Abstract
We investigate the long-time behavior of solutions to a stochastically forced one-dimensional Navier–Stokes system, describing the motion of a compressible viscous fluid, in the case of linear pressure law. We prove existence of an invariant measure for the Markov process generated by strong solutions. We overcome the difficulties of working with non-Feller Markov semigroups on non-complete metric spaces by generalizing the classical Krylov–Bogoliubov method, and by providing suitable polynomial and exponential moment bounds on the solution, together with pathwise estimates.
Similar content being viewed by others
1 Introduction
We consider the question of existence of an invariant measure for the stochastically forced Navier–Stokes equations of compressible fluid flows in one space dimension. Formulated on the unit interval (0, 1), the equations read
where \(\rho =\rho (t,x)\) is the fluid density and \(u=u(t,x)\) is the fluid velocity. System (1.1)–(1.2) is appropriately written in dimensionless form, and is supplemented by the initial conditions
where \(\rho _0,u_0\) are assigned functions defined on the interval [0, 1], with the (strictly positive) initial density having normalized mass
In view of mass conservation, we obviously have
Moreover, we assume homogeneous Dirichlet boundary conditions for the velocity u, namely
Above, the driving noise is given by a collection of independent white noise processes, suitably colored in space (see (2.1) below), while \(A>0\) is a dimensionless parameter, inversely proportional to the Mach number. When considering strong (in the deterministic sense) solutions to (1.1)–(1.5), the natural phase space for our system is
The main result of this article is the following.
Theorem 1.1
For every fixed \(A>0\), the Markov semigroup \(\{{\mathcal {P}}_t\}_{t\ge 0}\) associated to (1.1)–(1.5) possesses an invariant probability measure \(\mu _A\in {\mathfrak {P}}(\mathcal {X}_{L^2})\). Furthermore,
Here, \(\mathcal {X}_{L^2}\) denotes the set \(\mathcal {X}\) in (1.6), endowed with the \(L^2\times L^2\) metric, and \(\Vert \sigma \Vert _{L^\infty }\) is given by (2.2) below.
The existence of statistically stationary states to randomly driven systems in fluid dynamics is of basic importance from both mathematical and experimental viewpoints. On the one hand, the existence of an invariant measure provides information on the generic long-time behavior of the system. On the other hand, under ergodicity assumptions, it provides a link between experimental observations (for example, in turbulence theory) and theoretical predictions.
In the context of fluid dynamics, a satisfactory theory of invariant measures has been developed for two-dimensional, incompressible flows [6, 7, 14, 22, 26, 27, 36, 37, 41, 42, 52]. In three dimensions, we mention the works [11, 12, 23], and especially [21], in which solutions which are strictly stationary stochastic processes are constructed, but the concept of invariant measure as a steady state is not well defined due to the absence of the Markov property. In this sense, the situation for multi-dimensional compressible flows is similar. To the best of our knowledge, the only relevant articles are [24], in which an invariant measure is constructed for an approximation of the Lagrangian version of (1.1)–(1.2) and for a different pressure law, and the more recent [4], in which statistically stationary solutions (but no invariant measure) are constructed. There are, however, results for both the dissipative and non-dissipative one-dimensional Burgers equation [10, 13, 25, 44, 51].
In this article, we prove existence of an invariant measure for the compressible Navier–Stokes system. The main advantage in working in one space dimension is that the equations are globally well-posed and can be solved pathwise, leaning on various works in the deterministic setting [28, 29, 31,32,33,34, 43, 46]. For the global solvability in the stochastic case, we mention [48, 49], valid in one space dimension, and [2, 5, 45, 47, 50] for the multi-dimensional case. For the deterministic set-up, we refer to the classical references [16, 38], and to [18, 19] for long-time behavior results.
In turn, (1.1)–(1.5) generates a proper Markov semigroup, and the concept of an invariant measure can be defined in a standard manner. However, there are several obstructions on the path towards the proof of Theorem 1.1. The main ingredients of our approach and contributions to the existing theory on invariant measures can be summarized as follows:
-
An \(L^2\)-based continuity result via the derivation of polynomial and exponential moment bounds As it is is clear from its statement, the topology on \(\mathcal {X}\) plays an essential role. The metric space \(\mathcal {X}_{L^2}\) is not complete (hence not Polish), due to both the open condition \(\rho >0\) and the fact that the Sobolev space \(H^1\) is not a closed subspace of \(L^2\). This choice is dictated by the available continuous dependence estimates (cf. Theorem 2.7), which make use of the so-called relative entropy functional and essentially provide an \(L^2\)-based continuity result. Such a result crucially depends on polynomial and exponential moment bounds of the solution to (1.1)–(1.5), which are carried out in detail in Sect. 2.1, and on careful pathwise estimates (see Sect. 2.2).
-
Introduction of a (larger) class of functions on \(\mathcal {X}_{L^2}\) which is invariant under the Markov semigroup \({\mathcal {P}}_t\) Closely related to the issue mentioned above, a second difficulty lies in the fact that the Markov semigroup associated to (1.1)–(1.5) is not known to be Feller (cf. Sect. 3.1), namely \({\mathcal {P}}_t\) may not map \(C_b(\mathcal {X}_{L^2})\) (the real-valued, continuous bounded functions on \(\mathcal {X}_{L^2}\)) into itself. This is again attributable to the mismatch between the set \(\mathcal {X}\) and the \(L^2\) topology, and causes problems when trying to apply the classical Krylov–Bogoliubov procedure to prove existence and, in particular, invariance of a probability measure constructed as a subsequential limit of time-averaged measures. We circumvent this issue by defining a class of functions on \(\mathcal {X}_{L^2}\) that is slightly larger that \(C_b(\mathcal {X}_{L^2})\), which is invariant under the Markov semigroup \({\mathcal {P}}_t\), and still well-behaved in the above limiting procedure.
