Abstract
In this article, we consider the Cauchy problem of the dissipative quasi-geostrophic equation in critical Fourier–Besov spaces. Using the bilinear-type fixed point theory, we obtain the global well-posedness results and we prove the existence and uniqueness of analytic solutions with small initial data belonging to the critical Fourier–Besov spaces \({\mathscr {F}}B_{p,q}^{1-2\alpha +\frac{2}{p'}} \). In addition, we show the stability of global solutions.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Introduction
In this article, we consider the following Cauchy problem for the two-dimensional quasi-geostrophic equation with subcritical dissipation \(\alpha >1 / 2\).
where \(\alpha >1 / 2\) is a real number, \(k>0\) is a dissipative coefficient. The scalar function v(x, t) represents the potential temperature, and \(u_v=\left( -\mathfrak {R}_{2} v, \mathfrak {R}_{1}v\right) \) is the divergence free velocity \((\nabla \cdot u_v=0)\) which is determined by the Riesz transformation of v. \(\Lambda \) is the operator defined by the fractional power of \(-\Delta \):
and more generally
where \({\mathscr {F}}\) is the Fourier transform.
For the purpose of simplicity, we make the assumption that \(k=1\) since we are interested in the dissipative case.
For fluid dynamics partial differential equations with singular data in various spaces, there is a large body of literature on global-in-time well-posedness, where the smallness conditions are considered in the norms of critical spaces (i.e., the norm is invariant under the scaling of the equation/system). For example, for Navier–Stokes equations, 2D quasi-geostrophic equations, and related models, we have well-posedness results in the critical case of the following spaces: Lebesgue space \(L^{p}\) [33], Marcinkiewicz space \(L^{p, \infty }\) [15, 46], Sobolev spaces \(H^{s}\) [25], Lei-Lin spaces \({\mathscr {X}}^{s}\) [11, 13, 14], Besov spaces \({B}_{p, q}^{s}\) [47], Triebel-Lizorkin spaces \({F}_{p, q}^{s}\) [18], Morrey spaces \({M}^{\lambda }_{p}\) [34], Besov–Morrey spaces \({\mathscr {N}}_{p, \lambda , q}^{s}\) [38], Fourier–Besov spaces \({\mathscr {F}} {\mathscr {B}}_{p, q}^{s}\) [31], Fourier–Herz spaces \({\mathscr {F}} {\mathscr {B}}_{1, q}^{s}={\mathscr {B}}_{1, q}^{s}\) [17], Fourier–Besov–Morrey spaces \({\mathscr {F}} {\mathscr {N}}_{p, \lambda , q}^{s}\) [1,2,3,4,5,6, 23, 26], \( B M O ^ { - 1 }\) [35], and pseudomeasure spaces \({\mathscr {P}}{\mathscr {M}}\) [40], among others. Moreover, in some of the above references, one can find results on regularity, decay and/or asymptotic behavior of solutions, such as the works [15, 16, 20, 26, 33, 34, 38, 46].
It is clear that the case \(\alpha \) small is mathematically interesting to understand the effects of a milder dissipation in the model, in a similar spirit to studies that consider small viscosities. Moreover, although often considered a toy model, the 2D quasi-geostrophic equation with \(\alpha =\frac{1}{2}\) presents some structural similarities with 3D Navier–Stokes as well as connections with geophysical models. In fact, the dissipative term being then of the form \(\Lambda v\), the last one models the Ekman pumping observable in ocean flows for example. This phenomena appears when the winds blowing at the surface of the oceans compete with the Coriolis force (deflecting the water at the surface of the oceans, to the right in the northern hemisphere and to the left in the southern hemisphere) causing the water to rise to the surface. Mathematically, the power \(\alpha =\frac{1}{2}\) corresponds to the index for which the nonlinear term and the dissipation are of the same order (in the sense that \(u_v\) and \(\Lambda v\) are two operators deriving once). The previous remarks motivate the following definition:
Definition 1
We define the following 3 cases:
-
If \(0 \le \alpha <\frac{1}{2}\), then Eq. (1) is called supercritical
-
If \(\alpha =\frac{1}{2}\), then Eq. (1) is called critical
-
If \(\frac{1}{2}<\alpha \le 1\), then Eq. (1) is called subcritical
Numerous works have been published in recent years on the study of the quasi-geostrophic equation, specifically the mathematical questions that naturally arise in the study of the Cauchy problem, namely questions of regularity, existence, blow-up, uniqueness, of solutions (strong, weak, mild). When we want to study questions about the existence (global or local) of solutions, it is important to know the scaling of the equation and to look for the critical spaces. For the quasi-geostrophic equation (1), we have the following scaling: if v solves (1) with initial data \(v_{0}\), then \(v_{\gamma }\) with \(v_{\gamma }(x, t):=\gamma ^{2\alpha -1} v\left( \gamma x, \gamma ^{2\alpha } t\right) \) is also a solution to (1) with the initial data
Any Banach space \( Y\subset {\mathscr {S}}^{\prime }\left( {\mathbb {R}}^{n}\right) \) whose norm is invariant under the scaling (2), i.e.,
is called critical for Eq. (1). We notice that the spaces \(L^{\frac{2}{2 \alpha -1}}, \; {\dot{H}}^{2-2 \alpha }\), \(\dot{{\mathscr {B}}}_{p, q}^{1-2\alpha +\frac{2}{p}}\) where \(p, q \in [1, \infty ]\) and \( {\mathscr {F}}B_{p,q}^{1-2\alpha +\frac{2}{p'}} \) with \(p, q \in [1, \infty ]\) are critical for (1). In this respect, there are several papers on global and local existence for (1) in different critical spaces. For example:
-In the subcritical case, corresponding to the case \(\frac{1}{2}<\alpha \le 1\), the global existence and uniqueness questions are well understood. This case corresponds to values of \(\alpha \) for which the dissipation is an operator that derives strictly more than once, while the nonlinear term contains exactly one derivative. Constantin and Wu proved in [17] that for any regular initial data the solution remains regular for any time. We can recall for example the results in Lebesgue spaces \(L^{\frac{2}{2 \alpha -1}}\) for \(\alpha >\frac{1}{2}\) by Carrillo and Ferreira [16], Sobolev spaces \({\dot{H}}^{2-2 \alpha }\) for \(\alpha >\frac{1}{2}\) by Ju [32]. In 2015, Benameur and Benhamed [13] proved the global existence of (QG) for small initial data and the local existence for large initial data in the Lei-Lin critical space \({\mathscr {X}}^{s}\) for \(s=1-2 \alpha \), which are defined as follows [13]:
For \(s\in {\mathbb {R}}\), \( {\mathscr {X}}^{s}\left( {\mathbb {R}}^{n}\right) =\left\{ f \in {\mathscr {S}}^{\prime }\left( {\mathbb {R}}^{n}\right) ;\int _{{\mathbb {R}}^{n}}\vert \xi \vert ^{s}\vert {\widehat{f}}(\xi )\vert d\xi < \infty \right\} , \) with the norm
Recently, Benhamed and Abusalim [14] studied the asymptotic behavior of (1) in Lei-Lin space \({\mathscr {X}}^{1-2\alpha }\) with subcritical dissipation. Many other related results can be found in [12, 16, 21, 24].
