity is studied along with the possibility of representing them as the union of all complete bounded trajectories.
Bibliography: 45 titles.
Export citation and abstract BibTeX RIS
This research was partially supported by the Russian Science Foundation under grant no. 19-71-30004 (§§6–8 below) and by the Engineering and Physical Sciences Research Council under grant no. EP/P024920/1. |
§ 1. Introduction
In this paper we study the non-autonomous damped wave equation
in the bounded subdomain of with periodic boundary conditions. Here is the unknown function, is the Laplacian with respect to the variable , is a positive constant, is a given non-linearity which is assumed to be of quintic growth ( as ) and to satisfy certain natural conditions (stated in (4.2) below), and is a given external force which is an -valued measure of locally finite total variation and is assumed to be uniformly bounded on bounded time intervals: (see §2 for definitions of key function spaces).
Dispersive or/and dissipative semilinear wave equations of the form (1.1) model various oscillatory processes in many areas of modern mathematical physics including electrodynamics, quantum mechanics, non-linear elasticity, and so on, and constitute a topic of constant interest (see [26], [2], [40], [12], [36], [38], [35], and the references therein).
The basic property of these equations is the so-called energy equality
which can formally be obtained by multiplying (1.1) by and integrating with respect to and . Here
and . This identity motivates the natural choice of the energy phase space and the class of energy solutions (as solutions for which the energy functional is finite) and also makes it possible to control the energy norm of the solution. In particular, if the non-linearity has a sub-quintic or quintic growth rate, then due to the Sobolev embedding theorem , the energy space is given by . In the supercritical case where with , we need to take in order to guarantee the finiteness of the energy integral.
It is believed that the analytic properties and the dynamics as of solutions of the damped wave equations (1.1) depend strongly on the growth rate of the non-linearity as . Indeed, in the most studied case of a cubic () or sub-cubic () growth rate the above energy equality norm is sufficient to get the well-posedness of the boundary-value problem in the energy space and also the dissipativity and additional regularity of solutions, as well as to develop the corresponding attractor theory in both the autonomous and non-autonomous cases (see [1], [2], [12], [16], [25], [26], [29], [40], [43], and the references therein).
The case of super-cubic but sub-quintic growth rate () is a bit more complicated since the well-posedness of energy solutions is still an open problem (at least in the case of bounded domains). However, this problem can be overcome by using solutions slightly more regular than energy solutions, for example, solutions for which the mixed space-time norm is finite for any . These are the so-called Shatah–Struwe (or Strichartz) solutions, for which the corresponding initial boundary-value problem is known to be well-posed. The existence of such solutions is strongly based on the Strichartz estimates for the linear wave equation (see Theorem 2.1 below), which are now available not only for the whole space, but also for bounded domains with Dirichlet or Neumann boundary conditions (see [3], [6], [9], [10], [36], [37], [38]). Moreover, crucial for the attractor theory is the following energy-to-Strichartz estimate for such solutions:
where is a monotone increasing function which is independent of and the solution . In the sub-quintic case this estimate is a straightforward consequence of the linear Strichartz estimate and perturbation arguments. In turn, (1.3) enables us to establish the dissipativity of in the Strichartz norm based on standard energy estimates. Since finiteness of this norm is enough for uniqueness, the control obtained gives well-posedness, dissipativity, and the existence of global/uniform attractors as in the classical cubic case; see [14], [22], and [18] for the cases of , , and a bounded domain endowed with Dirichlet boundary conditions, respectively (see also [32] for the case of a damped wave equation with fractional damping).
In contrast to this, very little is known about the solutions of (1.1) in the case of supercritical (superquintic) growth rate of the non-linearity . In this case the situation is in some sense close to the 3D Navier–Stokes problem, namely, we have global existence of weak energy solutions, for which we do not know uniqueness, and local existence of more regular solutions, for which we do not know global existence. It is expected that smooth solutions may blow up in finite time even in the defocusing case, but to the best of our knowledge there are no such examples. In this case the existing attractor theory is related to multivalued semigroups or/and the so-called trajectory dynamical systems and trajectory attractors (see [12], [11], [29], [44], and the references therein).
We now turn to the most interesting borderline case of a critical quintic non- linearity , which is the main object of our study in this paper. In this case, the energy-to-Strichartz estimate (1.3) no longer follows from the Strichartz estimate for the linear equation (at least in a straightforward way), so the proof of global existence for Shatah–Struwe solutions is usually based on so-called non-concentration arguments and the Pohozhaev–Morawetz equality; see [3], [15], [19]–[21], [34], [33], [35], and [38] (see also [9] and [10] for the case of bounded domains with Dirichlet or Neumann boundary conditions). This approach enables us to construct a Shatah–Struwe solution such that the -norm is finite for all , but does not let us get any control of the norm in terms of the energy norm or verify that the Strichartz norm does not grow as . This is clearly not sufficient for constructing and studying attractors. Indeed, without a uniform control of the Strichartz norm as , this extra regularity may a priori be lost in the limit, and the attractor may contain solutions which are less regular than Shatah–Struwe solutions (and for which we do not have a uniqueness theorem). Thus, a uniform control of the Strichartz norm is crucial for constructing an attractor theory.
This problem was partly overcome in [18], where the asymptotic regularity for autonomous damped quintic wave equations in bounded domains of was established, along with the existence and regularity of global attractors. The method proposed there is heavily based on the existence of a global Lyapunov function and on the related convergence of the trajectories to the set of equilibria, and for this reason cannot be extended to the non-autonomous case. Moreover, to the best of our knowledge there have been no results so far on strong attractor theory for quintic wave equations in the non-autonomous case.
The main aim of our paper is to give a comprehensive study of non-autonomous quintic wave equations in the case of periodic boundary conditions. In order to do so, we first prove the energy-to-Strichartz estimate (1.3) for the Shatah–Struwe solutions of (1.1) for the quintic case as well. Therefore, the following theorem can be considered as our first main result.
Theorem 1.1. Let the non-linearity satisfy the assumptions (4.2), and let the external force belong to . Then the problem (1.1) is globally well-posed in the class of Shatah–Struwe solutions, any such solution satisfies the energy- to-Strichartz estimate (1.3), and the following dissipative estimate holds:
where the positive constant and the monotone increasing function are independent of , , and the solution .
The non-trivial part here is exactly to establish the energy-to-Strichartz estimate (1.3) (the rest is a standard consequence of this estimate and the classical dissipative energy estimate). To do so we start with the analogous energy-to-Strichartz estimate for the Shatah–Struwe solutions of the quintic wave equation in the whole space :
which was proved in [4] (see also [39] for the explicit expression of the function ), and we extend it to the non-autonomous case
This extension uses an approximation of the external force by sums of Dirac -measures and a representation of the solution for such external forces via solutions of the autonomous equation. This approach can be interpreted as an analogue of the Duhamel formula for the non-linear equation and is of independent interest. We would like to emphasise that this method requires one to consider measure-driven equations of the form (1.1) as an intermediate step, even if we want to verify the estimate (1.4) only for regular external forces (see §5 for details). This is one of the motivations for us to consider measure-driven damped wave equations. Of course, measure-driven equations are interesting and important in themselves; we mention here only that they are widely used in the theory of stochastic partial differential equations (see [24] and the references therein). Note also that no analogue of the energy-to-Strichartz estimate for the case of equation (1.5) in bounded domains (with Dirichlet or Neumann boundary conditions) is known so far, and this is the main reason for our choice of periodic boundary conditions.
We now turn to attractor theory. We first note that the dissipative estimate (1.4) implies in a standard way the existence of a uniform attractor for equation (1.1) in the weak topology of the energy space (see §6). However, new difficulties arise when we try to describe the uniform attractor in terms of bounded complete trajectories related to (1.1). We recall that, following the general theory developed in [12], [11], in order to obtain such a description we need to study not only equation (1.1), but also all its time shifts as well as their limits in the corresponding topology. In our case it is natural to take the closure of the set of all time shifts of the initial measure in the weak-star topology generated by the duality
where stands for the continuous functions with compact support. Namely, we introduce the group of time shifts via , and we define the hull of the given measure as follows:
(see §6 for more details). Then the general theory predicts the representation
where is the set of complete (defined for all ) bounded (in ) solutions of equation (1.1) with the right-hand side . Again, according to the general theory, this representation will hold if the solution operators (which map the initial data to the Shatah–Struwe solution of the problem (1.1) with right-hand side ) are weak-star continuous as maps from to .
Unfortunately, in contrast to the standard situations considered in [12], the map may be discontinuous in the case of measure-driven equations. As shown in §6, this may destroy (and destroys in concrete examples given there) the representation formula (1.6). More precisely, the attractor may become larger than the union of all bounded complete trajectories. In order to avoid this pathology, we found necessary and sufficient conditions for the measure which guarantee the continuity of the map . In particular, because of these restrictions the measures must have zero discrete part. For this reason we call these measures weakly uniformly non-atomic measures (see §6 for details). Thus, we have proved the following result.
Theorem 1.2. Let the assumptions of Theorem 1.1 hold, and let the measure be weakly uniformly non-atomic. Then the weak uniform attractor has the representation (1.6).
We would like to recall that the representation formula (1.6) is one of the key tools for the further study of the attractor (and is crucial for our study of the compactness of weak attractors in stronger topologies; see §7). Unfortunately, this formula fails for generic measures , which makes the theory constructed not entirely satisfactory. We expect that the problem can be resolved by using the trajectory approach, and we will return to this question in a forthcoming paper.
We would also like to mention that measure-driven equations appear naturally in attractor theory even if we start from a regular external force (the natural class of external forces from the point of view of Strichartz estimates). Indeed, we cannot guarantee in general that the hull is a subset of , and the appearance of Borel measures which are not absolutely continuous with respect to Lebesgue measure in the hull looks unavoidable. This is a second motivation for us to consider measure-driven damped wave equations from the very beginning.
As the next step, we study the existence of a uniform attractor for (1.1) in the strong topology of the energy space . Clearly, the assumption alone is not enough for this (see the examples given in [45]), so we need to impose some extra conditions on the measure to get this result. In this paper we introduce, following [45], two classes of right-hand sides, the so-called space-regular measures and the time-regular measures. Roughly speaking, these classes consist of the measures which can be approximated (in ) by measures which are smooth in space or in time, respectively; see Definition 7.1. The intersection of these classes coincides with the class of translation-compact external forces introduced in [12].
