Home Stability of pyramidal traveling fronts in time-periodic reaction–diffusion equations with degenerate monostable and ignition nonlinearities
Article Open Access

Stability of pyramidal traveling fronts in time-periodic reaction–diffusion equations with degenerate monostable and ignition nonlinearities

  • Yuan-Hao Liu , Zhen-Hui Bu EMAIL logo and Suobing Zhang
Published/Copyright: March 31, 2025

Abstract

This article is concerned with the stability of time-periodic traveling fronts for reaction–diffusion equations with time-periodic degenerate monostable and ignition nonlinearities. Based on the existence of pyramidal traveling fronts established by Bu et al., we characterize the pyramidal traveling front as a combination of planar fronts on the lateral surfaces and establish some important estimates. Then, we show the asymptotic stability of pyramidal traveling fronts by using the sub-super solution method combined with the comparison principle.

MSC 2010: 35K57; 35C07; 35B35; 35B40

1 Introduction

In this article, we investigate the following reaction–diffusion equation:

(1.1) u t ( x , t ) = Δ u ( x , t ) + f ( u ( x , t ) , t ) , x R 3 , t > 0 ,

where x = ( x , y , z ) , u t = u t , and Δ = 2 x 2 + 2 y 2 + 2 z 2 . The nonlinear term f satisfies the following hypotheses:

  1. f ( u , t ) C 1 + P , P 2 ( [ 1 , 2 ] × R , R ) , f ( u , t ) = f ( u , t + T ) , where P ( 0 , 1 ) and T > 0 .

  2. f ( u , t ) = 0 for ( u , t ) ( [ 0 , θ 0 ] 1 ) × R , where θ 0 [ 0 , 1 ) . Moreover, f u ( 1 , t ) < 0 , f u ( 0 , t ) = 0 , and f ( u , t ) > 0 for ( u , t ) ( θ 0 , 1 ) × R .

Clearly, the nonlinearity term f is a degenerate monostable nonlinearity when θ 0 = 0 , and f is known as an ignition nonlinearity when θ 0 > 0 . The reaction–diffusion equation (1.1) is an important type of parabolic partial differential equation [9,16,20,34].

For the nonlinear parabolic equation (reaction–diffusion equation) and the variations, the well-posedness of solutions, including existence, uniqueness, and dynamical behavior, has been investigated by many authors. For example, Xu and Su [40] studied the semilinear pseudo-parabolic equation. Xu and Su [40] proved the global existence, nonexistence, and asymptotic behavior of solutions with low initial energy and the critical initial energy by introducing a family of potential wells. Moreover, they also obtained finite-time blowup with high initial energy by the comparison principle. Based on the model described in [40], and considering a porous medium, Lian et al. [17] considered a model with a singular potential term x s u t . They obtained the global existence, asymptotic behavior, and blowup of solutions with the low initial energy. Moreover, they estimated the upper bound of the blowup time for the low initial energy not equal to 0. And at the same time, they also proved the finite-time blowup and estimated the upper bound of the blowup time for the high initial energy. Furthermore, Wang and Xu [35] investigated the initial boundary-value problem of semilinear pseudo-parabolic equations with nonlocal terms. By introducing the nonlocal term, the equation gives insight into biological and chemical problems where conservation properties predominate. Wang and Xu [35] established the existence, uniqueness, and asymptotic behavior of the global solution and the blowup phenomena of solution with subcritical initial energy. These results could be extended parallelly to the critical initial energy. In addition, they also studied the blowup phenomena of solution with supercritical initial energy. Xu et al. [39] investigated the initial boundary-value problem of a class of reaction–diffusion systems with nonlinear coupled source terms. They gave the sufficient initial conditions of global existence, long time decay, and finite-time blowup of solutions for the low initial energy case and critical initial energy case. For the high initial energy case, they first proved the possibility of both global existence and finite-time blowup of solutions and then obtained some sufficient initial conditions of finite-time blowup and global existence of solutions, respectively. For more studies on nonlinear parabolic equations, we refer to [10,25] and references therein.