-In the critical case \(\alpha =\frac{1}{2}\), one of the first results (valid also in the supercritical case \(0 \le \alpha <\frac{1}{2}\)) was that of Resnick in 1995 [44] in which he proved the global existence of weak solutions with arbitrary large initial data \(L^{\infty } \cap L^{2}\) and satisfying a "Leray" type energy inequality. In 2006, Marchand [42] extended Resnick’s weak solution existence result. He proved that for any initial data \(v_{0} \in L^{p}\) with \(p>\frac{4}{3}\) and for any \(t \in (0, T)\), there exists a global weak solution verifying the following energy inequality:
If \(v _{0} \in H^{-1 / 2}\) then there exists a solution \(\left. v \in L^{1 / 2}\left( (0, T), H^{-1 / 2}\right) \cap L^{2}, H^{\alpha -1 / 2}\right) \) satisfying for all \(t \in (0, T)\) the energy inequality
-In the supercritical case \((0 \le \alpha <\frac{1}{2})\). We have global existence results when the initial data are small or local existence when the initial data are large. A. Córdoba and D. Córdoba proved in [22] that there exists a global solution if the initial data are in the Sobolev space \(H^{m}\) where \(m<2\). Chae and Lee in [19] showed that if the initial data are small in the critical space \(\dot{{\mathscr {B}}}_{2,1}^{2-\alpha }\) then there exists a unique global solution. This result was improved by Bae in [8] where he showed that it is sufficient that the initial data are small in \(\dot{{\mathscr {B}}}_{p, q}^{1+\frac{2}{p}-2 \alpha }\) with \(p<\infty \).
Writing the Cauchy problem (1) in integral form and solving the new problem using the fixed point method, we get the global solution. On the other hand, by setting: \({\tilde{V}}(t, x):=e^{ \sqrt{t} \vert D \vert ^{\alpha }} v(t, x)\) and considering the corresponding equations of \({\tilde{V}}(t, x)\), we also prove the Gevrey regularity of the solution in the same spaces, that is, \({\tilde{V}}(t, x)\) belong to the same spaces with v(t, x). Furthermore, we show that the solution v presents the asymptotic property \(\Vert v(t)\Vert _{{\mathscr {F}}B_{p,q}^{1-2\alpha +\frac{2}{p'}} }\longrightarrow 0\) as \(t \longrightarrow \infty .\) For that, we use standard interpolation in the Fourier–Besov space \({\mathscr {F}}B_{p,q}^{1-2\alpha +\frac{2}{p'}}\), energy estimates in \(L^{2}\), and Young’s inequality of convolutions, among others. The definition of the Fourier–Besov space is the Fourier transformation form of the Besov space. Exactly, analogous to that of a Besov space, the norm of the Littlewood–Paley blocks is taken in the frequency space instead of the physical space, that is, the homogeneous Fourier–Besov norm is defined as
To solve the equation (1), we consider the following equivalent integral equation coming from Duhamel’s principle
where \({\mathscr {S}}_{\alpha }:=e^{-t \Lambda ^{2\alpha }}\) denotes the fractional heat semigroup operator, which can be regarded as the convolution operator with the kernel \(K_t(x)={\mathscr {F}}^{-1}(e^{-t \vert \xi \vert ^{2\alpha }}),\) and
First, we show the existence of global solutions for (1).