-
Derivation of lower and upper bounds of the density with the correct time-averaged growth Lastly, the lack of instantaneous smoothing in the density equation (1.1) constitutes an obstacle towards compactness estimates for time-averaged measures. However, an energy structure already exploited in [32, 43] allows to obtain a dissipation term involving \(\Vert (\log \rho )_x\Vert _{L^2}\), thanks to the linear pressure law adopted here. Consequently, upper and lower bounds on the density (which, in 1D, can be proven with more general pressure laws) have the correct time-averaged growth and provide suitable tightness estimates for time-averaged measures (cf. Sect. 3.2).
-
The present article is, according to our knowledge, the first that presents rigorous results on the existence of invariant measures for compressible flows.
Remark 1.2
(The low Mach number limit) One may be tempted to study the behavior of the measures \(\mu _A\) as \(A\rightarrow \infty \), in the spirit of compressible-incompressible limits [3, 15, 17, 39, 40]. From (1.7), it is clear that, as \(A\rightarrow \infty \), the density component of the measures concentrates on sets such that \(\log \rho =0\), namely, \(\rho =1\). However, (1.1) implies that \(u_x=0\), and due to the homogeneous boundary conditions (1.5), we deduce that also \(u=0\). This is inconsistent with the second equation, as long as \(\sigma \ne 0\). By replacing \(\sigma \) with \(A^{-\eta }\sigma \), for any \(\eta >0\), it is easily seen from (1.7) that \(\mu _A\rightarrow \delta _{(\rho =1,u=0)}\) as \(A\rightarrow \infty \), essentially describing rigid body motion.
1.1 Extensions and Further Developments
Theorem 1.1 is in fact true for pressure laws of the type
and any polynomial growth as \(\rho \rightarrow \infty \). The case of pure power law \(p(\rho )=\rho ^\gamma \), for \(\gamma >1\), remains open. Another important question is the uniqueness (and hence ergodicity) and the attracting properties of the measure \(\mu _A\). It is worth to point out that the problem is highly degenerate, since the noise only acts on one component of the phase space. We conjecture that uniqueness holds, since the unforced system evolves (at a very slow rate) towards the unique steady state \((\rho ,u)=(1,0)\).
1.2 Outline of the Article
Section 2 is dedicated to the properties of solutions to the compressible Navier–Stokes equations. Energy estimates and moment bounds are proven in Sect. 2.1, pathwise estimates in Sect. 2.2, while uniqueness and continuous dependence on data are discussed in Sects. 2.3 and 2.4 , respectively. In Sect. 3 we prove Theorem 1.1, setting up the Markovian framework and discussing existence of invariant measure for non-Feller Markov processes on non-complete metric spaces.
1.3 Notation and Conventions
Throughout the paper, c will denote a generic positive constant independent of A, whose value may change even in the same line of a certain equation. In the same way, \(C_1, C_2, \ldots \) denote specific large deterministic constants. We will denote by \(\mathcal {X}_{L^2}\) the set \(\mathcal {X}\) endowed with \(L^2 \times L^2\) metric. Notice that \(\mathcal {X}_{L^2}\) is clearly a metric space. When we want to indicate the set \(\mathcal {X}\) endowed with the \(H^1\times H^1\) metric, we write \(\mathcal {X}_{H^1}\). By \(\mathcal {B}(\mathcal {X}_{L^2})\), we denote the family of Borel subsets of \(\mathcal {X}_{L^2}\). With the symbol \(M_b(\mathcal {X}_{L^2})\) (resp. \(C_b(\mathcal {X}_{L^2})\)) we refer to the set of all real valued bounded measurable (resp. bounded continuous) functions on \(\mathcal {X}_{L^2}\). Finally, \({\mathfrak {P}}(\mathcal {X}_{L^2})\) is the set of all probability measures on \(\mathcal {X}_{L^2}\).
2 Existence and Uniqueness of Solutions
In view of the particular form of the stochastic forcing in (1.2), the proof of existence and uniqueness of pathwise solutions borrows many ideas from the deterministic case. Fix a stochastic basis
We write the noise term in (1.2) as
The sequence \(W(t) =\{W^k(t)\}_{k\in {\mathbb {N}}}\) consists of independent copies of the standard one-dimensional Wiener process (Brownian motion). As such, for each k, \(\mathrm{d}W^k(t)\) is formally a white noise which, in particular, is stationary in time. Throughout the article, we will assume
In particular, this implies
By considering the change of variable \({{\tilde{u}}}=u-\sigma W\), a pathwise approach for the existence of solutions (even in the multidimensional case) has been carried out in [20] (see also [47,48,49] and the more recent [5] for local existence of strong solutions). In the deterministic setting and in one space dimension, the local existence of strong solutions has been proved in [46], while their global existence can be found in [43], in the case of pressure law of the type \(\rho ^\gamma \), for \(\gamma >1\). That this result can be extended to our case is a consequence of the estimates provided in the work [32] on (deterministic) weak solutions and their regularization properties. We summarize these observations in the theorem below.
Theorem 2.1
Fix a stochastic basis \(\mathcal {S}=(\Omega , \mathcal {F}, {\mathbb {P}}, \{\mathcal {F}_t\}_{t\ge 0}, W)\). For any \((\rho _0,u_0)\in \mathcal {X}\), there exists a unique \((\rho (t;\rho _0), u(t;u_0))\) satisfying (1.1)–(1.5), in the time integrated sense and with the regularity
almost surely. Moreover, the map \(t\mapsto (\rho (t;\rho _0), u(t;u_0))\) is weakly continuous, and for every \(t\ge 0\),
almost surely.
The bounds in the Theorem 2.1 (see also Lemma 2.2) for the density show that neither vacuum states nor concentration states can occur, no matter how large the initial datum is. This is one of several important differences between the Navier–Stokes equations and the inviscid Euler equations, for which vacuum states may in fact occur for large initial data and for certain equations of state (cf. [8, 9]). It is also relevant in this regard that solutions of the Navier–Stokes equations show certain instabilities when vacuum states are allowed (cf. Hoff and Serre [30]).