The following result is verified in §7.
Theorem 1.3. Let the assumptions of Theorem 1.2 hold, and in addition let the measure be space- or time-regular. Then there exists a uniform attractor for equation (1.1) in the strong topology of the energy space , and it coincides with the weak attractor constructed above.
As in [45], we utilise the energy equality and the so-called energy method (see also [5], [30]) to verify the asymptotic compactness.
Furthermore, we also verify that the uniform attractor is more smooth if the external forces are more smooth. As usual, in order to do so it is enough to verify that belongs to the higher energy space for some small positive . The additional regularity can be obtained using standard bootstrap arguments. To get this higher regularity, we follow mainly [43] and use the following consequence of the Kato–Ponce inequality:
which holds for (see §10). This enables us to prove the following result (in §8).
Theorem 1.4. Let the assumptions of Theorem 1.1 hold, and in addition let
for some . Then the attractor is a bounded set in the higher energy space . Moreover, the analogous result holds also if is sufficiently smooth in time.
Finally, for the convenience of the reader we collect in §9 some standard facts and concepts in the theory of vector-valued measures and related functions of bounded variation.
§ 2. Function spaces and preliminaries
In this section, we introduce some notation which will be used throughout the paper, and we state some classical results on solutions of linear wave equations. We start with function spaces.
Let be a domain of with a smooth boundary. As usual, the Lebesgue spaces of -integrable functions on are denoted by , . In the particular case we use the notation . For any , we denote by the classical Sobolev space of distributions whose derivatives up to order belong to . The closure of in the space is denoted by . In the case , we will write instead of in order to simplify the notation. The negative Sobolev spaces are defined as dual spaces:
In the case , , we define the fractional space , , to be the restriction of the Bessel potentials space to the domain . We recall that the norm in is defined by
where stands for the Fourier transform of (for instance, see [41] for more details). In particular, the fractional Laplacian gives an isomorphism between the spaces and :
Note that this formula remains true in the spatially periodic case, when . In the general case where is a bounded domain some restrictions appear due to the boundary conditions (see [41]).
Below we will also make wide use of the classical Sobolev embedding theorem
and the interpolation inequality
where , , , and
We will also need spaces of functions of mixed space-time regularity. For instance, the natural norms in the spaces and are given by
and
respectively. The index 'loc' or '' stands for the local or uniformly local topology, respectively. For instance,
and
Finally, to treat external forces, we need the space of vector measures with values in and with finite total variation, and the associated spaces of functions of bounded variation (see §9 for more details). Namely, denotes the locally convex space of -valued Borel measures on such that the restrictions of to every finite segment belong to . Similarly,
The spaces and are defined similarly.
We now recall standard results about the solutions of the linear wave equation
in the energy phase spaces
For simplicity, we state the results for the spatially periodic case, although most of the results stated below remain true in the case of general bounded domains as well.
Theorem 2.1. Let the initial data satisfy , let , and let for some . Then there is a unique solution of the problem (2.2). Also, the solution belongs to the space and the following estimate holds:
where the constant does not depend on and .
The proof of this theorem can be found in [3], [35], or [38], for example.
To conclude this section, we state the analogue of the above estimate for the damped linear wave equation
where . This estimate will be crucially used later on in order to obtain an additional regularity of uniform attractors.
Corollary 2.2. Let and for some . Then the solution of (2.4) satisfies the estimate
where the positive constants and are independent of , , and .
Proof. Indeed, due to the isomorphism (2.1) it is sufficient to verify (2.5) just for . For simplicity we also assume that . Multiplying (2.4) by , where is sufficiently small, and arguing in the standard way (see [12], for example), we arrive at the inequality
for some positive constants and . After that, we rewrite (2.4) in the form of the equation (2.2) with the right-hand side
and apply the estimate (2.3) for the Strichartz norm on the time interval , . This gives us the estimate
We claim that (2.7) implies the estimate (2.5) for the Strichartz norm. Indeed, we may assume without loss of generality that (if this condition is not satisfied, then we can always increase by a suitable to satisfy this assumption and put for ). In this case, using the concavity of the function and (2.7), we get that for some
Finally, replacing by , we get the desired estimate for the Strichartz norm and finish the proof of the corollary.
§ 3. Measure-driven damped wave equation: the linear case
In this section we consider the linear wave equation
on a three-dimensional torus , where the damping parameter is and, in contrast to the previous section, is a measure. All the results in this section are actually valid not only for the case of periodic boundary conditions but also for Dirichlet or Neumann boundary conditions when is a smooth bounded domain (although this result is not necessary for our purposes). We suppose here that
where is the space of -valued Borel vector measures on with finite total variation (see §9 for more details).
We start with the definition of an energy solution of (3.1) which is a bit more delicate since, in contrast to the usual case, the time derivative may have jumps produced by atoms of the measure .
Definition 3.1. A function such that (where ) is an energy solution of the problem (3.1) on if:
1) it satisfies the equation in the sense of distributions, that is, for any test function
2) is weakly left-continuous at every point as an -valued function;
3) the initial conditions are satisfied in the sense that
Remark 3.2. Since and , the function is weakly continuous as a function with values in , , so the initial data for is well defined. The situation with the derivative is a bit more delicate since it can be discontinuous. Namely, from Definition 3.1 we see that the distributional derivative satisfies
and this functional can clearly be extended by continuity to any . For this reason,
This, together with the fact that , implies that
Since any -function has left and right limits at every point (see §9), the function also has left and right limits and at any point (in the weak topology of ) as well as the limits and . Thus, assumption 2) of the definition makes sense, and the second part of the initial conditions 3) for is also well-defined. However, since is not dense in , the values and remain undefined (as well as the values of at the jump points).
In order to avoid this ambiguity and to be able to define the dynamical process associated with our problem (see §6), we choose a weakly left-continuous representative on from the equivalence class of by default. Then the value is also well defined, and the value is determined by the first part of the initial conditions 3).
Remark 3.3. Note that any energy solution of the problem (3.1) has the following property:
(in the case we just assume that ). Indeed, integrating by parts in (3.3) and using (9.22) to handle the most complicated term, which involves measures, we get that
where . Therefore, almost everywhere for some . Using now the assumption that is left-continuous, together with the obvious fact that is also left-continuous, and taking into account the initial data, we conclude that
The desired formula (3.5) is an immediate consequence of (3.7).
The formula proved shows, in particular, that the function will be weakly continuous as a function with values in if the measure is non-atomic. Moreover, multiplying (3.7) by , integrating with respect to , integrating back by parts, and using the initial conditions, we return to the distributional formulation (3.3). Thus, the identities (3.3) and (3.7) are equivalent and we may verify (3.7) instead of (3.3). We will make essential use of this observation later.
In the next step we write out an explicit formula for the solution of (3.1). We start with the homogeneous case . Then the solution is given by
where with periodic boundary conditions,
and . The corresponding solution semigroup in the energy phase space is then defined by
The following result is well known and can be verified by straightforward calculations.
Lemma 3.4. The operators are bounded in and satisfy the estimate
where the non-negative constants and can depend on , and if .
Furthermore, in the regular case where the measure is absolutely continuous ( for some ), the solution of the non-homogeneous equation is given by the Duhamel formula:
The next theorem shows that the analogue of this formula holds in the general case as well.
Theorem 3.5. Let and , and let the external force belong to . Then the problem (3.1) has a unique energy solution on . This solution is given by
where denotes the density of with respect to the measure (see (9.17)).
Furthermore, the following energy estimate holds:
for some constant depending only on .
Proof. We note first that due to Lemma 3.4 the function is well defined and belongs to the space and satisfies the energy inequality (3.12) (here we have implicitly used that ). The weak left-continuity of as well as the fact that it satisfies the initial data is also an immediate consequence of (3.11).
In order to check that it satisfies the equation in the sense of distributions, we expand into a Fourier series associated with the eigenfunctions of the operator . Namely, let and be the eigenvectors and eigenvalues (enumerated in non-decreasing order) of , respectively, and let be the orthogonal projection onto the linear subspace spanned by the first eigenvectors. Also, let . Then
and, by Lemma 9.6 and the estimate (3.12),
Thus, it is enough to verify that for every the function is a distributional solution of the ordinary differential equation
But this can be done in a straightforward way by using the integration by parts formula (9.22) (with ) and the properties of the Duhamel integral (we leave the rigorous proof of this to the reader). Thus, the function is indeed the desired energy solution.
Finally, let and be two energy solutions. Then since both these functions are weakly continuous in , and their derivatives are weakly left-continuous and have the same jumps according to (3.5), we conclude that is weakly continuous in , where . In addition, solves the homogeneous problem (3.1) with and zero initial data. It is well known that such a solution is unique, so and uniqueness is also verified.
Corollary 3.6. Let the assumptions of Theorem 3.5 hold. Then the energy solution is in and , where is the support of the discrete part of the measure , or equivalently, the set of points of discontinuity of the distribution function . Moreover, both the limits and exist for every in the strong topology of .
Indeed, this follows immediately from the analogous statement for the finite- dimensional part and from the uniform smallness of the function proved in the theorem.
Corollary 3.7. Assume that, in addition, the measure is non-atomic ( for all ). Then the solution is in . Moreover, the energy equality holds:
for all (here and below, is written instead of , since the integrals over and coincide for all non-atomic measures).
As usual, (3.14) is proved first for the finite-dimensional function , where it is standard since is continuous in time and therefore can be approximated by smooth functions. Then by passing to the limit as one gets the desired energy equality also for the infinite-dimensional case (using the fact that the are uniformly small).
Remark 3.8. The identity (3.14) can be rewritten in the form
In particular, the function is absolutely continuous in time. However, the energy itself is not necessarily absolutely continuous since the singular part of the measure is not assumed to vanish.