Theorem 1
(Well-posedness) Let \(1\le p<\infty \), \(1\le q\le \infty \) and \(\frac{1}{2}<\alpha <1+\frac{1}{p'}\). There exists a constant \(\gamma >0\) such that for any \(v_0\in {\mathscr {F}}B_{p,q}^{1-2\alpha +\frac{2}{p'}} \) satisfying \(\Vert v_0\Vert _{{\mathscr {F}}B_{p,q}^{1-2\alpha +\frac{2}{p'}} }<\gamma \), the equation (1) admits a unique global solution
such that
where \(X_{p,q}={\mathscr {L}}^{\infty }\left( {\mathbb {R}}^{+}, {\mathscr {F}}B_{p,q}^{1-2\alpha +\frac{2}{p'}} \right) \cap {\mathscr {L}}^{1}\left( {\mathbb {R}}^{+}, {\mathscr {F}}B_{p,q}^{1+\frac{2}{p'}} \right) .\)
Remark 1
Taking \(n=2,\, a_{21}=1=-a_{12}, a_{11}=a_{22}=0, \left| P_{j}(\xi )\right| =\vert \xi \vert (j=1,2\), \(\beta =1\) ) in the expressions (1.2) to (1.5) in [26], we obtain the 2DQG (1). Then, using Theorem 3.1 in [26], we obtain a well-posedness result in Fourier–Besov–Morrey spaces \(\mathscr {F}\mathscr {N}_{p, \lambda , \infty }^{1-2\alpha +\frac{\lambda }{p}+\frac{2}{p'}}\) for (1) in the subcritical case, which is related to Theorem 1 but is different. It is worth noticing that in [26] it is used the persistence norm \(B C\left( {\mathbb {R}}^{+}, {\mathscr {F}} {\mathscr {N}}_{p, \lambda , \infty }^{1-2\alpha +\frac{\lambda }{p}+\frac{2}{p'}}\right) \) while Chemin–Lerner type norms are employed in our Theorem 1, namely the norms of \({\mathscr {L}}^{\infty }\left( {\mathbb {R}}^{+}, {\mathscr {F}} B_{p, q}^{1-2\alpha +\frac{2}{p'}}\right) \) and \({\mathscr {L}}^{1}\left( {\mathbb {R}}^{+}, {\mathscr {F}} B_{p, q}^{1+\frac{2}{p'}}\right) \).
Theorem 2
(Space analyticity) Under the assumptions of Theorem 1, the solution obtained in Theorem 1 is analytic in the sense that \(e^{\sqrt{t} \vert D \vert ^{\alpha }}v\in {X}_{p,q}\) and
where \(e^{ \sqrt{t} \vert D \vert ^{\alpha }}v={\mathscr {F}}^{-1}(e^{\sqrt{t}\vert \xi \vert ^{\alpha }}{\hat{v}})\).
As an application of the analyticity of solutions obtained in the above theorem, we can get the time decay estimates of solutions.
Theorem 3
(Decay estimate) Under the assumptions of Theorem 2, for any \(\sigma >0\), the global solution \(v \in {X}_{p, q}\) and \(e^{ \sqrt{t} \vert D \vert ^{\alpha }}v\in {X}_{p, q}\) achieved from Theorem 1 satisfies the following time decay estimate:
where \(C_{\alpha ,\sigma }\) is a constant depending on \(\alpha \) and \(\sigma \).
Next, we present an asymptotic stability result for solutions in the context of Fourier–Besov spaces.
Theorem 4
(Asymptotic stability) Let \(\frac{2}{3}<\alpha < 1\), \( 1 \le p,q \le 2\), and \(v\in {\mathscr {C}}\left( {\mathbb {R}}^{+}, {\mathscr {F}}B_{p,q}^{1-2\alpha +\frac{2}{p'}} \right) \) be a global solution of (1) given by Theorem 1. Then,
Remark 2
Theorems 1 and 4 extend the results of [13, 14] from Lei-Lin spaces to Fourier–Besov spaces. In fact, for \(p=q=1\), \({\mathscr {F}}B_{1,1}^{s}={\mathscr {X}}^{s}\) (see Remark 3).
2 Preliminaries
The definitions and notations used in this paper are presented as follows:
-
(1)
The Lei-Lin space is defined by
$$\begin{aligned} {\mathscr {X}}^{s}\left( {\mathbb {R}}^{n}\right) :=\left\{ f \in {\mathscr {S}}^{\prime }\left( {\mathbb {R}}^{n}\right) :\Vert f\Vert _{{\mathscr {X}}^{s}\left( {\mathbb {R}}^{n}\right) }=\int _{{\mathbb {R}}^{n}}\vert \xi \vert ^{s}\vert {\widehat{f}}(\xi )\vert \mathrm {d} \xi <\infty \right\} , \end{aligned}$$where \({\mathscr {S}}^{\prime }\left( {\mathbb {R}}^{n}\right) \) is the set of tempered distributions;
-
(2)
Let \(1 \le p,q \le \infty \), the mixed Lebesgue-sequence space \(l^{q}(L^p)\) consists of all sequences \(\{f_i\}_{i\in {\mathbb {Z}}}\) of measurable functions in \({\mathbb {R}}^{n}\) such that \(\Vert \{f_i\}_{i\in {\mathbb {Z}}}\Vert _{l^{q}(L^p)}<\infty .\) For \(\{f_i\}_{i\in {\mathbb {Z}}}\in l^{q}(L^p)\) we define
$$\begin{aligned} \Vert \{f_i\}_{i\in {\mathbb {Z}}}\Vert _{l^{q}(L^p)}:= \left( \sum _{j \in {\mathbb {Z}}}\left\| f_i\right\| _{L^p}^{q}\right) ^{\frac{1}{q}}; \end{aligned}$$ -
(2)
we will use \(\Vert \cdot \Vert _{X}\) to denote \(\Vert \cdot \Vert _{X\left( {\mathbb {R}}^{n}\right) }\);
-
(3)
For the convenience of description, we use \({\mathscr {F}}B_{p, q}^{s}\) to denote the homogeneous Fourier–Besov spaces, \({\dot{H}}^{s}\) to denote the usual homogeneous Sobolev space;
-
(4)
For f, we denote \(u_f:=\left( -\mathfrak {R}_{2} f, \mathfrak {R}_{1}f\right) \), where \(\mathfrak {R}_{j}=\partial _{x_{j}}(-\Delta )^{-1 / 2}\), \(j=1,2\) are the Riesz transforms;
-
(5)
The convolution product of a suitable pair of functions f and g on \({\mathbb {R}}^{n}\) is given by
$$\begin{aligned} (f * g)(x):=\int _{{\mathbb {R}}^{n}} f(y) g(x-y) d y; \end{aligned}$$ -
(6)
For any subset A of a set E, the symbol \(\chi _{A}\) denote the characteristic function of A defined by \(\chi _{A}(x)=1 \quad \) if \(x \in A, \quad \chi _{A}(x)=0 \quad \) elsewhere;
-
(7)
Let X, Y be Banach spaces, we denote
$$\begin{aligned} \Vert v\Vert _{X \cap Y}:=\Vert v\Vert _{X}+\Vert v\Vert _{Y} \quad \text {;} \quad \Vert (v, w)\Vert _{X}:=\Vert v\Vert _{X}+\Vert w\Vert _{X}; \end{aligned}$$ -
(8)
Throughout the paper, C stands for a generic positive constant, which may be different from line to line, and \(A \lesssim B\) means that there is a constant \(C>0\) such that
$$\begin{aligned} A \le C B; \end{aligned}$$ -
(9)
\(p^{\prime }\) is the conjugate of p satisfying \(\frac{1}{p}+\frac{1}{p^{\prime }}=1\) for \(1 \le p \le \infty ;\)
The proofs of the results discussed in this work are based on a dyadic partition of unity in the Fourier variables, known as the homogeneous Littlewood–Paley decomposition. We present briefly this construction below. For more detail, we refer the reader to [10].