As far as the regularity of solutions is concerned, we need quantitative estimates and moment bounds that are new to the best of our knowledge. These estimates are contained in the the Propositions 2.3 and 2.4 below, and, in particular, agree with the regularity expressed in the theorem above.
2.1 Energy Estimates
For a pair \((\rho ,u)\in \mathcal {X}\), we define the entropy function in the classical way as
In the one-dimensional setting, it turns out the strong solutions are also well-behaved with respect to the modified energy functional
Several quantities are controlled by \(\mathcal {E}\). In particular, uniform estimates on \(\mathcal {E}\) entail lower bounds on \(\rho \).
Lemma 2.2
Assume that \((\rho ,u)\in \mathcal {X}\) are such that \(\mathcal {E}(\rho ,u)<\infty \). Then
and
Moreover,
Proof
As a preliminary observation we note that (pointwise)
Moreover, since \(\rho \in H^1\) and
we have
Therefore
By continuity of \(\rho \) and (2.7), we can choose \(x_0\in (0,1)\) such that \(\rho (x_0)=1\) and write
Hence,
Consequently,
From (2.8), the first bound (2.3) follows immediately. Regarding (2.4), we combine the upper bound in (2.3) with (2.8) to deduce that
Finally, (2.5) follows from the lower bound in (2.3). \(\square \)
It is fairly clear from the above estimates that if \((\rho ,u)\in \mathcal {X}\), then \(\mathcal {E}(\rho ,u)<\infty \). Viceversa, if \((\rho ,u)\) are smooth functions such that \(\mathcal {E}(\rho ,u)<\infty \), then \((\rho ,u)\in \mathcal {X}\). The above lemma provides a quantification of this dichotomy.
It is crucial for us to prove various estimates on strong solutions to the compressible Navier–Stokes equations, keeping particular attention to the dependence on the parameter \(A>0\). We collect these estimates in the next two propositions, whose proofs are carried out in the subsequent sections.
Proposition 2.3
Consider a solution \((\rho ,u)\) to (1.1)–(1.5) with initial data \((\rho _0,u_0)\in \mathcal {X}\) in (1.3) such that
Then there hold the entropy inequality
and the energy inequality
for all \(t\ge 0\). Moreover, the exponential martingale estimate
holds for every \(R\ge 0\) and for
As a consequence, for every \(m\ge 1\), there exists a constant \(c_m>0\) such that
for every \(T\ge 1\), and the exponential moment bound holds
q for all \(T\ge 1\).
The next result is instead a collection of similar, yet pathwise, estimates.
Proposition 2.4
Consider a solution \((\rho ,u)\) to (1.1)–(1.5) with initial data \((\rho _0,u_0)\in \mathcal {X}\) such that \(\mathcal {E}(\rho _0,u_0)<\infty \). For every \(T\ge 0\) and almost surely, there holds
In particular,
Moreover,
In the above the constants \(C_i\) can be explicitly computed.
The proof of (2.15) is simply a consequence of (2.3). We begin with Proposition 2.3.
2.1.1 Entropy Estimates
According to the Itō lemma
Hence, integrating by parts,
Moreover, by integration by parts and using only (1.1), we have
Combining the above computations we find
so that, for every \(t\ge 0\), we find the entropy balance
Notice that by (1.4) and (2.2), we have
so that
This is precisely (2.9).
2.1.2 Energy Estimates
Let \(\ell = \rho ^{-1/2}\) so that
As such
Notice that this computation only makes use of (1.1). On the other hand
Combining the above two equations and weighting appropriately we infer
and hence
We sum together (2.17) and half of (2.18) to finally obtain
Hence
so that
and therefore (2.10) holds.
2.1.3 Exponential Estimates
It remains to prove (2.11), which follows from (2.19) and an application of the exponential martingale estimate
valid for any continuous martingale \(\{Z(t)\}_{t\ge 0}\) with quadratic variation \(\langle Z\rangle (t)\). Indeed, consider the time integrated version of (2.19), namely
where Z(t) is the martingale
with quadratic variation
In view of the Poincaré-like inequality [32]
the mass constraint (1.4) and (2.2), we have
and
As a consequence, the quadratic variation of Z(t) can be estimated as
Now, from (2.21) we infer that the functional
satisfies with probability one the inequality
where we conveniently fix the constant \(\gamma _0>0\) as
In this way, from (2.23) it follows that
and thus (2.24) implies
In turn, from (2.20) and the above (2.25) we deduce that
for every \(R\ge 0\). This is exactly (2.11).
2.1.4 Polynomial and Exponential Moments
We now focus on (2.12), which is, in fact, an easy consequence of (2.11). Indeed, (2.26) implies that
where \(\Psi \) is the functional defined in (2.24). Using that, for a non-negative random variable Z, any \(m\ge 1\) and any constant \(c>0\), there holds
we infer from (2.27) that for some constant \(c_m>0\), independent of \(\gamma _0, T\) and \(\mathcal {E}(\rho _0,u_0)\) there holds
which is precisely (2.12). For (2.13), the idea is similar. Indeed, for a non-negative random variable Z and any \(\delta ,c>0\), we also have that that
Hence, for any \(\delta <\gamma _0\), we deduce from (2.27) that
and (2.13) simply follows by choosing \(\delta =\gamma _0/2\). We concluded the proof of Proposition 2.3.
2.2 Pathwise Estimates
By setting \({{\tilde{u}}}=u-w\), with \(w=\sigma W\) we obtain from (1.1) to (1.2) that the pair \((\rho ,{{\tilde{u}}})\) satisfies the system
where
with initial datum \((\rho _0,{{\tilde{u}}}_0)=(\rho _0,u_0)\). Notice that from (2.28) we can also write the velocity equation alone as
which will also be useful in the sequel. Without explicit reference, we will use (2.2) multiple times to bound the various norms of w appearing in the estimates.