In this article, we are more concerned with the global solution of parabolic partial differential equations. That is, a solution with a definition for all times t R which will not blowup in finite-time. Among various global solutions to the parabolic partial differential equations, traveling wave solutions have been extensively studied due to their favorable mathematical properties. The most classic one is the planar traveling wave solution. The planar wave solution refers to the solutions in which the level sets are hyperplanes parallel to each other, see [8,9,18]. To describe diverse phenomena in a multidimensional space, many scholars have begun to pay their attention on nonplanar traveling fronts, which have important applications in multi-dimensional curved flames [3] and multi-dimensional chemical waves [23,24]. For the reaction–diffusion equation (systems) (1.1) wtih f ( u , t ) = f ( u ) in a two-dimensional space, Bonnet and Hamel [3] established the existence of two-dimensional V-shaped traveling curved fronts for ignition equations. Ninomiya and Taniguchi [21,22] established the existence and global stability of traveling curved fronts in the Allen-Cahn equations. Wang and Bu [6,37] established the existence of and stability of V-shaped traveling fronts for reaction–diffusion equations with combustion and degenerate Fisher-KPP nonlinearities. Wang [36] studied the existence and stability of curved traveling fronts in bistable systems. In the three-dimensional space, Taniguchi [31,32] investigated the existence and stability of pyramidal traveling fronts in the Allen-Cahn equations. Taniguchi [33] also studied the traveling fronts with convex polyhedral shapes in bistable reaction–diffusion equations. Wang and Bu [37] obtained the existence of pyramidal traveling fronts in reaction–diffusion equations with combustion and degenerate Fisher-KPP nonlinearities. For more results about nonplanar traveling fronts, we refer to [7,1113,15]. Recently, the reaction–diffusion equations with time-periodic nonlinearities have attracted the attention of scholars. Wang and Wu [38] studied the existence and stability of periodic traveling curved fronts in reaction–diffusion equations with bistable time-periodic nonlinearity. Bao et al. [1] established the existence and stability of time-periodic traveling curved fronts in the periodic Lotka-Volterra competition-diffusion system. Recently, Zhang et al. [42] obtained the existence and stability of curved fronts in time-periodic reaction–diffusion equations with monostable nonlinearity. Sheng [29] and Sheng et al. [30] obtained the existence and stability of periodic pyramidal traveling fronts of bistable reaction–diffusion equations with time-periodic bistable nonlinearity. For more results about periodic traveling fronts, we refer to [2,4,14,2628,41].

Motivated by [5,41,42], this article studies the traveling fronts of reaction–diffusion equations with time-periodic degenerate monostable and ignition nonlinearities. We begin by giving some important estimates. In particular, we use the periodic planar traveling fronts and the periodic V-shaped traveling fronts to characterize some qualitative properties of the periodic pyramidal traveling fronts. Furthermore, using these estimates, along with the super-sub solution method and the comparison principle, we show the asymptotic stability of the periodic pyramidal traveling fronts. Different from the reaction–diffusion equations without periodicity, the presence of the time variable t in the nonlinear term f complicates the analysis. This is the biggest difficulty of this article. Before presenting the main results of this article, we first introduce some results on the planar traveling fronts and pyramidal traveling fronts.

According to [2,2628], equation (1.1) admits a periodic planar traveling front ϕ C 2,1 ( R 2 × R ) with the wave speed c * > 0 under assumptions (F1) and (F2), which satisfies ϕ ξ ( ξ , t ) > 0 ,

(1.2) ϕ t + c * ϕ ξ ϕ ξ ξ f ( ϕ , t ) = 0 , ( ξ , t ) R 2 , ϕ ( , t ) = 0 , ϕ ( + , t ) = 1 uniformly in t R , ϕ ( ξ , t + T ) = ϕ ( ξ , t ) , ( ξ , t ) R 2

and

(1.3) lim ξ ϕ ξ ( ξ , t ) ϕ ( ξ , t ) = c * > 0 , lim ξ ϕ ξ ξ ( ξ , t ) ϕ ( ξ , t ) = c * 2 uniformly in t R .