Let \(f \in S^{\prime }\left( {\mathbb {R}}^{n}\right) .\) Define the Fourier transform as
and its inverse Fourier transform as
Let \(\varphi \in S\left( {\mathbb {R}}^{d}\right) \) be such that \(0 \le \varphi \le 1\) and \({\text {supp}}(\varphi ) \subset \left\{ \xi \in {\mathbb {R}}^{d}: \frac{3}{4} \le \vert \xi \vert \le \frac{8}{3}\right\} \) and
We denote
and
We now present some frequency localization operators:
and
where \({\dot{\Delta }}_{j}={\dot{S}}_{j}-{\dot{S}}_{j-1}\) is a frequency projection to the annulus \(\{\vert \xi \vert \sim 2^{j}\}\) and \({\dot{S}}_{j}\) is a frequency to the ball \(\{\vert \xi \vert \lesssim 2^{j}\}\).
From the definition of \({\dot{\Delta }}_{j}\) and \({\dot{S}}_{j}\), one easily derives that
The following Bony paraproduct decomposition will be applied throughout the paper.
where
We give the definition of the homogeneous Fourier–Besov spaces \({\mathscr {F}}{B}_{p, q}^{s}\) [45].
Definition 2
(Homogeneous Fourier–Besov spaces) Let \(1 \le p, q \le \infty \) and \(s \in {\mathbb {R}}.\) The homogeneous Fourier–Besov space \({\mathscr {F}}{B}_{p, q}^{s} \) is defined as the set of all distributions \(f \in {\mathscr {S}}^{\prime } \backslash {\mathscr {P}}, \; {\mathscr {P}}\) is the set of all polynomials, such that the norm \(\Vert f\Vert _{{\mathscr {F}}{B}_{p, q}^{s} }\) is finite, where
where \(\widehat{{\dot{\Delta }}_{j}f}=\varphi _j {\hat{f}}.\)
Remark 3
[36] Notice that in the case \(p=q\) we have an equivalent norm on \({\mathscr {F}}{B}_{p, q}^{s}\), that is,
Lemma 1
The derivation \(\partial ^{\alpha }_{\xi }: {\mathscr {F}}B^{s+\vert \alpha \vert }_{p,q} \rightarrow {\mathscr {F}}B^{s}_{p,q} \) is a bounded operator.
Proof
We have
where in (7) we used the fact that \(\vert \xi \vert \sim 2^j\) for all \(j \in {\mathbb {Z}}.\) \(\square \)
Remark 4
As a consequence of Lemma 1, we have the following estimates:
Proposition 5
[23] (Sobolev-type embedding) For \(p_{2} \le p_{1}\) and \(s_{2} \le s_{1}\) satisfying \(s_{2}+\frac{n}{p_{2}}=s_{1}+\frac{n}{p_{1}}\), we have the continuous inclusion
for all \(1 \le r_{1} \le r_{2} \le \infty \).
The definition of mixed space-time spaces is given below.
Definition 3
[23] Let \(s\in {\mathbb {R}},\;1\le p<\infty ,\; 1\le q,\rho \le \infty , \;\) and \(I=[0,T),\;T\in (0,\infty ]\). The space-time norm is defined on u(t, x) by
and denote by \({\mathscr {L}}^{\rho }(I,{\mathscr {F}}B_{p,q}^{s})\) the set of distributions in \(S'({\mathbb {R}}\times {\mathbb {R}}^{n})/{\mathscr {P}}\) with finite \(\Vert .\Vert _{{\mathscr {L}}^{\rho }(I,{\mathscr {F}}B_{p,q}^{s})}\) norm.
According to Minkowski inequality, it is easy to verify that
where \(\displaystyle \Vert u(t, x)\Vert _{L^{\rho }\left( I; {\mathscr {F}}B_{p,q}^{s}\right) }:=\left( \int _{I}\Vert u(\tau , \cdot )\Vert _{{\mathscr {F}}B_{p, q}^{s}}^{\rho } d \tau \right) ^{1 / \rho }.\)
At the end of this section, we will recall an existence and uniqueness result for an abstract operator equation in a Banach space that will be used to show Theorem 1 in the sequel. For the proof, we refer the reader to see [9, 41].