The analogue of this formula can be written in the general case, when the discrete part of does not vanish. However, in this case one should be careful with the definition of the integral , since the function makes jumps exactly at the points where is discontinuous. Moreover, since by (3.7) the function is continuous, the only problematic term is . This integral makes sense as a Lebesgue–Stiltjes integral. But the value of the integral thus defined is inconsistent with the energy identity. Indeed, in our case is left-continuous at the jump points and therefore
However, arguing a bit more carefully (for example, approximating by smooth functions or comparing the values of the energy functional before and after a jump), we see that the correct formula must be
which corresponds to the choice (see also [17]). This gives the following natural interpretation of the problematic integral:
which is consistent with the energy equality. We will return to this topic in a forthcoming paper.
We end this section by establishing Strichartz-type estimates for the measure- driven wave equation using approximations of the measure by absolutely continuous measures (functions in ).
Theorem 3.9. Let , let the initial data be in , and let the external force be in . Then the energy solution of the problem (3.1) has the estimate
where the constant depends on and but not on and .
Proof. Let be the distribution function of (see (9.5)). Let be smooth approximations of constructed as in Proposition 9.11 and consider the following sequence of approximations of :
We note that by construction (see Proposition 9.11) we have
In addition, for all (see Remark 9.12). Using the standard energy estimate and (3.19), we see that
The last estimate together with (3.18) implies that converges to some weak-star in as . We need to show that is an energy solution of the problem (3.1). Indeed, arguing in the standard way we see that strongly in , and therefore is weakly continuous in and .
To verify that is an energy solution, it is enough to take the pointwise limit in the equality
and get (3.7). Thus, is an energy solution of (3.1), and by uniqueness .
To obtain the desired Strichartz estimate, we apply Theorem 2.1 to equation (3.17) and get that
The last estimate allows us to assume without loss of generality that converges to weakly in as . Weak lower semicontinuity of the norm implies the desired estimate (3.16) and finishes the proof of the theorem.
Remark 3.10. Since the energy estimate lets us control the -norm of , we can replace the -norm on the left-hand side of (3.16) by any intermediate Strichartz norm, for instance, by the -norm.
§ 4. The quintic wave equation: well-posedness and dissipativity in the energy norm
In this section, we discuss the properties of solutions for our main object of study — the damped quintic wave equation
on the 3D torus . Since the results presented below are either well known or straightforward adaptations of well-known results to the case of external forces generated by measures, we restrict ourselves to giving only a brief exposition (more details can be found in [12], [18], [9], and [10]).
We assume that , , and the non-linearity has the following structure:
We start our exposition by giving the analogue of Definition 3.1 for an energy solution in the non-linear case.
Definition 4.1. A function such that (where ) is an energy solution of the problem (4.1) on an interval if:
1) it satisfies equation (4.1) in the sense of distributions, that is, for any test function ,
2) is weakly left-continuous at every point as an -valued function;
3) the initial conditions are satisfied in the sense that
As in the linear case (see Remarks 3.2 and 3.3), it can be shown that
and, in particular, the difference between two energy solutions of (4.1) (corresponding to the same ) belongs to . In addition, exactly as in the linear case, the equality (4.3) is equivalent to
The presence of the non-linear term does not make any difference here, due to the embedding theorem and the growth assumption .
The next step is a theorem on the solvability of the problem (4.1) in the class of energy solutions.
Theorem 4.2. Let and , and let the non-linearity satisfy (4.2). Then in the sense of the above definition for every there exists at least one energy solution which satisfies the estimate
where the monotone increasing function and the constant are independent of , , , and .
Proof. Indeed, let us start with the case where the measure is regular, that is, . In this case the assertion of the theorem is standard: the existence of a weak solution can be obtained, for example, using Galerkin approximations. Uniform energy estimates for the Galerkin approximations can be deduced just by multiplying the equation by for some positive , and the validity of the energy estimate for the solution of (4.1) is then established by passing to the limit in the Galerkin approximations (see [12] and the references therein for details).
Consider now the general case, when the measure can be singular. In this case, we approximate by regular measures using the special approximations constructed in Proposition 9.11 (see also Remark 9.12). Namely, the sequence is uniformly bounded in and weak-star convergent to the measure in , and the converge to for every . Let be an energy solution of (4.1), with replaced by . Then by the uniform energy estimate we may assume without loss of generality that
Due to the compactness of the embedding
we conclude that strongly in and therefore almost everywhere in . Moreover, since is uniformly bounded in , the convergence almost everywhere implies that
The convergence established lets us pass to the limit as in the equations (4.3) and prove that the limit function solves (4.1) in the sense of distributions.
Finally, in order to verify left-continuity (find a suitable representative in the equivalence class), it is sufficient to pass to the pointwise limit in equation (4.4) for the solutions , and the theorem is proved.
The existence of weak energy solutions can be proved similarly not just for quintic non-linearities. The only difference is that the energy space should be properly defined in the case of rapidly growing non-linearities. Namely, if the non-linearity grows like with , then one should take as the energy space (see [12] for details). However, to the best of our knowledge, the uniqueness of such a solution is not known for .
Moreover, for the quintic case we also do not know whether or not any energy solution satisfies the energy estimate (4.5). In order to overcome this problem, we introduce (following [4], [34], and [33]) the so-called Shatah–Struwe solutions and use the Strichartz estimates.
Definition 4.3. An energy solution is a Shatah–Struwe solution of the problem (4.1) if, in addition,
Note that, since
for any Shatah–Struwe solution we have .
The next theorem establishes the uniqueness of such solutions.
Theorem 4.4. Let and be two Shatah–Struwe solutions of the problem (4.1) which correspond to different initial data and the same measure . Then the following estimate holds:
where and the constant is independent of and . In particular, the Shatah–Struwe solution is unique.
Proof. Indeed, let . Then the function is weakly continuous in , since and make the same jumps determined by the discrete part . This function solves the equation
Since and , multiplication by can be justified in the standard way and gives
Moreover, using again the estimate , the Hölder inequality, and the embedding , we get that
and the Gronwall inequality finishes the proof of the theorem.
The next corollary is crucial for our proof of asymptotic compactness.
Corollary 4.5. Let the assumptions of Theorem 4.2 hold, and in addition let the measure be non-atomic (that is, for all ). Then for every Shatah– Struwe solution , the energy functional is a continuous -function of time and the following energy equality holds for all :
In particular, .
Proof. Indeed, since , the term can be treated as a regular measure. Thus, according to Corollary 3.7, we may write
Since , the term involving the non-linearity is well-defined. Moreover, arguing in the standard way, we get that the function is absolutely continuous and
Thus, the energy equality is proved. The fact that the energy functional is continuous follows immediately from this equality. Finally, the fact that follows from the energy equality in a straightforward way using the energy method.
We now discuss the existence of Shatah–Struwe solutions.
Proposition 4.6. Let the assumptions of Theorem 4.2 be satisfied. Then for any there exists a unique global Shatah–Struwe solution , and it satisfies the energy dissipative estimate (4.5).
The proof of existence is standard (see [35], [9], and [10] for details). First, based on the Strichartz estimate (3.16) for the linear equation and treatment of the non-linearity as a perturbation, one establishes local existence. Then by using the so-called Pohozhaev–Morawetz identity and non-concentration arguments one establishes that the Strichartz norm cannot blow up in finite time, and this gives global existence. The presence of the measure on the right-hand side does not produce any essential difficulties. We do not give a detailed proof here since in the next section we give an alternative proof and estimate the Strichartz norm without using non-concentration arguments.
§ 5. Quintic wave equation: energy-to-Strichartz estimates
As we have already mentioned, the global existence result for Shatah–Struwe solutions based on non-concentration arguments (and stated in Proposition 4.6) does not give any control of the Strichartz norm in terms of and the corresponding norms of the initial data and the external forces. In particular, we do not have any control of the behaviour of this norm as , which in turn leads to essential problems in attractor theory (see [18] for details). The aim of this section is to estimate this Strichartz norm in terms of the energy norm and a suitable norm of the external forces. Since we already know a dissipative estimate for the energy norm, this result will give us the desired dissipative estimate for the Strichartz norm. Our approach is crucially based on the following result for solutions of the homogeneous quintic wave equation in the whole of .
Proposition 5.1. There exists a monotone increasing function such that any Shatah–Struwe solution of the quintic wave equation
in the whole space has the estimate
The proof of this estimate can be found in [4] (see also [39] for an explicit expression for ).
Clearly, the estimate (5.2) on the whole line cannot hold in the case where is a bounded domain. However, its finite-time analogue remains true for .
Corollary 5.2. There exists a monotone increasing function such that any Shatah–Struwe solution of the quintic wave equation (5.1) with periodic boundary conditions has the estimate
Indeed, this estimate follows immediately from (5.2) and the result on a finite speed of propagation for wave equations (see [35]). To the best of our knowledge, the question of the validity of (5.3) in the case of general bounded domains remains open.
We are now ready to state the key result of this section.
Theorem 5.3. Let , let the non-linearity satisfy (4.2), and let the external force be in . Then the Shatah–Struwe solution of the problem
satisfies the estimate
where the monotone non-decreasing function is independent of the choice of the initial data and .
Proof. We suppose first that (that is, ). The general case will be considered later. To verify the desired estimate, we consider a sequence approximating the solution , where solves the problem (5.4) with the external force instead of , and the sequence of discrete measures is provided by Theorem 9.21:
and . Note that the solution should solve the homogeneous problem for and has jumps of the time derivative at finitely many points :
Thus, the existence and uniqueness of follows immediately from the analogous result for the homogeneous problem (5.1) and we need not use Proposition 4.6 here. Moreover, by Theorem 4.2 we have the uniform energy estimate
for some monotone increasing function . Hence by passing to a subsequence if necessary and using the fact that uniformly for all (due to the special choice of explained in Theorem 9.21), we may assume that is weak-star convergent to the weak energy solution of (5.4) (see the proof of Theorem 4.2). Thus, we only need to verify the uniform estimate of the Strichartz norms for the solutions . Then passing to the limit as will give us the desired estimate for as well.
Note that we can get the Strichartz estimate for the solution just by applying the estimate (5.3) on every time interval and using the fact that the energy norm is under control. However, this is not enough, since the resulting estimate will clearly depend on . So we need to proceed a bit more carefully.