For β ( 0 , 1 ) , we can easily obtain that Π ( β c * ) = ( β c * ) 2 c * ( β c * ) < 0 . In addition, there exist positive constants L 1 , L 2 , L 3 , and β 1 such that

(1.4) L 1 e c * ξ ϕ ( ξ , t ) , ϕ ξ ( ξ , t ) , ϕ ξ ξ ( ξ , t ) L 2 e c * ξ , ξ < 0 , t R , ϕ ( ξ , t ) 1 , ϕ ξ ( ξ , t ) , ϕ ξ ξ ( ξ , t ) L 3 e β 1 ξ , ξ > 0 , t R .

The aim of this article is to investigate the stability of pyramidal traveling fronts of (1.1). Assuming that the traveling wave solution propagates along the z direction with speed c ( c > c * ) . Then, we can set u ( x , y , z , t ) = v ( x , y , z + c t , t ) = v ( x , y , z , t ) . For convenience, we denote v ( x , y , z , t ) simply as v ( x , y , z , t ) . By substituting v into (1.1), it follows that

(1.5) v t = v x x + v y y + v z z c v z + f ( v , t ) , x R 3 , t > 0 ,

(1.6) v ( x , 0 ) = v 0 ( x ) , x R 3 .

We write the solution of (1.5) and (1.6) as v ( x , t ; v 0 ) . A function V ( x , t ) is called a periodic traveling front with speed c if it satisfies

(1.7) [ V ] V t V x x V y y V z z + c V z f ( V , t ) = 0 , x R 3 , t R ,

(1.8) V ( x , y , z , t ) = V ( x , y , z , t + T ) , x R 3 , t > 0 .

Now, we introduce the notion of a pyramid. Let n 3 and consider { θ j } 1 j n taken in a counterclockwise order, satisfying θ j + 1 θ j < π for all 1 j n 1 and θ 1 θ n < π , that is,

cos θ j sin θ j + 1 cos θ j + 1 sin θ j > 0 , 1 j n 1 , cos θ n sin θ 1 cos θ 1 sin θ n > 0 .

Define m * = c 2 c * 2 c * . For each ( x , y ) R 2 , let

h j ( x , y ) = m * ( x cos θ j + y sin θ j ) , 1 j n , h ( x , y ) = max 1 j n h j ( x , y ) = m * max 1 j n ( x cos θ j + y sin θ j ) ,

then { x R 3 z = h ( x , y ) } is a pyramid in R 3 . The projections of the lateral surfaces of the pyramid onto the x y plane are denoted by

Ω j = { ( x , y ) R 2 h ( x , y ) = h j ( x , y ) } , j = 1 , 2 , , n .

The boundary of Ω j is denoted as Ω j . Set E = j = 1 n Ω j . The lateral surfaces of the pyramid are denoted as

S j = { ( x , y , z ) R 3 z = h j ( x , y ) , ( x , y ) Ω j } , j = 1 , 2 , , n .

The edges of the pyramid are represented as

Γ j = S j S j + 1 , 1 j n 1 , S n S 1 , j = n .

And Γ = j = 1 n Γ j represents the set of all edges of the pyramid. Define

D ( γ ¯ ) = { x R 3 dist ( x , Γ ) γ ¯ } , γ ¯ > 0

and

V ̲ ( x , t ) = max 1 j n ϕ c * c ( z + h j ( x , y ) ) , t = ϕ c * c ( z + h ( x , y ) ) , t .

It is easy to verify that V ̲ is a subsolution of equation (1.5).

The existence of time-periodic pyramidal traveling fronts for reaction–diffusion equations with time-periodic degenerate monostable nonlinearity has been established by Bu et al. [4]. For the case of time-periodic ignition nonlinearity, one can easily check with a similar proof. More precisely, we have the following theorem.

Theorem 1.1

[4, Theorem 1.1] Assume that (F1) and (F2) hold. For each c > c * , equation (1.1) admits a time-periodic nonplanar traveling front V ( x , t ) satisfying (1.5) and (1.6). Moreover,

(1.9) lim γ ¯ + sup x D ( γ ¯ ) , t [ 0 , T ] V ( x , t ) V ̲ ( x , t ) V ̲ β ( x , t ) = 0 , β ( 0 , 1 )

and

V z ( x , t ) > 0 , x R 3 , t R .