Proposition 6
Let X be a Banach space with norm \(\Vert .\Vert _{X}\) and \(B:X\times X\longmapsto X\) be a bounded bilinear operator satisfying
for all \(u,v\in X \) and a constant \(\eta >0\). Then, if \(0<\varepsilon <\frac{1}{4\eta }\) and if \(y\in X\) such that \(\Vert y\Vert _{X}\le \varepsilon \), the equation \(x:=y+B(x,x)\) has a solution \({\overline{x}}\) in X such that \(\Vert {\overline{x}}\Vert _{X}\le 2 \varepsilon \). This solution is the only one in the ball \({\overline{B}}(0,2\varepsilon )\). Moreover, the solution depends continuously on y in the sense: if \(\Vert y'\Vert _{X}< \varepsilon ,\;x'=y'+B(x',x')\), and \(\Vert x'\Vert _{X}\le 2\varepsilon \), then
3 Well-posedness
In this section, we present the proof Theorem 1. We have learned from Sect. 1 that the integral form of the equation (1) is as follows:
where \({\mathscr {S}}_{\alpha }(t)v_0:={\mathscr {F}}^{-1}(e^{-t \vert \xi \vert ^{2\alpha }}\widehat{v_0})\) and
Lemma 2
(Linear estimate) Let \(v_{0}\in {\mathscr {F}}B_{p,q}^{1-2\alpha +\frac{2}{p'}}\) for \(1\le p<\infty \) and \(1\le q\le \infty \). Then there exists a constant \(C>0\) such that
Proof
Since \({\text {supp}} \varphi _{j} \subset \left\{ 2^{j-1} \le \vert \xi \vert \le 2^{j+1}\right\} \), we get
Then, applying the Minkowski inequality, we obtain
Similarly,
This finishes the proof of Lemma 2. \(\square \)
Lemma 3
(Bilinear estimate) Under the hypothesis of Theorem 1, there exists a constant \(\eta > 0\) such that
for all \(v,\psi \in X_{p,q}.\)
Proof
Recalling that \(u_v=\left( -\mathfrak {R}_{2} v, \mathfrak {R}_{1}v\right) .\) Since \(\widehat{\mathfrak {R}_{j}v(\xi )}=-i \frac{\xi _j}{\vert \xi \vert }{\widehat{v}}(\xi ), j=1,2,\) then \(\nabla \cdot u_v=0\) and
Applying \({\dot{\Delta }}_{j} \) to \({\mathscr {B}}(v, \psi )\), then doing Fourier transform on it and taking the \(L^{p}\)-norm, using Minkowski’s inequality, we find that
where we have used the fact that \(\vert \xi \vert \sim 2^{j}\).
According to Bony’s paraproduct decomposition and quasi-orthogonality property for Littlewood–Paley decomposition, for fixed j, we find that
Then
Then, by the triangle inequalities in \(L^p\) and in \(\ell ^q\), we have
By applying Hölder’s inequality and Young’s inequality, we get
Therefore,
Taking the \(L^{\infty }\)-norm of \(I_1\) in time, we can conclude that
Multiplying (15) by \(2^{j(1-2\alpha +\frac{2}{p'})}\), and taking \(\ell ^{q}\)-norm with index j, using Minkowski’s inequality and Young’s inequality, we can see that
Analogously, we have
Next, we have
Therefore,
Taking the \(L^{\infty }\)-norm of \(I_3\) in time, we can conclude that
Multiplying (16) by \(2^{j(1-2\alpha +\frac{2}{p'})}\), and taking \(\ell ^{q}\)-norm with index j, using Young’s inequality, we can see that
where the condition \(\alpha <1+\frac{1}{p'}\) ensures that the series \(\sum _{i\le 3}2^{i(2-2 \alpha +\frac{2}{p'})}\) converges.
Hence,
When taking the \(L^1\)-norm in time to \(\left\| \widehat{ {\dot{\Delta }}_j {\mathscr {B}}(v, \psi )}\right\| _{L^p}\), we will obtain similar results that
This finishes the proof of Lemma 3.
From the above results and according to Proposition 6 (the fixed point theorem), we can prove Theorem 1. \(\square \)
4 Analyticity of global solutions
A number of researchers have focused on the analyticality of the solutions, especially with regard to the Navier–Stokes equations, for further information, see [7] and its references. In this section, we will show the Gevrey class regularity for (1) in the Fourier–Besov–Morrey space.
The next lemma is helpful in demonstrating spatial analyticality.
Lemma 4
Let \(0<s \le t<\infty \) and \(0\le \alpha \le 1\). Then the following inequality holds
for all \(\xi ,\ y\in {\mathbb {R}}^{n}\).
4.1 Proof of Theorem 2
Let \({\tilde{V}} =e^{\sqrt{t} \vert D \vert ^{\alpha }}v\) and using the integral equation (10), we obtain
Then
Let \(\rho =1\) (or \(\infty \)), multiplying with \(2^{j (1-2\alpha +\frac{2}{p\prime }}\varphi _j\) and taking \({\mathscr {L}}^{\rho }(I,L^p)\)-norm we can obtain
where \(e^{\sqrt{t}\vert \xi \vert -\frac{1}{2} t\vert \xi \vert ^{2}}=e^{-\frac{1}{2}\left( \sqrt{t}\vert \xi \vert -1\right) ^{2}+\frac{1}{2}} \le e^{\frac{1}{2}}\) and Lemma 4 are used. The remaining part of the proof is similar to the proof of Theorem 1, therefore the details can be omitted.