2.2.1 Pathwise Energy Estimates
We begin by proving (2.14). A lengthy calculation analogous to that in Sect. 2.1.2 shows that
where
Recalling (2.29), we may rewrite the above term as
We now integrate by parts multiple times. Clearly,
Moreover, expanding all the derivatives and integrating by parts, the second term above reduces to
Concerning the third term, similar arguments lead to
Therefore, collecting all of the above we find that
and using (1.4), (2.22), (2.6) and standard inequalities we deduce that
implying that
Hence, (2.14) follows from the standard Gronwall lemma, together with the fact that
2.2.2 \(H^1\) Estimates on the Velocity
To prove (2.16), we multiply (2.30) by \(u_{xx}\) and integrate by parts, to obtain the identity
Now,
and
Concerning the last term, we find
Thus, (2.31) and the above estimates entail
In view of (2.14), (2.15) and Lemma 2.2, we can apply the Gronwall lemma to the above inequality and deduce (2.16).
2.3 Uniqueness of Strong Solutions
Let \((\rho ,u)\) and (r, v) be two solutions to (1.1)–(1.2). The relative entropy between \((\rho ,u)\) and (r, v) is defined as the functional
It is well-known that the above functional is not a distance, and it is not even symmetric. Nonetheless \(\mathcal {H}_{r}(\rho , u | r, v)=0\) if and only if \((\rho ,u)=(r,v)\). More precisely, the following well-known facts hold.
Lemma 2.5
Let \((\rho ,u)\in \mathcal {X}\) and \((r,v)\in \mathcal {X}\). Then
and
Proof
The proof is based on the fact that the function \(\xi \mapsto F(\xi ):=\xi \log \xi -\xi +1\) is convex. In particular,
Moreover, due to (1.4), the relative entropy can be rewritten as
Therefore, using Taylor’s theorem, on the one hand we have the lower bound
while on the other hand the upper bound reads
Using the above bounds in the expression for \(\mathcal {H}_{r}\) yields precisely (2.32)–(2.33), and the proof is over. \(\square \)
We then have the following uniqueness result.
Theorem 2.6
Let \((\rho ,u)\) and (r, v) be two solutions of (1.1)–(1.5), corresponding to the same initial condition \((\rho _0,u_0)\in \mathcal {X}\). Then
Proof
We compute the time derivative of \(\mathcal {H}_{r}(\rho , u | r, v) \). Firstly, notice that (1.2) can be rewritten as
so that the difference of the velocities satisfies
Therefore, making use of (1.1) and (2.34), we have that
On the other hand, due to (1.1) once more, we compute
Therefore, combining (2.35) and (2.36), we arrive at
We now estimate each term in the right-hand side above. We preliminary notice that for any smooth function g such that \(g(x_0)=0\) for some \(x_0\in [0,1]\), there holds the inequality
where \(\rho \) can be replaced by any positive function with mass at most 1 (and, in particular, by r). Indeed,
Concerning the first term in the right-hand side of (2.37), using standard inequalities and (2.32) we obtain
where the last line follows from (2.32). For the second term, we use (2.22) and (2.38) to deduce that
Hence, from (2.37) it follows that
Thanks to Proposition 2.4, the term multiplying \(\mathcal {H}_{r}(\rho , u | r, v)\) in the right-hand side above is locally integrable in time. We now apply the standard Gronwall lemma to (2.39) and use the fact that
by assumption to conclude that the two solutions are indistinguishable. \(\square \)
2.4 Continuous Dependence of Solutions
Regarding the continuous dependence of solutions, the picture is more complicated, and can be proven in the following situation.
Theorem 2.7
Let \(\{(\rho ^n,u^n)\}_{n\in {\mathbb {N}}}\) be a sequence of solutions to (1.1)–(1.5) with initial data \(\{(\rho _0^n,u_0^n)\}_{n\in {\mathbb {N}}}\) such that
and let \((\rho ,u)\) be a a solution to (1.1)–(1.5) with initial datum \((\rho _0,u_0)\in \mathcal {X}\). If
then
for all \(t\ge 0\).
Proof
We use the approach in Theorem 2.6. Arguing in the same way as we did to derive (2.39), we arrive at
Due to the lack of symmetry of the relative entropy, it is important that we consider the above quantity, and not the other possible \(\mathcal {H}_{r}(\rho , u | \rho ^n, u^n)\). Let
In view of Proposition 2.4, we have n-independent almost sure bounds of the form
and
The point is that the latter bound does not depend on \( \Vert (u^n_0)_x\Vert _{L^2}\), which is not, in general, under control uniformly in \(n\in {\mathbb {N}}\). Thanks to Lemma 2.5, the fact that \(\mathcal {E}(\rho ^n_0, u_0^n)\le R\) and (2.41), we have that
Thus, the standard Gronwall lemma applied to (2.42) implies
for all \(t\ge 0\). Hence, a further application of Lemma 2.5 leads to the conclusion of the proof. \(\square \)
3 Invariant Measures
In this section, we prove of the main results of this article, namely the existence of invariant measures to (1.1)–(1.2). As mentioned earlier, we will work in the phase space
However, in view of Theorem 2.7, the correct topology does not seem to be the natural one induced by the product norm in \(\mathcal {X}\), since Lemma 2.5 suggests that continuous dependence for our system holds in a weaker sense. We make this precise here below.
3.1 The Markovian Framework
We denote by \(\rho (t;\rho _0), u(t;u_0)\) the unique solution to (1.1)–(1.2), with (deterministic) initial data \((\rho _0, u_0)\in \mathcal {X}\). For a set \(B\in \mathcal {B}(\mathcal {X}_{L^2})\), we define the transition functions
for any \(t\ge 0\). For \(t\ge 0\), define the Markov semigroup
The usual semigroup properties
and
follow from the existence and uniqueness results for our system. The goal is to prove the existence of a probability measure \(\mu \in {\mathfrak {P}}(\mathcal {X}_{L^2})\) that is invariant under \({\mathcal {P}}_t\), namely such that
Now, it is clear from the estimates in Propositions 2.3 and 2.4 that
namely, \({\mathcal {P}}_t\) is well defined on measurable bounded functions. The Feller property, namely that \({\mathcal {P}}_t\) maps \(C_b(\mathcal {X}_{L^2})\) into \(C_b(\mathcal {X}_{L^2})\) is more delicate, and we are not able to prove it at the moment. In particular, it does not follow from Theorem 2.7, due to the additional requirement (2.40). However, Theorem 2.7 suggests the following definition.