Now, we give the main result of this article.

Theorem 1.2

Let β 0 , 1 8 . Assume that the initial value v 0 ( x ) C ( R 3 , [ 0 , 1 ] ) satisfies

v 0 ( x ) V ̲ ( x , 0 ) , x R 3

and

(1.10) lim γ ¯ + sup x D ( γ ¯ ) v 0 ( x ) V ̲ ( x , 0 ) V ̲ β ( x , 0 ) = 0 ,

then the solution v ( x , t ; v 0 ) of (1.5) and (1.6) satisfies

lim k + sup x R 3 v ( x , t + k T ; v 0 ) V ( x , t ) V ̲ β ( x , t ) = 0 .

2 Preliminaries

In this section, we establish some key results. First, we present the existence and stability of V-shaped traveling fronts for time-periodic reaction–diffusion equations in R 2 with ignition and degenerate monostable nonlinearities, which were proved by Zhang et al. [41,42]. Let w ( ξ , η , t ; w 0 ) be the solution of

(2.1) w t w ξ ξ w η η + s w η f ( w , t ) = 0 , ( ξ , η ) R 2 , t > 0 ,

(2.2) w ( ξ , η , 0 ) = w 0 ( ξ , η ) , ( ξ , η ) R 2

for s > c * . V * ( ξ , η , t ; s ) called a periodic traveling front with speed s > c * if

(2.3) ( V * ) t ( V * ) ξ ξ ( V * ) η η + s ( V * ) η f ( V * , t ) = 0 , ( ξ , η ) R 2 , t > 0 ,

(2.4) V * ( ξ , η , t ) = V * ( ξ , η , t + T ) , ( ξ , η ) R 2 , t > 0 .

It is clear that ϕ c * s η ± s 2 c * 2 c * ξ , t satisfy (2.3) and (2.4), and hence

V ̲ * ( ξ , η , t ) max ϕ c * s η s 2 c * 2 c * ξ , t , ϕ c * s η + s 2 c * 2 c * ξ , t = ϕ c * s η + s 2 c * 2 c * ξ , t

is a subsolution of (2.1). In addition, we have ( V ̲ * ) η ( ξ , η , t ) > 0 , V ̲ * ( ξ , η , t ) = V ̲ * ( ξ , η , t + T ) , and V ̲ * ( ξ , η , t ) is strictly monotone increasing in η . The following two results are the existence and stability of such traveling fronts.

Theorem 2.1

[41, Theorem 1.2], [42, Theorem 1.1] Assume that ( F 1 ) and ( F 2 ) hold. Then, for any s > c * , there exists a curved front V * ( ξ , η , t ) satisfying (2.3) and (2.4), V * ( ξ , η , t ) > V ̲ * ( ξ , η , t ) and

lim R + sup ξ 2 + η 2 > R 2 , t [ 0 , T ] V * ( ξ , η , t ) V ̲ * ( ξ , η , t ) ( V ̲ * ( ξ , η , t ) ) β = 0 , β ( 0 , 1 ) .

Theorem 2.2

[41, Theorem 1.4], [42, Theorem 1.3] Assume that ( F 1 ) and ( F 2 ) hold. Let s > c * . Let V ( ξ , η , t ) be the curved front satisfying (2.3) and (2.4). If w 0 ( ξ , η ) C ( R 2 , [ 0 , 1 ] ) satisfies

lim R + sup ξ 2 + η 2 > R 2 , t [ 0 , T ] w 0 ( ξ , η ) V ̲ * ( ξ , η , 0 ) V ̲ * β ( ξ , η , 0 ) = 0

for any β ( 0 , 1 8 ) , then the solution w ( ξ , η , t ; w 0 ) of (2.1) and (2.2) satisfies

lim t + sup ( ξ , η ) R 2 w ( ξ , η , t ; w 0 ) V * ( ξ , η , t ) V ̲ * β ( ξ , η , t ) = 0 .

Second, we introduce the mollified pyramid and present some of its properties. Set ρ ˜ ( r ) C [ 0 , + ) satisfies

ρ ˜ ( r ) > 0 , ρ ˜ r ( r ) 0 if r 0 , ρ ˜ ( r ) = 1 if 0 r 1 2 , ρ ˜ ( r ) = e r if r > 0 large enough , R 2 ρ ˜ ( x 2 + y 2 ) d x d y = 1 .