Consider the approximations , , of the solution which solve (5.4) with the same initial data and with the external forces
Then on the one hand,
uniformly with respect to and by Theorem 4.2. On the other hand, it is clear that for all . Moreover,
and the functions and , , solve the linear homogeneous problem (5.1) with the initial data
In particular, by Corollary 5.2 and the estimate (5.8),
Finally, we introduce the functions and , . Then obviously,
and the functions , , satisfy the equations
Note also that for . To estimate the Strichartz norms of the functions , we use that
and therefore by the Hölder inequality and the Sobolev embedding we have
Multiplying (5.11) by and using (5.12), we now get that
and the Gronwall inequality together with the control (5.9) gives us that
We are now ready to apply the standard Strichartz estimate to the linear equation (5.11) and get that
Finally, according to (5.10),
Thus, the theorem is proved in the particular case .
We now consider the general case , which can be derived from the estimate obtained using more or less standard perturbation theory arguments. First we recall the following simple lemma, which can be verified using convexity arguments (see [42] and also [32]).
Lemma 5.4. Let be a monotone increasing function, let , and let . Then there exists a smooth monotone increasing function such that
where is determined by only.
We rewrite (5.4) in the form
and apply the already proved estimate (5.5) on the interval where will be determined later. This gives us the estimate
Since the function has a sub-quintic growth rate, the Hölder inequality gives us that
for some positive exponent . Inserting this estimate into the previous one and using Lemma 5.4, we get that
Important here is that the function is independent of . Fixing to be small enough, we derive from (5.17) that
for some new monotone increasing function . Since the energy norm of the solution is under control, we may apply this estimate on the intervals , , and so on. This gives us the desired control
with some monotone increasing function . Since the -norm of is controlled by the -norm of , we can get control of the -norm of using the Strichartz estimate for the linear equation. Thus, the theorem is proved.
As a consequence of Theorem 5.3, we obtain the desired dissipative Strichartz estimate for the solutions of the non-linear damped wave equation (1.1), which is crucial for what follows.
Corollary 5.5. Let the non-linearity satisfy (4.2), and let the external force be in . Then for any and any initial data the problem (1.1) has a unique Shatah–Struwe solution , and the following estimate holds: for ,
for some constant and some monotone increasing function which are independent of , , and .
Proof. This result easily follows if one applies the estimate (5.5) to the equation (1.1) on , treating as the right-hand side, and then combines the resulting estimate with the dissipative energy estimate (4.5) and the estimate in Lemma 5.4.
Remark 5.6. Since the -norm of the solution together with the energy norm allow us to control the -norm of the non-linearity , by applying the Strichartz estimates for the linear equation and treating as an external force we get a dissipative estimate for other Strichartz norms of , namely,
where and depends on but not on and .
Remark 5.7. We recall that the Strichartz estimates for non-homogeneous linear dispersive equations are usually derived from the homogeneous equations using duality arguments and the so-called Christ–Kiselev lemma (see [38] and the references therein). In contrast to this, the approach proposed in the proof of Theorem 5.3 works directly for non-linear (and even critically non-linear) problems and can be treated as a generalisation of the Christ–Kiselev lemma to the non-linear case. We believe that this approach will also be useful for other dispersive equations.
§ 6. Damped wave equation: weak uniform attractors
We start with basic definitions of non-autonomous dynamical systems (adapted to the measure-driven case). For a more detailed treatment and recent advances see [12] and [45]. Let us first recall the key definitions and concepts related to attractor theory, beginning with the autonomous case.
Definition 6.1. Let be a Hausdorff topological space, and let , , be a one-parameter semigroup on it. Also, let be a family of sets with the property that if and , then . The sets are said to be bounded.
A set is called an absorbing set for the semigroup if for any there exists a time such that
A set is called an attracting set for the semigroup if for every neighbourhood and every there exists a time such that
Finally, a set is a global attractor for the semigroup if:
1) is compact and bounded () in ;
2) is an attracting set for ;
3) is a minimal set with the properties 1) and 2).
Property 3) of the global attractor is usually formulated as strict invariance with respect to , but keeping in mind the non-autonomous case, we prefer to state it as minimality (see [12] for more details). To state the existence result for the autonomous case we need one more definition.
Definition 6.2. The semigroup is (sequentially) asymptotically compact on a set if, for any sequences and , the sequence is precompact in .
Proposition 6.3. Let the semigroup possess an absorbing set . Assume also that:
1) the topology induced on by the inclusion is metrizable and complete (that is, is a complete metric space);
2) the semigroup is asymptotically compact on .
Then has a global attractor .
In addition, if the operators are continuous on for every fixed , then the attractor is strictly invariant: , and is generated by all bounded trajectories defined for all :
where .
The proof of this proposition is standard and the details can be found in [12].
Since we are mainly interested in the non-autonomous equations, we recall below how the above concepts and results can be extended to the non-autonomous case. The first difference is that the solution operators form not a semigroup but a so-called dynamical process, which is a two-parameter family , , acting in the phase space and satisfying the relations
The operator is understood as a solution operator which maps the initial data at the time to the solution at the time .
Definition 6.4. Let be a Hausdorff topological space, and let , , be a dynamical process on it. Also, let be a family of sets such that if and , then . The sets are said to be bounded.
A set is called a uniformly absorbing set for the process if for any there exists a time such that
A set is called a uniformly attracting set for the process if for every neighbourhood and every , there exists a time such that
Finally, a set is called a uniform attractor for the process if:
- 1)is compact and bounded in ;
- 2)is a uniformly attracting set for ;
- 3)is a minimal set with the properties 1) and 2).
Below, will usually be a Banach space (or even a Hilbert space) endowed with either the strong or the weak topology. The associated uniform attractor will be referred to as a strong or a weak unform attractor, respectively. In both cases consists of all bounded sets in the Banach space under consideration.
Generalisation of the concept of asymptotic compactness to the non-autonomous case is also straightforward.
Definition 6.5. A process is uniformly asymptotically compact on a set if, for any sequences such that and any sequence , the sequence is precompact in .
The existence theorem is generalised in a similar way (see [12] for details).
Proposition 6.6. Let the process possess a uniformly absorbing set . Assume also that:
1) the topology induced on by the inclusion is metrizable and complete (that is, is a complete metric space);
2) the process is uniformly asymptotically compact on .
Then has a uniform attractor .
We now return to our damped wave equation (1.1). Since, according to Corollary 5.5, for any , , and this problem has a unique Shatah–Struwe solution , we can introduce a family of dynamical processes , , in the energy phase space. However, since in contrast to the usual case, the trajectories may have jumps, we should be a bit more careful in order to preserve the property (6.2). In particular, here we use our convention that the trajectories are left-continuous. Therefore, we may let
where . Then it is not difficult to see that the operators thus defined are indeed dynamical processes in the energy space , so we can study their uniform attractors. We fix as the family of all bounded (in the usual sense) subsets of our energy space (it is a Banach space, hence bounded sets are well defined). Then the estimate (5.18) guarantees the existence of a uniformly absorbing set. Moreover, it can be taken in the form of a ball
Recall that in this section we are mainly interested in weak uniform attractors, so we endow the space with the weak topology and denote the resulting locally convex space by . Since is a reflexive Banach space, the absorbing set is compact and metrizable in the weak topology of , hence all the assumptions of Proposition 6.6 are automatically satisfied, and we have proved the following result.
Theorem 6.7. Let the assumptions of Corollary 5.5 hold. Then for every the dynamical process has a uniform attractor in , which is called a weak uniform attractor for the equation (1.1).
As the next step we describe the extension of the key representation formula (6.1) to the case of uniform attractors. To this end we use (following [12]) the reduction of the dynamical process to a semigroup acting on an extended phase space. Namely, we introduce the group of shifts acting on the space of measures according to the formula
Then it is not difficult to verify that the above dynamical process satisfies the translation identity (that is, the cocycle property)
In order to fix a suitable topology on the space , we recall that is the dual space of , where means the space of continuous functions with compact support endowed with the inductive limit topology. Denote by the space endowed with the associated weak-star topology. Then by the Banach– Alaoglu theorem the unit ball of is precompact and metrizable in the topology of . We recall that in this topology if and only if
for any . We are now ready to define the hull of the measure to be the closure in the weak-star topology of the set of all shifts of :
where means the closure in . Obviously, the set endowed with the weak-star topology is a compact metric space and the group of shifts
acts continuously on .
Remark 6.8. Note that, in contrast to the usual case, the norm in the space is not lower semicontinuous in the induced weak-star topology, that is, there exist sequences bounded in such that weak-star in but
For this reason the unit ball in is not closed in . On the other hand, from (9.28) it is not difficult to prove that
which is enough for our purposes. In particular, the weak-star closure of the unit ball of in is a subset of the ball in of radius . Note also that the constant is sharp. Indeed, consider the sequence of measures . Then it is obvious that is weak-star convergent to , but
Now let be a family of dynamical processes associated with the damped wave equation (1.1) with right-hand sides . Then the extended phase space for the problem (1.1) is defined by
and the associated semigroup on acts as follows:
Indeed, the semigroup property for is an immediate consequence of the translation identity (6.5).
The key general idea is to relate the above uniform attractor for the dynamical process to the global attractor of the extended semigroup and, in particular, to describe the structure of using the representation (6.1) for the autonomous case. Namely, we endow the extended phase space with the topology induced by the embedding , and we fix bounded sets in as follows: is bounded if and only if is bounded in (here and below, is the projection onto the first component of the direct product ). Then by the estimate (5.18) and the inequality
the set
with defined by (6.4) is a compact metrizable absorbing set for the extended semigroup , and therefore by Proposition 6.3 the semigroup possesses a global attractor . The next theorem gives the desired description of the structure of the uniform attractor for the damped wave equation (1.1), under some extra continuity assumptions.
Theorem 6.9. Let the assumptions of Theorem 6.7 hold, and in addition let the maps be continuous (in the weak topology) as maps from to for all with . Then
and, moreover,
where is the so-called kernel of the process in the terminology of [12].
The proof of this result in a general setting can be found in [12].
Note that, in contrast to the usual case, the continuity assumption is not satisfied for general measures . Namely, the following result holds.