Definition 3.1
Let \(\phi \in M_b(\mathcal {X}_{L^2})\). We say that \(\phi :\mathcal {X}\rightarrow {\mathbb {R}}\) belongs to the class \(\mathcal {G}\) if
whenever
From the above definition and Lemma 2.2, it is clear that
It turns out that \(\mathcal {G}\) is also invariant under \({\mathcal {P}}_t\).
Lemma 3.2
The class \(\mathcal {G}\) is invariant under the Markov semigroup \({\mathcal {P}}_t\), namely
Proof
Let \(\{(\rho _0^n,u_0^n)\}_{n\in {\mathbb {N}}}\subset \mathcal {X}_{L^2}\) be a sequence, complying with (3.1), namely
In light of Theorem 2.7,
Moreover, in view of (2.10) and (3.2), there exists a constant \({\bar{M}}:={\bar{M}}(t,M,\sigma )\) independent of \(n\in {\mathbb {N}}\) such that
which, by Chebyshev’s inequality, it implies that
for any \(R>0\). Fix \(\phi \in \mathcal {G}\), and let
To show that \({\mathcal {P}}_t \phi \in \mathcal {G}\), we are required to prove that for any fixed \(t>0\) there holds
Writing \(\mathcal {E}_n=\mathcal {E}(\rho (t;\rho ^n_0), u(t;u^n_0))\) for short, we have by definition that
Since \(\phi \in \mathcal {G}\), we also deduce by (3.2) and (3.3) that
and therefore by the bounded convergence theorem that
Let us now arbitrarily fix \(\varepsilon >0\), and pick \(R_\varepsilon =4{\bar{M}}M_\phi /\varepsilon \). Invoking (3.6), we deduce that
Moreover, from (3.7) there exists \(n_\varepsilon \in {\mathbb {N}}\) such that
Thus,
and the proof is over. \(\square \)
3.2 Tightness of Time-Averaged Measures
We prove the existence of an invariant measure for \({\mathcal {P}}_t\) via the classical Krylov–Bogoliubov procedure. However, since we are working in a non-complete metric space, and with a Markov semigroup that is not Feller, some details do not follow directly from the well-known theory. In what follows, the parameter \(A>0\) appearing in (1.2) is fixed, but we suppress the dependence on it of all the quantities (except for bounds), in order to keep the notation as simple as possible. As initial conditions to our problem, we fix \(\rho _0=1\) and \(u_0=0\). Notice that the corresponding energy vanishes, namely \(\mathcal {E}(1, 0)=0\). For \(T>0\), define the time-averaged measure on \(\mathcal {X}\) by
where \(B\in \mathcal {B}(\mathcal {X}_{L^2})\). The first step is to prove tightness of the family \(\{\mu _T\}_{T>0}\). We will make use of the following lemma.
Lemma 3.3
Assume that \(\rho \in H^1\) is such that \(\rho > 0\) and
Then, for \(x\in [0,1]\), we have the pointwise bounds
and, moreover
Proof
Firstly, we notice since \(\rho \) is continuous and has mean 1, there exists \(x_0\in (0,1)\) such that \(\rho (x_0)=1\). Therefore, for every \(x\in [0,1]\) we have
As a consequence,
proving (3.9). Moreover,
The claim then follows from (3.9). \(\square \)
As a consequence, tightness of \(\{\mu _T\}_{T>0}\) follows in a straightforward manner.
Proposition 3.4
The family of probability measures \(\{\mu _T\}_{T>0}\subset {\mathfrak {P}}(\mathcal {X}_{L^2})\) is tight. Hence, there exists a subsequence \(T_j\rightarrow \infty \) and a measure \(\mu \in {\mathfrak {P}}(\mathcal {X}_{L^2})\) such that
Proof
For any fixed \(R,S\ge 1\), define the sets
and
Note that since \(C_S\) is bounded in \(H^1\times H^1_0\) and is closed, it is compact in \(\mathcal {X}_{L^2}\). By Lemma 3.3, we have that if \((\rho ,u)\in K_R\), then
In particular, if \(R\ge 1\), we obtain
which therefore translates into the set inclusion
In view of (3.13) and Chebyshev’s inequality we have
In light of the energy inequality (2.10) and the fact that \(\mathcal {E}(1, 0)=0\), it follows that
Hence, the family \(\{\mu _T\}_{T>0}\) is tight, and therefore there exists a subsequential limit \(\mu \in {\mathfrak {P}}(\mathcal {X}_{L^2})\). Note that this uses the direction of Prokhorov’s theorem that does not require completeness (see [1, Theorem 6.1]). Now, since \(C_{S_R}\) is a closed set, the Portmanteau theorem implies
Notice again that this does not require the metric space \(\mathcal {X}_{L^2}\) to be complete (see [1, Theorem 2.1]). Thus
The proof is therefore concluded. \(\square \)
3.3 Invariance of the Limit Measure
We aim to prove the following proposition.
Proposition 3.5
For every fixed \(A>0\), the Markov semigroup \(\{{\mathcal {P}}_t\}_{t\ge 0}\) associated to (1.1)–(1.5) possesses an invariant probability measure \(\mu _A\in {\mathfrak {P}}(\mathcal {X}_{L^2})\). Furthermore,
The main issue here is that \({\mathcal {P}}_t\) is not known to be Feller, and therefore we cannot directly apply (3.10) to \({\mathcal {P}}_t\phi \). As a first step, we extend property (3.10) to the class \(\mathcal {G}\).