5 Time decay of mild solutions: proof of Theorem 3
Using the definition of the Fourier–Besov spaces, we have
Suppose the function \(g(x)=x^{\sigma } e^{-\sqrt{t} x^{\alpha }}\), where \(x \ge 0\), \(\alpha \) is a constant greater than 0. From the derivation of the function g, one can infer that \(g(x) \le g((\frac{x}{\alpha \sqrt{t}})^{\frac{1}{\alpha }}) \le \) \(C_{\alpha ,\sigma } t^{-\frac{\sigma }{2 \alpha }}\). Therbey,
6 Asymptotic stability
We will establish some crucial lemmas in the proof of Theorem 4.
Lemma 5
Let \(\frac{2}{3}<\alpha <1\) and \(1 \le p, q \le 2\). Then we have
Proof
Using the definition of the Fourier–Besov spaces, and Hölder’s inequality, we have
From Remark 3, one can see that \( {\dot{FB}}_{2,2}^{\alpha }={\dot{H}}^{\alpha }\) and \({\dot{FB}}_{2,2}^{0}=L^{2}\). Then, by Taking M such that \(2^{M}=\left( \frac{\Vert v\Vert _{{\dot{H}}^{\alpha }}}{\Vert v\Vert _{L^{2}}}\right) ^{1/\alpha }\), we obtain
\(\square \)
Lemma 6
Let \(\frac{1}{2}<\alpha <1\), \( 1 \le p<\infty \) and \(1\le q\le 2\). Then we have
Proof
To prove this lemma, we will follow the method described in the proof of Lemma 3, for fixed j, we have
One can write
Then
Multiplying by \(2^{2j(1-\alpha )}\), and taking \(l^2\)-norm of both sides in the above estimate, we obtain
Similarly,
Therefore, by the Young inequality for series, we get
where we have used Proposition 5 and the fact that \(\alpha <1.\)
Now we deal with the third term \(\mathrm {R}_3\), Again employing Young’s inequality and Hölder’s inequality and the fact that \(\alpha <1\), we obtain
Thus,
Combining \(\mathrm {R}_1\), \(\mathrm {R}_2\) and \(\mathrm {R}_3 \), we conclude the desired result. \(\square \)
6.1 Proof of Theorem 4
To show the asymptotic stability for the global solution, we follow ideas from [28], in which the authors used standard interpolation in the Fourier space and energy estimates in \(L^{2}\). (see also [12, 14, 45]).
Let \(\varepsilon >0\), such that \(\varepsilon \le \sigma \). For \(m \in {\mathbb {N}}\), setting
Clearly \({\mathscr {F}}^{-1}\left( \chi _{\mathrm {S}_{m}} {\hat{v}}_{0}\right) \rightarrow v_0\) in \({\mathscr {F}}B_{p,q}^{1-2\alpha +\frac{2}{p'}}.\) Then, there is \(m_0\in {\mathbb {N}} \) such that
Let m be a fixed integer such that \(m\ge m_0.\)
Set \(v_{0, m}={\mathscr {F}}^{-1}\left( \chi _{\mathrm {S}_{m}} {\hat{v}}_{0}\right) , w_{0,m} =v_{0}-{\mathscr {F}}^{-1}\left( \chi _{\mathrm {S}_{m}} {\hat{v}}_{0}\right) .\) Therefore, we have proved that
Next, we consider the system
For all \(m\ge m_0\), we have \(\left\| w_{0,m} \right\| _{{\mathscr {F}}B_{p,q}^{1-2\alpha +\frac{2}{p'}} } \le \frac{\varepsilon }{2}\). So, we can conclude from Theorem 1 that there exists a unique global solution \(w_m \in {\mathscr {C}}\left( {\mathbb {R}}^{+}, {\mathscr {F}}B_{p,q}^{1-2\alpha +\frac{2}{p'}} \right) \cap {\mathscr {L}}^{1}\left( {\mathbb {R}}^{+}, {\mathscr {F}}B_{p,q}^{1+\frac{2}{p'}} \right) \).
In addition, \(\forall t\ge 0\) we have
Put
with v is the solution of the Eq. (1). Then, \(v_m\) is a solution of the system:
Furthermore, \(v_m\in {\mathscr {C}}\left( {\mathbb {R}}^{+}, {\mathscr {F}}B_{p,q}^{1-2\alpha +\frac{2}{p'}} \right) \cap {\mathscr {L}}^{1}\left( {\mathbb {R}}^{+}, {\mathscr {F}}B_{p,q}^{1+\frac{2}{p'}} \right) .\) By taking the inner products in \(L^{2}({\mathbb {R}}^{2})\) with \(v_m\) and integrating by parts, we can get
Since
Thus, Cauchy–Schwarz inequality, the estimate (14), Lemma 6, and Young’s inequality give
where in (22) we used that \(\left\| w_m \right\| _{{\mathscr {F}}B_{p,q}^{1-2\alpha +\frac{2}{p'}} }\le \frac{\varepsilon }{2}.\)
Consequently,
Integrating with respect to time, we obtain
By Gronwall’s lemma, we get
where the following fact is used in the last inequality: \(\displaystyle \int _{0}^{t}\left\| w_m \right\| ^{2}_{{\mathscr {F}}B_{p,q}^{1-\alpha +\frac{2}{p'}} }\le C.\)
Indeed:
Since \(q\le 2\), then by Minkowski’s inequality (9) we have
Thus, using (25) and Hölder’s inequality for series, we get
Combining (23) and (24), we obtain
Now, applying Lemma 5 we deduce
Therefore,
where we have used (24).
Finally, by integrating in time between 0 and \(\infty \), we obtain
Let us consider the following subset of \([0,\infty [\):
Then, for all \(\frac{1}{2}<\alpha <1,\)
where \(\mu \left( {{\mathscr {A}}_{\varepsilon }}\right) \) is the Lebesgue measure of \({\mathscr {A}}_{\varepsilon }\).