Proposition 6.10. Let the assumptions of Theorem 6.7 hold. Then the continuity assumption of Theorem 6.9 holds if and only if
for every sequence such that weak-star in and for every fixed .
Proof. Indeed, let (6.12) be satisfied. We need to prove that converges weakly to as in and in . Let be the corresponding Shatah–Struwe solutions. Then by the uniform dissipative estimate (5.18) we may assume without loss of generality that weak-star in . Thus, we only need to pass to the limit in (4.4). Namely, taking into account that , we see that this equality is
where . Obviously, the limit function satisfies (1.1) in the sense of distributions, and the passage to the limit in (6.13) is also straightforward due to the condition (6.12).
Let us now check necessity. We first check the necessity of the condition that for all . Indeed, let for some . Since the number of jumps is at most countable, we may assume that . Consider the sequence and let , where . Clearly, as and we may assume without loss of generality that weak-star in . Moreover, by the Helly selection theorem, we may also assume that weakly in for almost all . Let . Then two cases are a priori possible.
1) on a subset of of positive measure, and continuity obviously fails.
2) almost everywhere. In this case by passing to the limit in (6.13), say, in , we get that
and the continuity of also fails.
Thus, the necessity of the first condition is proved.
The necessity of the second condition can be proved similarly, but even more simply, since we need not shift measures and we can pass to the limit directly in (6.13).
This proposition reduces finding necessary and sufficient conditions for weak continuity of the dynamical process associated with (1.1) to verifying the conditions (6.12), which are purely measure theoretic and can be completely analysed. To formulate the definitive criterion we need the following definition.
Definition 6.11. A measure is weakly uniformly non-atomic if for every there exists a monotone increasing function such that
for all . The space of such measures is denoted by .
Then the following result holds.
Proposition 6.12. The conditions (6.12) are satisfied if and only if the initial measure is weakly uniformly non-atomic.
Proof. Assume that the conditions (6.12) hold and let be arbitrary. Consider the function defined by
Then in view of the first condition in (6.12) this function is continuous in for any fixed . On the other hand, by the second condition in (6.12) it is continuous in for any fixed . Thus, there is a point such that is jointly continuous at for any (in fact, there is a dense set of such points ; see [31] and the references therein, for example). Since is compact, we conclude that there exists a monotone increasing function such that
and . Finally, using the equalities
and , we deduce (6.15). Thus, the conditions (6.12) imply that is weakly uniformly non-atomic.
Now let be weakly uniformly non-atomic. Then it is not difficult to see using the Helly selection theorem (see Theorem 9.13 and Corollary 9.18) that
where the functions are the same as in (6.15). Then the first assumption of (6.12) is immediate, and the second is a standard corollary of the Arzelà theorem.
Thus, we have proved the following theorem, which can be considered as the main result of this section.
Theorem 6.13. Let the assumptions of Theorem 6.7 hold, and in addition let . Then the weak uniform attractor of equation (1.1) satisfies (6.10) and (6.11).
Indeed, this is an immediate consequence of Theorem 6.9 and Propositions 6.10 and 6.12.
We now give some examples clarifying the conditions imposed on the external force.
Example 6.14. We start with the case of regular measures , where . Then
Thus, (and is even strongly uniformly non-atomic) and the above theory works. Moreover, in this case
so that all measures in the hull are regular.
This will not be the case if we consider so-called normal external forces in , which were introduced in [27] to study uniform attractors for parabolic equations (see also [45] for more details). We recall that is normal if there is a monotone increasing function such that and
In this case we still have (and even ), and the theory works. However, in this case the hull may contain measures with non-zero singular part. According to the Dunford–Pettis theorem (see §9) the condition ensuring the embedding is a bit stronger:
where is any (Lebesgue) measurable set in and is its Lebesgue measure.
The condition (6.17) can be weakened as follows:
which still ensures that .
Example 6.15. We now give two more exotic examples clarifying the nature of weakly non-atomic measures. We start with a scalar measure . To this end we fix a non-negative smooth function supported on such that , and we consider the delta-like sequence . Finally, we introduce the function
Clearly, this function belongs to . It is also not difficult to show that the th term of this function averages to zero. Thus, in particular, and
as . On the other hand, the total variation of this measure is
and we see that the th term here tends to the -function at . In particular,
Thus, , so that the assumption (6.18) does not imply (6.17), and the class of measures is indeed larger than .
The next example is somehow complementary to the previous one and gives an alternative construction in the infinite-dimensional space. Namely, let be a Hilbert space and let be an orthonormal basis in it. Let
Then clearly , and its total variation is
Thus, taking any and using the fact that , we see that . However, its total variation clearly does not belong to this space.
Our last example shows the pathology which may appear in the case where the condition fails.
Example 6.16. Consider the first-order ordinary differential equation
An example for the hyperbolic equation can be obtained similarly by adding the term , but the construction becomes less transparent, so we prefer to deal with the first-order equation. In this case the uniform attractor can be found explicitly. Namely, the external force now is , and its hull is
Moreover, using the comparison principle, for example, it is not difficult to see that every complete trajectory corresponding to an external force with satisfies the conditions
and hence
Finally, in the case the equation is monotone, hence , and in the case we have the autonomous regular attractor . Therefore,
We now consider a perturbed version of equation (6.20):
where is a sufficiently large number and is the same as in the previous example. Then since as in the weak-star topology, the hull of the external force is similar to the non-perturbed case:
Then, using the fact that the term on the right-hand side of (6.21) just generates in the solution of (6.20) a spike of height close to centred near if is large enough, we see that
Thus,
On the other hand, if we take with large enough, then we get a trajectory which is close to but with spikes of height close to . This shows that
Remark 6.17. We recall that the representation formula (6.10) plays a fundamental role in the theory of non-autonomous attractors (see [12], for example), so the last example shows that the theory of uniform attractors for general measures which we have developed is not entirely satisfactory, and we really need the restriction to have a reasonable theory.
So far, the problem of constructing a satisfactory attractor theory for general measures remains open. The most natural and straightforward idea here is to endow the space with a different topology in which the operators would be continuous with respect to . But unfortunately this does not work even in the scalar case. Indeed, we actually need a topology on the space of measures that has two properties:
1) the unit ball in is sequentially compact in ;
2) convergence in implies pointwise convergence of the distribution functions for any .
But such a topology does not exist! Indeed, consider the sequence . It clearly converges to zero in the weak-star topology and does not converge to zero in (since does not converge to zero). Note that convergence in plus uniform boundedness of a sequence implies its weak-star convergence (by the Helly theorem). Thus, we should have a subsequence which converges in to zero, which is impossible since does not tend to zero. This shows that the problem is deeper than one might expect.
Alternatively, it seems to us that the problem can be solved by passing from a dynamical process on the initial phase space to a so-called trajectory dynamical system which acts on pieces of trajectories endowed with a suitable space-time topology (for example, the weak topology of with ; see [12] and the references therein). We return to this problem in a forthcoming paper.
§ 7. Asymptotic compactness and strong uniform attractors
In this section we would like to address the question of the existence of a strong uniform attractor for equation (1.1). By definition, this is a uniform attractor for the dynamical process associated with this equation and acting in the energy phase space endowed with the strong topology (see Definition 6.4). In this section we always assume that
and therefore the weak uniform attractor always exists and has the description (6.11) by Theorem 6.13. It is also not difficult to see that the strong uniform attractor, if it exists, coincides with the weak one:
Moreover, in view of Proposition 6.6 the existence of a strong uniform attractor follows from the asymptotic compactness of the process . In fact, it is more convenient for us to verify instead the asymptotic compactness of the extended semigroup acting on the space , where the space is endowed with the strong topology (and remains endowed with the weak-star topology). Namely, we will verify that for any sequence such that and any sequences and , the sequence
is precompact in . By the translation identity, this implies the asymptotic compactness of the process . Actually, since under our conditions the extended semigroup is weakly continuous on for every fixed , one can prove that the asymptotic compactness of the semigroup and that of the process are equivalent, but we will not use this fact below.
Clearly, the assumption alone is not enough to get strong asymptotic compactness (see examples in [45]). In particular, as shown in [45], is also not enough for compactness even in the case of a linear damped wave equation. In order to state our extra assumptions on , we use the following classes of external forces introduced in [45].
Definition 7.1. Let . The measure is said to be space-regular if there exists a sequence such that
Similarly, is said to be time-regular if there exists a sequence such that (7.4) holds (here and below we identify a measure which is absolutely continuous with respect to Lebesgue measure with its density).
The following proposition gives the key property of the classes of measures introduced.
Proposition 7.2. Let be space-regular. Then for any and any there exists a such that
Moreover, every measure in is space-regular, and for every there exists a such that
Similarly, let be time-regular. Then for any and any there exists a such that (7.5) holds. Moreover, every measure in is time-regular and for every there exists a such that (7.6) holds.
The proof of this proposition is straightforward and is given in [45].
Remark 7.3. More details on the properties of space- or time-regular functions can be found in [45]. For instance, any time-regular measure belongs to (this follows, for example, from the Dunford–Pettis theorem; see Theorem 9.20). In contrast to this, space-regular measures may have discrete and singular components. It is also known that is simultaneously space- and time-regular if and only if it is translation-compact in .
The typical examples of space- or time-regular measures are or , , respectively. A typical example of a measure that is not space-regular is
where is an orthonormal basis in , say, generated by the Laplacian, and is the characteristic function of a set . An example of a measure that is not time-regular is even simpler: . Combining these two examples, we get a measure
that is neither space- nor time-regular. Nevertheless, , and as elementary calculations show, this inclusion ensures strong asymptotic compactness due to time averaging. Thus, the conditions introduced are not necessary for asymptotic compactness. Unfortunately, necessary and sufficient conditions are not yet known.
We are now ready to state and prove the main result of this section.
Theorem 7.4. Let the assumptions of Theorem 6.13 hold, and in addition let the external force be time-regular or space-regular. Then the dynamical processes associated with the problem (1.1) have a strong uniform attractor , which coincides with the weak attractor constructed above and admits the representations (6.10) and (6.11).