Lemma 3.6
Let \(\{\mu _{T_j}\}_{j\in {\mathbb {N}}}\subset {\mathfrak {P}}(\mathcal {X}_{L^2})\) be the convergent sequence in Proposition 3.4. Then
Proof
Let \(\phi \in \mathcal {G}\), arbitrarily fix \(R\ge 1\), and consider the compact sets \(C_{S_R}\) from (3.11), where \(S_R\) is given by (3.12). Since \(\phi \in \mathcal {G}\), \(\phi \) is bounded on \(\mathcal {X}_{L^2}\), so that
Moreover, the restriction of \(\phi \) to \(C_{S_R}\) is continuous (in the \(L^2\times L^2\) metric) on \(C_{S_R}\), for every \(R\ge 1\). Since \(C_{S_R}\) is a closed set, Tietze extension theorem (see e.g. [35]) guarantees the existence of a function \({{\tilde{\phi }}}_R\in C_b(\mathcal {X}_{L^2})\) such that
Now,
Fix \(\varepsilon >0\). By (3.14) and the above (3.17), there exists \(R_\varepsilon \ge 1\) such that
uniformly for \(j\in {\mathbb {N}}\), and, similarly, in light of (3.15),
Since \({\tilde{\phi }}_{R_\varepsilon }\in C_b(\mathcal {X}_{L^2})\), property (3.10) implies the existence of \(j_\varepsilon \in {\mathbb {N}}\) such that
Hence, for every \(\varepsilon >0\), there exists \(j_\varepsilon \in {\mathbb {N}}\) such that
which proves the claim. \(\square \)
We are now in the position to prove invariance of the limiting measure and the bound (3.16).
Proof of Proposition 3.5
As before, \(A>0\) is fixed and we suppress the various dependences on it. Let \(\phi \in C_b(\mathcal {X}_{L^2})\) be fixed. Thanks to Lemma 3.2,
In view of Lemma 3.6, we then have that
Hence, by (3.8), we conclude that
so that invariance follows from the arbitrariness of \(\phi \). Lastly, (3.16) is deduced directly from (2.10). Indeed, is \((\rho ^S,u^S)\in \mathcal {X}\) is a statistically stationary solution distributed as \(\mu \), the we can use (2.10) to derive the bound
which is precisely (3.16) after the usual change of variables. The proof is over. \(\square \)
References
Billingsley, P.: Convergence of Probability Measures, Second, Wiley Series in Probability and Statistics: Probability and Statistics. Wiley, New York (1999). A Wiley-Interscience Publication
Breit, D., Hofmanová, M.: Stochastic Navier–Stokes equations for compressible fluids. Indiana Univ. Math. J. 65(4), 1183–1250 (2016)
Breit, D., Feireisl, E., Hofmanová, M.: Incompressible limit for compressible fluids with stochastic forcing. Arch. Ration. Mech. Anal. 222(2), 895–926 (2016)
Breit, D., Feireisl, E., Hofmanova, M., Maslowski, B.: Stationary solutions to the compressible Navier–Stokes system driven by stochastic forces (2017). arXiv:1703.03177
Breit, D., Feireisl, E., Hofmanová, M.: Local strong solutions to the stochastic compressible Navier–Stokes system. Commun. Partial Differ. Equ. 43(2), 313–345 (2018)
Bricmont, J., Kupiainen, A., Lefevere, R.: Ergodicity of the 2D Navier–Stokes equations with random forcing. Commun. Math. Phys. 224(1), 65–81 (2001)
Bricmont, J., Kupiainen, A., Lefevere, R.: Exponential mixing of the 2D stochastic Navier–Stokes dynamics. Commun. Math. Phys. 230(1), 87–132 (2002)
Chang, T., Hsiao, L.: The Riemann Problem and Interaction of Waves in Gas Dynamics, Longman Scientific & Technical, Harlow. Wiley, New York (1989)
Chen, G.-Q., Glimm, J.: Global solutions to the compressible euler equations with geometrical structure. Commun. Math. Phys. 180, 153–193 (1996)
Da Prato, G., Debussche, A.: Differentiability of the transition semigroup of the stochastic Burgers equation, and application to the corresponding Hamilton-Jacobi equation. Atti Accad. Naz. Lincei Cl. Sci. Fis. Mat. Natur. Rend. Lincei (9) Mat. Appl 9(4), 267–277 (1999)
Da Prato, G., Debussche, A.: Ergodicity for the 3D stochastic Navier–Stokes equations. J. Math. Pures Appl. (9) 82(8), 877–947 (2003)
Da Prato, G., Debussche, A.: On the martingale problem associated to the 2D and 3D stochastic Navier–Stokes equations. Atti Accad. Naz. Lincei Rend. Lincei Mat. Appl. 19(3), 247–264 (2008)
Da Prato, G., Gatarek, D.: Stochastic Burgers equation with correlated noise. Stoch. Stoch. Rep. 52(1–2), 29–41 (1995)
Da Prato, G., Zabczyk, J.: Ergodicity for Infinite-Dimensional Systems, London Mathematical Society Lecture Note Series, vol. 229. Cambridge University Press, Cambridge (1996)
Desjardins, B., Grenier, E., Lions, P.-L., Masmoudi, N.: Incompressible limit for solutions of the isentropic Navier–Stokes equations with Dirichlet boundary conditions. J. Math. Pures Appl. (9) 78(5), 461–471 (1999)
Feireisl, E.: Dynamics of Viscous Compressible Fluids, vol. 26. Oxford University Press, Oxford (2004)
Feireisl, E., Novotný, A.: The low Mach number limit for the full Navier–Stokes–Fourier system. Arch. Ration. Mech. Anal. 186(1), 77–107 (2007)
Feireisl, E., Petzeltová, H.: Asymptotic compactness of global trajectories generated by the Navier–Stokes equations of a compressible fluid. J. Differ. Equ. 173(2), 390–409 (2001)
Feireisl, E., Petzeltová, H.: On the long-time behaviour of solutions to the Navier–Stokes–Fourier system with a timedependent driving force. J. Dyn. Differ. Equ. 19(3), 685–707 (2007)
Feireisl, E., Maslowski, B., Novotný, A.: Compressible fluid flows driven by stochastic forcing. J. Differ. Equ. 254(3), 1342–1358 (2013)
Flandoli, F., Gatarek, D.: Martingale and stationary solutions for stochastic Navier–Stokes equations. Probab. Theory Relat. Fields 102(3), 367–391 (1995)
Flandoli, F., Maslowski, B.: Ergodicity of the 2-D Navier–Stokes equation under random perturbations. Commun. Math. Phys. 172(1), 119–141 (1995)
Flandoli, F., Romito, M.: Markov selections for the 3D stochastic Navier–Stokes equations. Probab. Theory Relat. Fields 140(3–4), 407–458 (2008)
Gas, O.M.O.A.V.: Mesure invariante pour l’équation stochastique du mouvement d’un gaz visqueux en une dimension avec la discrétisation du domaine. Rev. Roum. Math. Pures Appl. 58(2), 149–162 (2013)