Besides, we have R 2 ρ ˜ ( x 2 + y 2 ) d x d y = 2 π 0 + r ρ ˜ ( r ) d r = 1 . Let ρ ( x , y ) = ρ ˜ ( x 2 + y 2 ) , then ρ C ( R 2 ) and R 2 ρ ( x , y ) d x d y = 1 . For all non-negative integers i 1 0 and i 2 0 satisfying 0 i 1 + i 2 3 , there exists a constant M * > 0 such that

D x i 1 D y i 2 ρ ( x , y ) M * ρ ( x , y ) , ( x , y ) R 2 ,

where D x i 1 = i 1 x i 1 and D y i 1 = i 2 y i 2 . Now, we define a function φ ( x , y ) as follows:

(2.5) φ ( x , y ) = R 2 ρ ( x x , y y ) h ( x , y ) d x d y = R 2 ρ ( x , y ) h ( x x , y y ) d x d y .

The set { x R 3 z = φ ( x , y ) } is called the mollifed pyramid of pyramid { x R 3 z = h ( x , y ) } . Let

(2.6) S ( x , y ) = c 1 + φ ( x , y ) 2 c * ,

where

φ ( x , y ) = ( φ x ( x , y ) , φ y ( x , y ) ) , φ ( x , y ) = φ x 2 ( x , y ) + φ y 2 ( x , y ) .

The following two lemmas were shown by Taniguchi [31], which give some important properties of the functions φ and S .

Lemma 2.3

Let φ and S be as in (2.5) and (2.6), respectively. For any fixed integers i 1 0 , i 2 0 , and i 1 + i 2 3 , one has

sup ( x , y ) R 2 D x i 1 D y i 2 φ ( x , y ) < K 1 for s o m e c o n s t a n t K 1 > 0 , h ( x , y ) < φ ( x , y ) h ( x , y ) + 2 π m * 0 + r 2 ρ ˜ ( r ) d r , φ ( x , y ) < m * , 0 < S ( x , y ) c c * , ( x , y ) R 2

and

lim λ + sup { S ( x , y ) ( x , y ) R 2 , dist ( ( x , y ) , E ) λ } = 0 , lim λ + sup { φ ( x , y ) h ( x , y ) ( x , y ) R 2 , dist ( ( x , y ) , E ) λ } = 0 .

Lemma 2.4

There exist two positive constants a 1 and a 2 such that

a 1 = inf ( x , y ) R 2 φ ( x , y ) h ( x , y ) S ( x , y ) sup ( x , y ) R 2 φ ( x , y ) h ( x , y ) S ( x , y ) = a 2 .

In addition, for integers i 1 0 and i 2 0 with 2 i 1 + i 2 3 , there exists a constant K 2 > 0 such that

sup ( x , y ) R 2 D x i 1 D y i 2 φ ( x , y ) S ( x , y ) < K 2

and

φ x x ( x , y ) , φ y y ( x , y ) m * M * , ( x , y ) R 2 .

Third, we consider the following equations:

Φ ¯ ( t ) f u ( 1 , t ) Φ ¯ ( t ) = Λ 0 Φ ¯ ( t ) , t > 0 , Φ ¯ ( t + T ) = Φ ¯ ( t ) , t > 0 ,

where Λ 0 = 1 T 0 T f u ( 1 , s ) d s > 0 . By a direct calculation, we have

Φ ¯ ( t ) = e Λ 0 t + 0 T f u ( 1 , s ) d s .

We define ν ( t ) = K 0 Φ ¯ ( t ) , P 1 = min t [ 0 , T ] ν ( t ) , and P 2 = max t [ 0 , T ] ν ( t ) , where K 0 is a positive constant such that P 1 > 1 .

Fourth, we present some properties of the reaction term f . By assumptions (F1) and (F2), we can choose ε 1 ( 0 , 1 ) small enough such that

(2.7) f u ( u , t ) f u ( 1 , t ) 1 2 Λ 0 , 1 ε 1 u 1 + ε 1 , t > 0 .