Proof. As explained before, we only need to verify the asymptotic compactness of the associated process in the strong topology of . To this end it is sufficient to verify the precompactness of the sequence (7.3), where , the are taken in a uniformly absorbing set , and . To verify this we will use the so-called energy method (see [5], [30]), which is based on the following elementary property: if in the Hilbert space and , then strongly. The proof is divided into two natural steps.
Step 1. In this step we use the weak continuity of the processes and the existence of a weak uniform attractor to obtain a good description of weak limit points of the sequence (7.3). The arguments given below actually re-prove the general representation formula (6.11) in the case of equation (1.1). Nevertheless, we decided to give these arguments here since they are crucial for our proof of asymptotic compactness.
Without loss of generality we may assume that (in the associated weak-star topology). Let us also introduce the solutions which correspond to this sequence:
Then by the dissipative estimate (5.18), the inequality (6.9), and the fact that the are uniformly bounded, the sequence satisfies the estimate
In particular, the sequence (7.3) is bounded, so that by passing to a subsequence if necessary we may assume that
for some . Moreover, without loss of generality, we may also assume that
for some function such that and . Passing to the limit as in the sense of distributions in the equations (1.1) for , we get in a standard way (for example, see [18] for details) that is a complete bounded solution of (1.1) with right-hand side , and since , the function has no jumps, so that is a Shatah–Struwe solution of (1.1). Therefore,
We need to check now that . To this end we establish some results on strong convergence of which will be essential in Step 2 below. First we note that is bounded in and is bounded in , so by compactness arguments,
The analogous result for is a bit more delicate since, in contrast to the standard case, the are not functions but measures. To overcome this problem, we deduce from (6.13) that
where we have implicitly used that is bounded in and that
Moreover, since , there exists a monotone increasing function such that and
Thus,
and the functions are equicontinuous as functions with values in . Since they are also uniformly bounded as functions with values in , the Arzelà theorem gives us that
Thus, strongly in and, in particular,
Step 2. In this step we verify that by passing to the limit in the corresponding energy equality. Crucial for this method is the fact that if , then any Shatah–Struwe solution of equation (1.1) satisfies the energy equality (see Corollary 4.5). Thus, the validity of taking the scalar product of the equation (1.1) with is justified, and testing the equation with does not require any extra justification. For this reason we may multiply the equation (1.1) for the solution by (following [18]), where is sufficiently small, to get that
where
and
Multiplying (7.16) by and integrating the resulting identity with respect to time from to , we get the energy identity in the integral form
where, to avoid dependence on in the lower limit of integration, we set for .
We want to pass to the limit as in (7.19). To this end, we first note that the weak convergence in and the compactness of the embedding imply that
In order to pass to the limit in the terms containing the non-linearity, we recall that has a positive coefficient in front of the leading quintic term (see (4.2)). Therefore,
for some . Moreover, the strong convergence in implies convergence almost everywhere (passing to a subsequence if necessary). This allows us to apply the Fatou lemma and get
Similarly, using the strong convergence in and the boundedness of in , we arrive at the inequality
Next, for sufficiently small the quadratic form is positive-definite and hence is convex and weakly lower semicontinuous. Therefore,
Let us now look at the right-hand side of (7.19). Since the are bounded in by assumption and tends to , the first term on the right-hand side of (7.19) vanishes.
Moreover, since the and are bounded in and , respectively, and strongly in , we have
Here we have also used the fact that weak-star in as well as that (in order to guarantee that weak-star in ).
Up to this point, we have nowhere used that is time- or space-regular. This will be essentially used to pass to the limit in the second term on the right-hand side of (7.19), namely, to show that
We assume for the moment that (7.26) has been verified and complete the proof of the theorem. Indeed, passing to a subsequence if necessary, we may assume that
Then by taking on both sides of (7.19) and using the inequalities obtained above together with the fact that
we arrive at the estimate
On the other hand, since is a Shatah–Struwe solution of the limit problem, it also obeys the energy equality
Combining (7.27) and (7.28) with the weak lower semicontinuity of the norm , we get the chain of inequalities
which implies the equality
Together with the already proved weak convergence , this proves strong convergence. Thus, to finish the proof of theorem, we only need to verify the identity (7.26). This is done in the following lemma.
Lemma 7.5. Let be a measure which is either time- or space- regular. Assume also that the sequence of functions is uniformly bounded and that strongly in . Then the equality (7.26) holds for every sequence such that weak-star in .
Proof. Let be time-regular. Then according to Proposition 7.2, for any there exist and measures such that
Moreover, since the hull is compact in the weak topology of , we may also assume that weakly in . In particular,
Since the functions are bounded in , we have
Thus, we only need to prove that
To verify this we use the fact that is smooth in time and that strongly in , so we may integrate by parts and get that
This proves the lemma in the case where is time-regular.
Assume now that is space-regular. Then as in the time-regular case we may approximate the measure by measures and fix a sequence such that (7.31) and (7.32) hold. As before, the desired convergence will be proved if we verify (7.33). However, since we do not assume that , this convergence may fail, and we need to proceed more carefully. Namely, let be a small number and let
Since weak-star in and strongly in , for any we have
Thus, to prove convergence we need to estimate
The first term on the right-hand side tends to zero as , and by (7.31) the second term satisfies the inequality
where the constant is independent of . Thus, we only need to prove that
uniformly with respect to all . Moreover, since is non-atomic, the function is continuous as a function with values in , and we only need to prove that
Finally, integration by parts together with the fact that gives us that
Thus, the convergence (7.38) is verified and the lemma is proved, together with Theorem 7.4.
§ 8. Smoothness of uniform attractors
The aim of this section is to verify that the uniform attractor of the damped wave equation (1.1) is more regular if the external force is more regular. We consider two model cases of additional regularity for , namely,
and
for some (small) positive . The main result of this section is the following theorem.
Theorem 8.1. Let the assumptions of Corollary 5.5 hold, and in addition let the measure satisfy (8.1) or (8.2). Then the dynamical process associated with equation (1.1) has a strong uniform attractor in the phase space (which coincides with the weak uniform attractor constructed in Theorem 6.7), and this attractor is bounded in the space for some small :
Remark 8.2. Note that (8.1) together with the assumption that implies that is a function of bounded variation with values in :
In particular, and therefore the uniform attractor has the representations (6.10) and (6.11). In contrast to this, when (8.2) holds, the measure may contain a non-zero discrete part and (6.11) is not necessarily satisfied.
To prove the theorem, we split the solution into three parts
where solves the linear wave equation
the function solves the auxiliary non-linear problem
where is a sufficiently large number, and the remainder solves the following problem with zero initial conditions:
We need to obtain good estimates for each of the three functions , , and . We start with the simplest case of , which satisfies the linear equation.
Lemma 8.3. Let the above conditions hold and let satisfy either (8.1) or (8.2). Then the solution of (8.5) satisfies the estimate
where is sufficiently small and the symbol denotes the space (if (8.1) holds) or (if (8.2) holds).
Proof. Indeed, in the case of (8.2) the estimate (8.8) is an immediate consequence of Theorem 3.5 and the estimate (3.16) applied to the function and also of elliptic regularity.
Now let (8.1) be satisfied. Then by differentiating (8.5) with respect to time and writing , we get that
Since is well defined and , we can apply Theorem 3.5 and the estimate (3.16) to this equation and get that
Now since , we get that the function is bounded in . Finally, using the fact that , we see that (8.8) is satisfied at least for . Of course, the restriction is artificial and can be easily removed, but the validity of (8.8) for some small positive is enough for our purposes.
In the next step we show that the function decays exponentially as .
Lemma 8.4. Let the above assumptions hold. Then the solution satisfies the estimate
where the positive constant and the monotone increasing function are independent of , , and .
Proof. Multiplying (8.6) by , where is sufficiently small, and arguing in the standard way, we obtain an analogue of the identity (7.16), where and the non-linearity is replaced by . Since satisfies (4.2), one can easily verify that for sufficiently large
and (7.16) implies that
Applying the Gronwall inequality and using the fact that
we end up with the desired estimate for the energy norm of . To get control of the Strichartz norm, we apply the energy-to-Strichartz estimate (5.5) to (8.6) and get that
Next we again use the fact that , which together with the fact that has at most quintic growth rate gives us the control
Finally, treating the term as an external force and applying the Strichartz estimate to the linear equation obtained, we arrive at the desired decaying Strichartz estimate for and thereby finish the proof of the lemma.
We are now ready to treat the most complicated -component of the solution . We do this in two steps: in the first step we get an estimate growing exponentially in time, which will be refined in the second step.
Lemma 8.5. Let the above assumptions hold. Then the solution satisfies the estimate
where is a sufficiently small positive exponent, and the monotone increasing functions and are independent of and , and of the concrete choice of and .
Proof. Treating the non-linearity in (8.7) as an external force and applying the -energy and Strichartz estimates to this linear equation, we get that
where is a sufficiently small number (see (2.5)). To estimate the non-linear term we use the key inequality (10.11), which gives
and
Using the Hölder inequality together with the control (8.8) for the -component, we get that
where and
Estimation of the three other terms in (8.12) containing the - and -norms of is analogous but even simpler due to the control (8.8). According to the estimates already obtained, we have
and substituting (8.13) into the right-hand side of (8.11), we arrive at the inequality
where the constant depends only on and . Introducing the function
and raising both sides of (8.15) to the power , we get finally that
for some new constant depending on and . The Gronwall inequality applied to this estimate, together with (8.14), gives the desired estimate (8.10) and finishes the proof of the lemma.
We now state (following [43]) a corollary of the estimates obtained that is crucial for what follows.
Corollary 8.6. Let the above assumptions hold and let , where is a uniformly absorbing set for the equation (1.1). Then for any there exists a splitting of the solution of (8.7) such that
and
where the constant depends only on (and is independent of , , , and ). Moreover,
where is independent also of .
Proof. Note that by the estimates (8.9) and (8.8) it is sufficient to construct the desired splitting for only. To do this we fix a large (actually ) and construct the splitting (8.4) at the points , , , and so on. Namely, let , , and denote the solutions of the problems (8.5), (8.6), and (8.7), respectively, where the initial moment of time is replaced by , and define
Then as elementary calculations based on (8.9) show, the function satisfies (8.17) and (8.19). In turn, the estimates (8.8) and (8.10), together with the dissipative estimate for the solution , guarantee that the function satisfies (8.18) and (8.19). Finally, to obtain the desired splitting of , we just need to take
and the corollary is proved.