Goldys, B., Maslowski, B.: Exponential ergodicity for stochastic Burgers and 2D Navier–Stokes equations. J. Funct. Anal. 226(1), 230–255 (2005)
Hairer, M.: Exponential mixing properties of stochastic PDEs through asymptotic coupling. Probab. Theory Relat. Fields 124(3), 345–380 (2002)
Hairer, M., Mattingly, J.C.: Ergodicity of the 2D Navier–Stokes equations with degenerate stochastic forcing. Ann. Math. (2) 164(3), 993–1032 (2006)
Hoff, D.: Global existence for 1D, compressible, isentropic Navier–Stokes equations with large initial data. Trans. Am. Math. Soc. 303(1), 169–181 (1987)
Hoff, D.: Global solutions of the equations of one-dimensional, compressible flow with large data and forces, and with differing end states. Z. Angew. Math. Phys. 49(5), 774–785 (1998)
Hoff, D., Serre, D.: The failure of continuous dependence on initial data for the Navier–Stokes equations of compressible flow. SIAM J. Appl. Math. 51, 887–898 (1991)
Hoff, D., Smoller, J.: Non-formation of vacuum states for compressible Navier–Stokes equations. Commun. Math. Phys. 216(2), 255–276 (2001)
Hoff, D., Ziane, M.: The global attractor and finite determining nodes for the Navier–Stokes equations of compressible flow with singular initial data. Indiana Univ. Math. J. 49(3), 843–889 (2000)
Hoff, D., Ziane, M.: Finite-dimensional attractors and exponential attractors for the Navier–Stokes equations of compressible flow. SIAM J. Math. Anal. 34(5), 1040–1063 (2003). (electronic)
Kazhikhov, A.V., Shelukhin, V.V.: Unique global solution with respect to time of initial-boundary-value problems for one-dimensional equations of a viscous gas. J. Appl. Math. Mech. 41, 273–282 (1977)
Kelley, J.L.: General Topology. D. Van Nostrand Company Inc, Toronto (1955)
Kuksin, S., Shirikyan, A.: Stochastic dissipative PDEs and Gibbs measures. Commun. Math. Phys. 213(2), 291–330 (2000)
Kuksin, S., Shirikyan, A.: Mathematics of Two-Dimensional Turbulence, vol. 194. Cambridge University Press, Cambridge (2012)
Lions, P.-L.: Mathematical Topics in Fluid Mechanics, Vol. 2, vol. 10. The Clarendon Press/Oxford University Press, New York (1998)
Lions, P.-L., Masmoudi, N.: Incompressible limit for a viscous compressible fluid. J. Math. Pures Appl. (9) 77(6), 585–627 (1998)
Masmoudi, N.: Examples of singular limits in hydrodynamics. In: Handbook of Differential Equations: Evolutionary Equations, vol. III, pp. 195–275. North-Holland, Amsterdam (2007)
Masmoudi, N., Young, L.-S.: Ergodic theory of infinite dimensional systems with applications to dissipative parabolic PDEs. Commun. Math. Phys. 227(3), 461–481 (2002)
Mattingly, J.C.: Exponential convergence for the stochastically forced Navier–Stokes equations and other partially dissipative dynamics. Commun. Math. Phys. 230(3), 421–462 (2002)
Mellet, A., Vasseur, A.: Existence and uniqueness of global strong solutions for one-dimensional compressible Navier–Stokes equations. SIAM J. Math. Anal. 39(4), 1344–1365 (2008)
Sinaĭ, Ya G.: Two results concerning asymptotic behavior of solutions of the Burgers equation with force. J. Stat. Phys. 64(1–2), 1–12 (1991)
Smith, S.A.: Random perturbations of viscous, compressible fluids: global existence of weak solutions. SIAM J. Math. Anal. 49(6), 4521–4578 (2017)
Solonnikov, V.A.: The solvability of the initial-boundary value problem for the equations of motion of a viscous compressible fluid. Zap. Naučn. Sem. Leningrad. Otdel. Mat. Inst. Steklov. (LOMI) 56, 128–142 (1976). 197. Investigations on linear operators and theory of functions, VI
Tornatore, E.: Global solution of bi-dimensional stochastic equation for a viscous gas. NoDEA Nonlinear Differ. Equ. Appl. 7(4), 343–360 (2000)
Tornatore, E., Yashima, H.F.: One-dimensional equations of a barotropic viscous gas with a not very regular perturbation. Ann. Univ. Ferrara Sez. VII (N.S.) 7(40), 137–168 (1996)
Tornatore, E., Yashima, H.F.: One-dimensional stochastic equations for a viscous barotropic gas. Ricerche Mat. 46(2), 255–283 (1998)
Wang, D., Wang, H.: Global existence of martingale solutions to the three-dimensional stochastic compressible Navier–Stokes equations. Differ. Integral Equ. 28(11–12), 1105–1154 (2015)
Weinan, E., Khanin, K., Mazel, A., Sinai, Ya.: Invariant measures for Burgers equation with stochastic forcing. Ann. Math. (2) 151(3), 877–960 (2000)
Weinan, E., Mattingly, J.C., Sinai, Ya.G.: Gibbsian dynamics and ergodicity for the stochastically forced Navier–Stokes equation. Commun. Math. Phys. 224(1), 83–106 (2001)
Acknowledgements
The authors would like to thank Jacob Bedrossian, Sandra Cerrai, Dave Levermore, Jonathan Mattingly, Sam Punshon-Smith and Mohammed Ziane for helpful discussions. We also thank the program “New Challenges in PDE: Deterministic Dynamics and Randomness in High Infinite Dimensional Systems”, held at the Mathematical Science Research Institute (MSRI) in the Fall 2015, where some initial steps in the work were formulated. MCZ was in part supported by the National Science Foundation under the Award DMS-1713886. K.T. gratefully acknowledges the support in part by the National Science Foundation under the Award DMS-1614964. NGH was in part supported by the National Science Foundation under the Award DMS-1313272 and Simons Foundation 515990.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix A: A General Result on Existence of Invariant Measures
Appendix A: A General Result on Existence of Invariant Measures
The results in Sect. 3 on the existence of invariant measures can be stated in greater generality as follows. Let \((\mathcal {X}, d_s)\) be a metric space, and assume we are given another metric \(d_w\) on \(\mathcal {X}\) such that
-
(1)
Any \(d_s\)-bounded subset of \(\mathcal {X}\) is \(d_w\)-precompact.