Using the estimates (28) and (29), we get \(\mu \left( {{\mathscr {A}}_{\varepsilon }}\right) <\infty \) and \(\displaystyle \mu \left( {{\mathscr {A}}_{\varepsilon }}\right) \lesssim \left( {\frac{2}{\varepsilon }}\right) ^{\frac{\alpha }{1-\alpha }}\int ^{\infty }_{0}\Vert v_m\Vert _{{\dot{H}}^{\alpha }}^{2}.\) For \(\eta >0,\) there exists \(t_0\in [0,\mu \left( {{\mathscr {A}}_{\varepsilon }}\right) +\eta ]\) such that \(t_0 \notin {\mathscr {A}}_{\varepsilon }.\)
Then,
Therefore, we have
Consequently,
Now, we consider the quasi-geostrophic equation starting at \(t=t_0\).
Using inequality (30) and Theorem 1, we infer that there exists a unique solution \(\theta \in {\mathscr {C}}\left( {\mathbb {R}}^{+}, {\mathscr {F}}B_{p,q}^{1-2\alpha +\frac{2}{p'}} \right) \cap {\mathscr {L}}^{1}\left( {\mathbb {R}}^{+}, {\mathscr {F}}B_{p,q}^{1+\frac{2}{p'}} \right) \) of the problem (1).
The existence and uniqueness of a solution to the quasi-geostrophic equation gives \(\forall t \ge 0 \quad \theta (t)=v\left( t_{0}+t\right) \). Then,
This completes the proof of Theorem 4.
Remark 5
If \(k\ne 1\), the results of Theorem 1 and Theorem 4 remain true, but with more complicated calculations.
Data availability statement
No datasets were generated or analysed during the current study.
References
Azanzal, A., Allalou, C., Abbassi, A.: Well-posedness and analyticity for generalized Navier-Stokes equations in critical Fourier-Besov-Morrey spaces. J. Nonlinear Funct. Anal. 2021, 1–14 (2021)
Azanzal, A., Abbassi, A., Allalou, C.: Existence of solutions for the Debye-Hückel system with low regularity initial data in critical Fourier–Besov–Morrey spaces. Nonlinear Dyn. Syst. Theory 21, 367–380 (2021)
Azanzal, A., Abbassi, A., Allalou, C.: On the Cauchy problem for the fractional drift-diffusion system in critical Fourier–Besov–Morrey spaces. Int. J. Optim. Appl. 1, 28–36 (2021)
Azanzal, A., Allalou, C., Melliani, S.: Well-posedness, analyticity and time decay of the 3D fractional magneto-hydrodynamics equations in critical Fourier-Besov-Morrey spaces with variable exponent. J. Elliptic Parabolic Equ. 8, 1–20 (2022). https://doi.org/10.1007/s41808-022-00172-x
Azanzal, A., Abbassi, A., Allalou, C.: Gevrey Class Regularity for the 2D Subcritical Dissipative Quasi-geostrophic Equation in Critical Fourier-Besov-Morrey Spaces. In: Melliani, S., Castillo, O. (eds) Recent Advances in Fuzzy Sets Theory, Fractional Calculus, Dynamic Systems and Optimization. ICPAMS 2021. Lecture Notes in Networks and Systems, vol 476. Springer, Cham (2023). https://doi.org/10.1007/978-3-031-12416-7-5
Azanzal, A., Allalou, C., Melliani, S.: Well-posedness and blow-up of solutions for the 2D dissipative quasi-geostrophic equation in critical Fourier–Besov–Morrey spaces. J. Elliptic Parabol. Equ. 8, 23–48 (2021). https://doi.org/10.1007/s41808-021-00140-x
Bae, H.: Existence and analyticity of Lei–Lin solution to the Navier–Stokes equations. Proceedings of the American Mathematical Society, 2887–2892 (2015)
Bae, H.: Global well-posedness of dissipative quasi-geostrophic equation in critical spaces. Proc. Am. Math. Soc. 136, 257–261 (2008)
Bahouri, H., Chemin, J.Y., Danchin, R.: Fourier Analysis and Nonlinear Partial Differential Equations. Springer, Amsterdam (2011)
Bahouri, H.: The Littlewood-Paley theory: a common thread of many words in nonlinear analysis. Eur. Math. Soc. Newsl. (2019)
Benameur, J.: Long time decay to the Lei-Lin solution of 3D Navier–Stokes equations. J. Math. Anal. Appl. 422, 424–434 (2015)
Benameur, J., Blel, M.: Long-time decay to the global solution of the 2D dissipative quasigeostrophic equation. In Abstract and Applied Analysis, 2012 (2012)
Benameur, J., Benhamed, M.: Global existence of the two-dimensional QGE with sub-critical dissipation. J. Math. Anal. Appl. 423, 1330–1347 (2015)
Benhamed, M., Abusalim, S.M.: Long time behavior of the solution of the two-dimensional dissipative QGE in Lei-Lin Spaces. Int. J. Math. Math. Sci. 2020 (2020)
Carrillo, J.A., Ferreira, L.C.F.: Self-similar solutions and large time asymptotics for the dissipative quasi-geostrophic equation. Monatsh. Math. 151(2), 111–142 (2007)
Carrillo, J.A., Ferreira, L.C.F.: The asymptotic behaviour of subcritical dissipative quasi-geostrophic equation. Nonlinearity 21, 1001–1018 (2008)
Cannone, M., Wu, G.: Global well-posedness for Navier–Stokes equations in critical Fourier–Herz spaces. Nonlinear Anal, 75 (2012)
Chae, D.: On the well-posedness of the Euler equations in the Triebel-Lizorkin spaces. Commun. Pure Appl. Math. J. Issued Courant Inst. Math. Sci. 55, 654–678 (2002)
Chae, D., Lee, J.: Global well-posedness in the super-critical dissipative quasi-geostrophic equation. Commun. Math. Phys. 233, 297–311 (2003)