Besides, by assumption (F1), we know that there exists a positive constant K ¯ 1 such that

(2.8) f u ( u 1 , t ) f u ( u 2 , t ) K ¯ 1 u 1 u 2 P , u 1 , u 2 [ 0 , 1 ] , t R .

Finally, we construct an auxiliary function ω ( x ) C ( R ) such that

ω ( x ) = 1 if x 1 , 0 < ω ( x ) < 1,0 < ω ( x ) < 1 if 1 < x < 1 , ω ( x ) = 0 if x 1 ,

which will be used in constructing supersolutions.

3 Key estimates

In this section, we provide some important estimates that will be used in the proof of asymptotic stability. First, we show that for any ξ and t , the planar traveling wave solution ϕ ( ξ , t ) and its initial value ϕ ( ξ , 0 ) can be controlled by each other.

Lemma 3.1

Let ϕ be a planar traveling front of equation (1.1) satisfying (1.2) and (1.3). Then, there exist constants a , b > 0 such that

a ϕ ( ξ , 0 ) ϕ ( ξ , t ) b ϕ ( ξ , 0 ) , ξ R , t R .

Proof

By the periodicity of ϕ , we only need to consider t [ 0 , T ] . Let g ( ξ , t ) ln ϕ ( ξ , t ) , then there exists θ ( 0 , 1 ) such that

ln ϕ ( ξ , t ) ϕ ( ξ , 0 ) = g ( ξ , t ) g ( ξ , 0 ) = g t ( ξ , θ t ) t .

By (1.2), we have

g t ( ξ , t ) = ϕ t ( ξ , t ) ϕ ( ξ , t ) = ϕ ξ ξ ( ξ , t ) c * ϕ ξ ( ξ , t ) + f ( ϕ ( ξ , t ) , t ) ϕ ( ξ , t ) .

From (1.3) and the fact that

lim ξ + ϕ ξ ( ξ , t ) ϕ ( ξ , t ) = 0 , lim ξ + ϕ ξ ξ ( ξ , t ) ϕ ( ξ , t ) = 0 , lim ϕ 1 f ( ϕ ( ξ , t ) , t ) ϕ ( ξ , t ) = 0 , lim ϕ 0 f ( ϕ ( ξ , t ) , t ) ϕ ( ξ , t ) = lim ϕ 0 f ϕ ( ϕ ( ξ , t ) , t ) 1 = 0 ,

we know that there exists a constant C such that g t ( ξ , t ) C . Thus, we have

ln ϕ ( ξ , t ) ϕ ( ξ , 0 ) C T .

Define a = e C T , b = e C T , then

a ϕ ( ξ , 0 ) ϕ ( ξ , t ) b ϕ ( ξ , 0 ) .

This completes the proof.□

From the definition of V ̲ , similar to Lemma 3.1, we can deduce the following corollary.

Corollary 3.2

There exist constants a , b > 0 , such that

a V ̲ ( x , 0 ) V ̲ ( x , t ) b V ̲ ( x , 0 ) , x R 3 , t R .

Set

m max 1 v 2 , t [ 0 , T ] f v ( v , t ) .

Let F ( x , t ) be a given continuous function with

(3.1) sup x R 3 , t > 0 F ( x , t ) m .

We consider the following Cauchy problem:

(3.2) W t N W F ( x , t ) W = 0 , x R 3 , t > 0 ,

(3.3) W ( x , 0 ) = W 0 ( x ) , x R 3 ,

where

N W W x x + W y y + W z z c W z .

Let

N 1 sup ζ R , t [ 0 , T ] ϕ ζ ( ζ , t ) ϕ ( ζ , t ) , A ˆ = m * max 1 j n cos θ j , B ˆ = m * max 1 j n sin θ j

and

K max { A ˆ , B ˆ , 1 } , N ˆ β N 1 c * + β 2 N 1 2 c * 2 c 2 + β 2 N 1 2 c * 2 c 2 A ˆ 2 + β 2 N 1 2 c * 2 c 2 B ˆ 2 .