-
(2)
If \(d_s(x_n,y_n)\rightarrow 0\) as \(n\rightarrow \infty \) for some \(x_n,y_n\in \mathcal {X}\), then \(d_w(x_n,y_n)\rightarrow 0\) as \(n\rightarrow \infty \).
We denote by \(\mathcal {X}_w\) the space \(\mathcal {X}\) endowed with the metric \(d_w\). Given \(x\in \mathcal {X}\), let \(\{X_t^x\}_{t\ge 0}\) be a Markov process on \(\mathcal {X}\) with \(X_0^x=x\), and define the time-averaged measures
where \(B\in \mathcal {B}(\mathcal {X}_w)\), the Borel \(\sigma \)-algebra of \(\mathcal {X}_w\). For \(t\ge 0\), define the Markov semigroup
We assume that the Markov semigroup is well defined on measurable bounded functions, namely the property that \({\mathcal {P}}_t : M_b(\mathcal {X}_w)\rightarrow M_b(\mathcal {X}_w)\) is well-defined.
Definition A.1
Let \(\phi \in M_b(\mathcal {X}_w)\). We say that \(\phi :\mathcal {X}\rightarrow {\mathbb {R}}\) belongs to the class \(\mathcal {G}\) if \(\phi \in C(K)\) for any \(d_s\)- bounded set \(K\subset \mathcal {X}\).
The following is a slight variant of the Krylov–Bogoliubov procedure to prove the existence of an invariant measure for \({\mathcal {P}}_t\).
Theorem A.2
Assume that for some initial condition \(x\in \mathcal {X}\) the following two properties hold.
-
(Tightness) For every \(\varepsilon \in (0,1]\), there exists a \(d_s\)-bounded and \(d_w\)-closed set \(K_\varepsilon \subset \mathcal {X}\) such that the time-averaged measures \(\{\mu ^x_T\}_{T>0}\) satisfy
$$\begin{aligned} \inf _{T>0} \mu ^x_T(K_\varepsilon )\ge 1-\varepsilon . \end{aligned}$$(A.2) -
(Almost Feller) For all \(t>0\),
$$\begin{aligned} {\mathcal {P}}_t C_b(\mathcal {X}_w)\subset \mathcal {G}. \end{aligned}$$(A.3)
Then there exists an invariant measure \(\mu \) for \({\mathcal {P}}_t\).
Proof
From the tightness property (A.2) and the fact that \(d_s\)-bounded sets are \(d_w\)-precompact, we infer the existence of a sequence \(T_j\rightarrow \infty \) and a measure \(\mu \in {\mathfrak {P}}(\mathcal {X}_w)\) such that
Let \(\varepsilon >0\), and let \(K_\varepsilon \) be the corresponding sets as above. Now, since \(K_\varepsilon \) is a \(d_w\)-closed set, the Portmanteau theorem and (A.2) imply
We now claim that, in fact,
To prove the claim, let \(\phi \in \mathcal {G}\) be arbitrarily fixed, and set
It is important to notice here that \(M_\phi \) does not depend on \(\varepsilon \). We can then apply Tietze extension theorem (see e.g. [35]) and deduce the existence of a function \({\tilde{\phi }}_\varepsilon \in C_b(\mathcal {X}_w)\) such that
Now,
By (A.2) and the above (A.7), we have
uniformly for \(j\in {\mathbb {N}}\), and, similarly, in light of (A.5),
Since \({\tilde{\phi }}_{\varepsilon }\in C_b(\mathcal {X}_{L^2})\), property (A.4) implies the existence of \(j_\varepsilon \in {\mathbb {N}}\) such that
Hence, for every \(\varepsilon >0\), there exists \(j_\varepsilon \in {\mathbb {N}}\) such that
which proves (A.6). We are now left to prove that \(\mu \) is invariant for \({\mathcal {P}}_t\). Let \(\phi \in C_b(\mathcal {X}_w)\) be fixed. Thanks to assumption (A.3),
In view of (A.6), we then have that
Hence, by (A.1), we conclude that
so that invariance follows from the arbitrariness of \(\phi \). \(\square \)
Rights and permissions
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
About this article
Cite this article
Coti Zelati, M., Glatt-Holtz, N. & Trivisa, K. Invariant Measures for the Stochastic One-Dimensional Compressible Navier–Stokes Equations. Appl Math Optim 83, 1487–1522 (2021). https://doi.org/10.1007/s00245-019-09594-x
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00245-019-09594-x