Constantin, P., Majda, A.J., Tabak, E.: Formation of strong fronts in the 2-D quasigeostrophic thermal active scalar. Nonlinearity 7(6), 1495–1533 (1994)
Constantin, P., Wu, J.: Behavior of solutions of 2D quasi-geostrophic equation. SIAM J. Math. Anal. 30, 937–948 (1999)
Córdoba, A., Córdoba, D.: A maximum principle applied to quasi-geostrophic equation. Commun. Math. Phys. 249, 511–528 (2004)
de Almeida, M.F., Ferreira, L.C.F., Lima, L.S.M.: Uniform global well-posedness of the Navier–Stokes–Coriolis system in a new critical space. Math. Z. 287, 735–750 (2017)
Dong, H., Li, D.: On the 2D critical and supercritical dissipative quasi-geostrophic equation in Besov spaces. J. Differ. Equ. 248, 2684–2702 (2010)
Duvaut, G., Lions, J.L.: Inéquations en thermoélasticité et magnétohydrodynamique. Arch. Ration. Mech. Anal. 46, 241–279 (1972)
Ferreira, L.C.F., Lima, L.S.M.: Self-similar solutions for active scalar equations in Fourier-Besov-Morrey spaces. Monatsh. Math. 175(4), 491–509 (2014)
Foias, C., Temam, R.: Gevrey class regularity for the solutions of the Navier–Stokes equations. J. Funct. Anal. 87, 359–369 (1989)
Gallagher, I., Iftimie, D., Planchon, F.: Non-blowup at large times and stability for global solutions to the Navier–Stokes equations. CR Math. Acad. Sci. Paris 334, 289–292 (2002)
Grujié, Z., Kukavica, I.: Space analyticity for the Navier–Stokes and related equations with initial data in \(L^p\). J. Funct. Anal. 152(2), 447–466 (1998)
Iwabuchi, T.: Global well-posedness for Keller-Segel system in Besov type spaces. J. Math. Anal. Appl. 379, 930–948 (2011)
Iwabuchi, T., Takada, R.: Global well-posedness and ill-posedness for the Navier–Stokes equations with the Coriolis force in function spaces of Besov type. J. Funct. Anal. 267, 1321–1337 (2014)
Ju, N.: Dissipative \(2 \rm D \) quasi-geostrophic equation: local well-posedness, global regularity and similarity solutions. Indiana Univ. Math. J. 56, 187–206 (2007)
Kato, T.: Strong \(L^p\)-solutions of the Navier–Stokes equation in \({\mathbb{R} }^m\), with applications to weak solutions. Math. Z. 187, 471–480 (1984)
Kato, Tosio: Strong solutions of the Navier-Stokes equation in Morrey spaces. Boletim da Sociedade Brasileira de Matemética-Bulletin/Brazilian Mathematical Society, 22 , 127-155 (1992)
Koch, H., Tataru, D.: Well-posedness for the Navier–Stokes equations. Adv. Math. 157, 22–35 (2001)
Konieczny, P., Yoneda, T.: On dispersive effect of the Coriolis force for the stationary Navier–Stokes equations. J. Differ. Equ. 250, 3859–3873 (2011)
Konieczny, P., Yoneda, T.: On dispersive effect of the Coriolis force for the stationary Navier–Stokes equations. J. Differ. Equ. 250, 3859–3873 (2011)
Kozono, H., Yamazaki, M.: Semilinear heat equations and the Navier–Stokes equation with distributions in new function spaces as initial data. Commun. Partial Differ. Equ. 19, 959–1014 (1994)
Lemarié-Rieusset, P.G.: The Navier-Stokes Problem in the 21st Century. CRC Press, New York (2018)
Liu, Q., Zhao, J., Cui, S.: Existence and regularizing rate estimates of solutions to a generalized magneto-hydrodynamic system in pseudomeasure spaces. Annali di Matematica 191, 293–309 (2012)
Lemarié-Rieusset, P.G.: Recent Developments in the Navier-Stokes Problem. CRC Press, New York (2002)
Marchand, F.: Existence and regularity of weak solutions to the quasigeostrophic equations in the spaces Lp or \(H^{-1/2}\). Commun. Math. Phys. 277, 45–67 (2008)
Pedlosky, J.: Geophysical Fluid Dynamics. Springer, New York, 710 (1987)
Resnick, S.: Dynamical problems in nonlinear advective partial differential equations. Ph.D. Thesis, University of Chicago (1995)
Xiao, W., Chen, J., Fan, D., Zhou, X.: Global well-posedness and long time decay of fractional Navier–Stokes equations in Fourier-Besov spaces. In Abstract and Applied Analysis Hindawi, 2014 (2014)
Yamazaki, M.: The Navier–Stokes equations in the weak-\(L^n\) space with time-dependent external force. Math. Ann. 317, 635–675 (2000)
Zhao, J.: Well-posedness and Gevrey analyticity of the generalized Keller-Segel system in critical Besov spaces. Annali di Matematica 197, 521–548 (2018)
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Azanzal, A., Allalou, C. & Melliani, S. Global well-posedness, Gevrey class regularity and large time asymptotics for the dissipative quasi-geostrophic equation in Fourier–Besov spaces. Bol. Soc. Mat. Mex. 28, 74 (2022). https://doi.org/10.1007/s40590-022-00468-x
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s40590-022-00468-x
Keywords
- Quasi-geostrophic equation
- Self-similar solutions
- Gevrey class regularity
- Large time asymptotics
- Fourier–Besov spaces