We are now ready to refine Lemma 8.5.
Lemma 8.7. Let the above assumptions hold and let . Then the solution of the problem (8.7) satisfies the estimate
where the constant is independent of , , and .
Proof. We refine the estimates (8.12) and (8.13) using the result of Corollary 8.6. To this end, we first note that we may assume without loss of generality that . Indeed, the extra term is easily controllable by the estimates obtained above in the energy space . Next, we write the difference as follows
The first term on the right-hand side of (8.22) is controlled exactly as in (8.12):
The third term is estimated similarly, using (10.4):
Thus, in view of the estimates (8.18), (8.8), (8.9), and (8.19) we have
So we only need to estimate the second term on the right-hand side of (8.22). To this end we use the assumption that and apply (10.10) to get that
Arguing as in (8.12), we then get that
where
It is important that the constant here is independent of . Therefore, due to (8.17) and (8.9) we have
and substituting the resulting estimates into (8.11), we finally get that the function satisfies a refined analogue of (8.16):
To derive the desired estimate (8.21) from (8.29), we need the following version of the Gronwall lemma.
Lemma 8.8. Let the function satisfy the estimate
for some constants and and a non-negative function . Then
The proof of this lemma follows almost word for word the proof of the usual Gronwall lemma and is thus omitted.
Applying the estimate (8.30) to (8.29) and using (8.28) (with the parameter fixed such that ), we derive the desired estimate (8.21) and thereby finish the proof of Lemma 8.7.
Now we are ready to complete the proof of the main theorem.
Proof of Theorem 8.1. According to the estimates (8.8), (8.9), and (8.21), the set
is a compact uniformly attracting set for the process in if is large enough. Thus, the process is uniformly asymptotically compact and possesses a uniform attractor in the strong topology of . Moreover, . This completes the proof of Theorem 8.1.
The next corollary gives the global well-posedness and dissipativity of the process in the higher energy space .
Corollary 8.9. Let the assumptions of Theorem 8.1 hold, and in addition let . Then the solution of (1.1) satisfies for all and the following estimate holds:
where the constant and the monotone increasing function are independent of , , , and .
Indeed, the proof of this estimate is based on the result of Corollary 8.6 and can be obtained as in the derivation of (8.21) (and even more simply, since we may take and pose the non-zero initial conditions directly for the -component). For this reason we leave the detailed proof of this corollary to the reader.
Remark 8.10. Note that Theorem 8.1 and Corollary 8.9 are formally proved only under the assumption that , and we do not know how to obtain more regularity of the attractor in one step even in the case where and are smooth. For instance, it would be interesting to get -regularity without using fractional Sobolev spaces. The problem is related to the restriction on the exponent in the key lemma, Lemma 10.2. However, higher regularity can be easily obtained in several steps using standard bootstrap arguments. Moreover, the most difficult step is exactly the first: to obtain the regularity of solutions for small positive . Since the non-linearity is no longer critical in , for further iterations one can use a standard linear decomposition to further improve the regularity. Namely, in (8.4) we may take and . For instance, if , we have
and therefore we need only one extra step of iterations to get the -regularity of the attractor.
§ 9. Appendix 1. -functions and vector measures
In this appendix we recall a number of more or less standard results concerning functions of bounded variation (-functions) with values in Banach spaces and the associated measures which are used throughout the paper. We restrict ourselves to the case where these functions take values in a separable Hilbert space (see [7], [8], [13], [28], and the references therein for more details).
Definition 9.1. A function is a function of bounded variation (a -function) if
where the supremum is taken over all finite and all partitions of the interval .
We also recall the elementary properties of the variation introduced:
- 1)if ;
- 2);
- 3);
- 4)the function is monotone increasing and ;
- 5)the function is continuous (left/right-continuous) at if and only if the same is true for .
Note that if is a -function, then is a scalar non-decreasing and bounded function, so it is continuous away from an at most countable set of points (indeed, for every the number of jumps of which are larger than must be finite). Therefore, property 5) guarantees that also has an at most countable number of discontinuities. Furthermore, due to monotonicity, the right/left limits and exist for all , and therefore simple arguments show that the right/left limits of also exist at every point in .
Definition 9.2. We denote by the Banach space of all -functions on with values in such that and is left-continuous at every interior point of (that is, may have a jump at ). The norm in this space is given by
As usual, every defines a vector-valued measure on the semiring generated by the subintervals of via the formulae
where , and we use the notation , and , . As usual, the assumption of left-continuity implies the -additivity of on the algebra generated by subintervals. Moreover, this measure can be extended in a unique way to the -algebra of Borel subsets of and gives a -additive Borel vector measure of finite total variation . We recall that for any Borel set the total variation is
where the supremum is taken over all countable disjoint Borel partitions of . It is also known that is a scalar positive -additive measure generated by the distribution function , that is, the formulae (9.2) remain valid if one replaces by on the left and by on the right. In particular, we have
Conversely, with any -additive -valued measure defined on the Borel -algebra of with finite total variation we can associate a -function in by the formula
Thus, there is an isomorphism between the space of Borel measures with bounded total variation endowed with the norm
and the space .
Furthermore, for every and every -measurable function such that , the Lebesgue integral is well defined. As usual, it is defined first for simple functions
where and the -measurable sets form a disjoint partition of , via the formula
Then it can be extended to any integrable function in the standard way by continuity (see [28] for details).
On the other hand, for every , we can also consider the Riemann– Stieltjes integral as the limit of the Riemann integral sums
where the limit is taken over all partitions and points , and where (the first term in the Riemann sum is suitably modified in order to preserve the additivity of the integral). It is well known that the Riemann–Stieltjes integral exists for at least every continuous function , and when it exists, it coincides with the Lebesgue (Lebesgue–Stieltjes) integral:
In addition, by the additivity of the Lebesgue integral, for every closed interval we have
at least for continuous functions .
We now recall that, by the standard properties of the Lebesgue integral, we have the inequality
In particular, for any
Thus, for any the linear functional
is a bounded linear functional on and . As in the scalar case (see [23]), the following vector version of the Riesz representation theorem holds (see [13] for details).
Theorem 9.3. Let be a separable Hilbert space and let . Then for any continuous linear functional there exists a function such that
and . In other words,
We now recall the concept of absolute continuity and the related Radon–Nikodým theorem for vector measures. For simplicity, we will consider only the case of a Hilbert space, where the Radon–Nikodým property is always satisfied.
Definition 9.4. For a Hilbert space and an interval , a measure is said to be absolutely continuous with respect to a scalar Borel measure with if the scalar measure given by (9.3) is absolutely continuous with respect to . The latter means that for every Borel set such that . We say that is absolutely continuous () if it is absolutely continuous with respect to Lebesgue measure on .
The vector-valued analogue of the Radon–Nikodým theorem is then as follows (for example, see [28]).
Theorem 9.5 (Radon–Nikodým). Let be a separable Hilbert space, let , and let . Then the measure is absolutely continuous with respect to a non-negative Borel measure if and only if there exists a function such that
for any Borel set , where the integral on the right-hand side is understood as a Bochner integral. Furthermore,
and, at least for continuous functions ,
for any Borel set .
There are two particular cases of this theorem which are of special interest for us. The first is when . Clearly, every measure is absolutely continuous with respect to . Therefore, in the case of a separable Hilbert space we can apply the Radon–Nikodým Theorem and conclude that there exists a function such that
Moreover, from (9.15) we see that
The above formulae allow us to express vector-valued measures in terms of scalar measures and integrable functions. In particular, from (9.17) we easily derive the following approximation result, which is crucial for our study of measure-driven partial differential equations.
Lemma 9.6. Let be a separable Hilbert space with an orthonormal basis , and let and . Let be the orthogonal projection onto the subspace spanned by the first basis vectors , and let . Then
Proof. Indeed, applying the projection to (9.17) and using (9.15), we find that
by the Lebesgue dominated convergence theorem.
The second, more standard, application of the Radon–Nikodým theorem is when is Lebesgue measure on . In this case absolutely continuous measures can be characterised via the analogous property of the corresponding distribution functions.
Definition 9.7. A function is said to be absolutely continuous, , if and only if for any there exists a such that for any finite sequence of pairwise disjoint sub-intervals of such that
we have
Then it is not difficult to see that if and only if its distribution function is in . We recall that by definition
As in the scalar case, the Radon–Nikodým theorem implies the following standard result.
Theorem 9.8. A function is absolutely continuous if and only if there exists a density such that
Furthermore, for any written in the form (9.20)
Remark 9.9. By the properties of the Bochner integral, every is differentiable almost everywhere on and , where is from (9.20). Moreover, this pointwise derivative coincides with the distributional derivative of .
Recall also that the pointwise derivative exists for any - function with values in a separable Hilbert space, but the analogue of the Newton–Leibniz formula (9.20) holds if and only if is absolutely continuous. Exactly as in the scalar case, any -function can be uniquely decomposed into three parts:
where is the discrete part (a step function), is the singular part (continuous but satisfying almost everywhere), and is the absolutely continuous part, which satisfies the Newton–Leibniz formula
and the analogous decomposition holds for the associated measures.
Remark 9.10. Recall also the standard integration by parts formula
which holds for every , and every subinterval . Of course, if , then should be replaced by . Actually, this formula remains true (after replacing the second term on the left-hand side by ) if the function is a -function with zero discrete part (). However, this is not straightforward when both and have non-zero discrete parts, since extra care is required to properly define the integral of a step function with respect to a Dirac measure (for example, see [17]). It should also be noted that by (9.22) the distributional derivative of the function is exactly the measure :
We now discuss the relation between the spaces and . For any we define the distribution function
Then it is obvious that is absolutely continuous, and thus the associated measure is also absolutely continuous. Hence, by the Radon–Nikodým theorem
for any Borel set . In particular,
Thus, the map is an isometric embedding of the space in the space , and the range of this linear operator is exactly the space of absolutely continuous measures. This lets us identify integrable functions with regular (absolutely continuous) measures.
The advantage of this embedding is that is dual to the separable Banach space , and hence its unit ball is weak-star compact. Thus, due to this embedding, the unit ball becomes weak-star precompact and (since this topology is metrizable on the unit ball) we can naturally identify weak-star limit points of bounded sequences in with vector measures of finite total variation. Moreover, the following statement holds.
Proposition 9.11. Let be a separable Hilbert space, let , and let be the unit ball in the space endowed with the total variation norm . Then
where means the closure of in the weak-star topology.
Proof. Let . We approximate the distribution function using the standard mollification procedure , where the positive kernels approximate the -function and
Then obviously and
Thus, , and since these measures are smooth, they belong to . Moreover, without loss of generality we may assume that is left/right-continuous at the endpoints and (otherwise we subtract the corresponding endpoint -measures and approximate them separately using one-sided approximating sequences of kernels).
Let be arbitrary. We need to prove that
Since the variations of the are uniformly bounded with respect to , it is enough to verify convergence for . In this case we can integrate by parts to get that
Then from the construction of it is easy to see that tends to at all points of continuity of . This fact implies the convergence of to , and by the Lebesgue dominated convergence theorem we also have the convergence to of the integral on the right-hand side.
Remark 9.12. Without loss of generality we may assume also that for all (including the jump points). Indeed, using the fact that the are continuous, and choosing the left-sided kernels (that is, such that ), we get the convergence for all . Thus, we may assume that for all .
For the next step we recall the characterisation of weak-star convergence in the case of scalar signed measures (see Proposition 8.1.8 in [8] 1), which is usually referred to as the Helly selection theorem.
Theorem 9.13. A sequence of scalar signed measures on a segment converges weak-star in to a measure if and only if:
1) ;
2) every subsequence in the sequence of distribution functions of the measures contains a further subsequence convergent to everywhere except on an at most countable set depending on the subsequence .
In the case when the measures are non-negative the second condition can be changed to:
) the whole sequence converges to the function at the points of continuity of .
Remark 9.14. We want to emphasise that, in contrast to the case of non-negative measures, weak-star convergence of signed measures does not imply pointwise convergence of the corresponding distribution functions on a dense set. In this connection we mention an example in [8]. On the interval we consider the sequence of measures , where the sequence of segments is formed by renumbering the segments , where for each . It is easy to see that converges to weak-star in , but the distribution functions do not converge to at any point in the open interval . Thus, the operator , regarded as a map from with the weak-star topology to , is not (sequentially) continuous for any fixed . For this reason the solution operator even of the simplest equation
is not continuous with respect to weak-star convergence of measures. This lack of continuity makes the corresponding attractor theory essentially more delicate.
The next theorem gives the analogue of the Helly selection theorem for vector measures, along with some further useful properties of weak-star convergence in .
Theorem 9.15. Let be a separable Hilbert space, let , and let be a sequence of vector measures. Also, let be the corresponding distribution functions. Then the following assertions are valid.
1. The sequence is weak-star convergent in to a measure if and only if it is bounded,
and every subsequence of the sequence contains a further subsequence which is weakly convergent in to the distribution function of the limit measure at all points in , with the exception of an at most countable subset depending on the choice of the subsequence .
2. Let be weak-star convergent to . Then for any segment the following inequality holds:
This inequality also holds when and are replaced by , or when and are replaced by .
Proof. 1. It is not difficult to see that weak-star convergence of the vector measures is equivalent to weak-star convergence of the scalar signed measures for any fixed , where
From the definitions of and , and from (9.5) we also see that
Then the first assertion is a standard corollary of Theorem 9.13 applied to the measures and of the fact that is separable.
2. To prove the second assertion, we first note that
for any . Together with the fact that the limit distribution is continuous everywhere except on an at most countable set, this inequality shows that it is sufficient to consider the case where and are points of continuity of the limit function .
Fix and fix a continuous function with norm on such that
Then by the continuity of at and we can extend to a continuous function on (which we also denote by ) without increasing its norm and in such a way that
where is sufficiently small and
Thus,
and by passing to the limit as , we get the desired inequality.
The case when and can be considered similarly.
Remark 9.16. Note that in general the weak-star convergence in does not imply weak-star convergence of the in if is a proper subinterval of . For this reason, the naive estimate
may fail and hence the second in (9.28) is essential. On the other hand, the sequence is bounded (and thus precompact) in . Therefore, by passing to a subsequence we may assume that weak-star in . However, even in this case we cannot get that . Instead, we can only prove that
for some depending of the choice of a subsequence. For instance, the sequence is weak-star convergent to in this space. Let . Then the restrictions in vanish identically, and hence . Consequently, .
The example constructed shows that the restriction operator of a measure to a proper subinterval is not continuous in general in the weak-star topology.
We now introduce the class of so-called uniformly non-atomic sets of measures, which allow us to overcome the discontinuity problem mentioned in Remark 9.14.
Definition 9.17. A set is strongly uniformly non-atomic if there exists a monotone increasing function such that and
Similarly, is weakly uniformly non-atomic if for every there exists a monotone increasing function satisfying such that
Corollary 9.18. Let be a separable Hilbert space, let , and let the sequence of measures be weak-star convergent to a measure .
Assume that the sequence is strongly uniformly non-atomic. Then the limit measure is also non-atomic, and the corresponding distribution functions and satisfy the inequalities
for all and , where is the same as in the definition of uniform non-atomicity for the family .
Assume that the sequence is weakly uniformly non-atomic. Then the limit measure is also non-atomic, and the corresponding distribution functions and satisfy the inequalities
for all , , and .
In both cases, for every , the scalar distribution functions converge to in .
Proof. Indeed, the first inequality in (9.37) follows from the inequality
the second is an immediate consequence of (9.28), and the convergence in follows from the Arzelà theorem and the Helly selection theorem stated above. The case of weakly uniformly non-atomic measures is treated similarly.
In particular, if the sequence is such that
then the weak-star limit measure is non-atomic.
Thus, under the assumptions of the above corollary, the discrete contribution of the -function vanishes. The next corollary gives a condition which guarantees that its singular part also vanishes. To this end we need one more definition.
Definition 9.19. A sequence of functions is equi- integrable if
for any Borel set (here is the Lebesgue measure of and is a monotone increasing continuous function which does not depend on and is such that ).
The next statement is a version of the Dunford–Pettis theorem for vector measures (see [7] for more details).
Theorem 9.20. Let be a separable Hilbert space, let , and let the sequence of measures be weak-star convergent in to a measure . Also let the corresponding distribution functions be absolutely continuous, so that . Then the are weakly convergent in if and only if they are equi-integrable. In this case the limit measure is also absolutely continuous and
We conclude this section with one more result related to the approximation of measures by sums of delta-measures; it plays an important role in the proof of one of our main results, Theorem 5.3.
Theorem 9.21. Let be a separable Hilbert space, and let and . Then there exists a sequence of discrete measures
with for all such that
and strongly in as and uniformly with respect to all . In particular, weak-star in .
Proof. We first note that, without loss of generality, we may assume that the measure is non-atomic (that is, that ). Indeed, in the general case we can split the measure into a discrete and a non-atomic part: , where
and . For these reasons we may consider and separately. In addition, the desired approximation for can obviously be chosen in the following form:
Thus, we assume from now on that and . Let
and define the sequence of measures by
Then by construction , and for any fixed
where is the largest such that . Since and is uniformly continuous, we have the uniform convergence . The weak-star convergence is an immediate consequence of this uniform convergence, and the theorem is proved.
Remark 9.22. Although approximation of measures by sums of delta-measures is a standard technical result which can be immediately obtained, say, from the Krein– Milman theorem, the convergence of to in the weak-star topology alone is not sufficient for our purposes due to the problems discussed in Remark 9.14. In contrast to the usual weak-star convergence, the result presented above has the extra important property that converges to pointwise and even uniformly in the strong topology of . This enables us to overcome the problem mentioned above. In particular, this uniform convergence implies that strongly in for every .
§ 10. Appendix 2. Key estimates in fractional Sobolev spaces
In this appendix we prove estimates for the -norm of the difference in terms of suitable norms of the functions and . These estimates are of fundamental importance for analysis of the regularity properties of uniform attractors. To this end, we need the following 'fractional Leibniz rule'.
Theorem 10.1 (Kato–Ponce inequality). Let and let , , , , be constants such that
Also, let the functions and on the -dimensional torus () satisfy the conditions
Then the product belongs to and satisfies the inequality
for some positive constant .
For the proof of this theorem see [3], for instance.
We apply this inequality to verify the following estimate.
Lemma 10.2. Let and let and be functions such that
Assume also that the function satisfies the inequality
for some constant . Then and the following estimate holds:
for some positive constant .
Proof. We apply the Kato–Ponce inequality to the function with the following exponents:
This gives us that
for some , where we have implicitly used the continuous embeddings
The resulting terms on the right-hand side can be estimated using standard interpolation inequalities. Indeed, using the growth assumption (10.3) we find that
for some . Also, we have
for some . Collecting (10.6), (10.8), and (10.9), we establish the required bound and complete the proof of the lemma.
Corollary 10.3. Let the assumptions of Lemma 10.2 be satisfied, and in addition let . Then
for some positive constant .
Consequently,
which gives the desired estimate.
Corollary 10.4. Let the function satisfy the inequality
and let the functions and satisfy (10.2) for some . Then the following estimate holds:
Indeed, according to (10.4) applied to , we have
Remark 10.5. The restriction in Lemma 10.2 is essential. Indeed, it is easy to see that (10.4) fails for . On the other hand, using the slightly sharper interpolation inequality
we see that (10.4) remains true for as well. We expect that it fails for although a rigorous proof of this fact is beyond the scope of our paper.
Part of this work was done while the first author held a postdoctoral position in the University of Cergy–Pontoise (France). He would like to thank N. Tzvetkov and A. Shirikyan for their hospitality and for interesting and stimulating discussions. The authors would also like to thank V. Chepyzhov, A. Mielke, and O. Smolyanov for stimulating discussions.
Footnotes
- 1
In [8] the author, following the tradition coming from probability theory, uses the notion of weak convergence of measures which coincides with the notion of weak-star convergence that we use here, following the terminology from functional analysis.