Lemma 3.3

Let W ( x , t ) be the solution of (3.2) and (3.3). Assume β ( 0 , 1 ) and the initial value W 0 ( x ) satisfies sup R 3 W 0 ( x ) V ̲ β ( x , 0 ) < + . Then, there exists a constant M > 0 such that for any γ ¯ > 0 ,

(3.4) sup D ( 2 γ ¯ ) W ( x , t ) V ̲ β ( x , t ) 8 M β e N ¯ t sup D ( γ ¯ ) W 0 ( x ) V ̲ β ( x , 0 ) + 3 M β e N ¯ t erfc 3 3 γ ¯ K β N 1 c * c t c t 4 t sup D ( γ ¯ ) c W 0 ( x ) V ̲ β ( x , 0 ) ,

where N ¯ N ˆ + m , erfc ( z ) 2 π z + e τ 2 d τ , and D ( γ ¯ ) c denotes the complementary set of D ( γ ¯ ) . In particular, if W 0 ( x ) = 0 in B ( x 0 , 3 R ) , then

(3.5) W ( x 0 , t ) V ̲ β ( x 0 , t ) 3 M β e N ¯ t erfc R K β N 1 c * c t c t 4 t sup R 3 W 0 ( x ) V ̲ β ( x , 0 )

for any x 0 R 3 and R > 0 .

Proof

Let w ˜ ( x , t ) = W ( x , t ) e m t , then we have

w ˜ t N w ˜ + [ m F ( x , t ) ] w ˜ = 0 , x R 3 , t > 0 , w ˜ ( x , 0 ) = W 0 ( x ) , x R 3 .

Let w ˆ ( x , t ) be the solution of the following equation:

w ˆ t N w ˆ = 0 , x R 3 , t > 0 , w ˆ ( x , 0 ) = W 0 ( x ) , x R 3 .

By (3.1) and the maximum principle, we have

(3.6) w ˜ ( x , t ) w ˆ ( x , t ) , x R 3 , t > 0 .

Now, we estimate w ˆ ( x , t ) . It is obvious that

w ˆ ( x , t ) = R 3 Γ ¯ ( x , t ) W 0 ( x x ) d x = R 3 Γ ¯ ( x x , t ) W 0 ( x ) d x

for x R 3 and t > 0 , where

Γ ¯ ( x , t ) = 1 ( 4 π t ) 3 2 exp x 2 + y 2 + ( z c t ) 2 4 t .

Define

h * ( x , y ) max 1 j n ( m * x cos θ j + y sin θ j )

and

Σ z z + A ˆ x x + B ˆ y y .

Clearly,

h * ( x , y ) A ˆ x + B ˆ y , ( x , y ) R 2 .

Since ϕ ( x + y , t ) e N 1 y is decreasing in y , one has ϕ ( x + y , t ) e N 1 y ϕ ( x , t ) for any y 0 . Following from Corollary 3.2, we can define M sup x R 3 , t [ 0 , T ] V ̲ β ( x , 0 ) V ̲ β ( x , t ) . Then, we have

(3.7) w ˆ ( x , t ) V ̲ β ( x , t ) = R 3 Γ ¯ ( x , t ) W 0 ( x x ) V ̲ β ( x , t ) d x = R 3 Γ ¯ ( x , t ) V ̲ β ( x x , t ) V ̲ β ( x , t ) V ̲ β ( x x , 0 ) V ̲ β ( x x , t ) W 0 ( x x ) V ̲ β ( x x , 0 ) d x = R 3 Γ ¯ ( x , t ) ϕ β c * c ( z z + h ( x x , y y ) ) , t ϕ β c * c ( z + h ( x , y ) ) , t V ̲ β ( x x , 0 ) V ̲ β ( x x , t ) W 0 ( x x ) V ̲ β ( x x , 0 ) d x R 3 Γ ¯ ( x , t ) ϕ β c * c ( z + z + h ( x , y ) + h * ( x , y ) ) , t ϕ β c * c ( z + h ( x , y ) ) , t V ̲ β ( x x , 0 ) V ̲ β ( x x , t ) W 0 ( x x ) V ̲ β ( x x , 0 ) d x R 3 Γ ¯ ( x , t ) M β exp β N 1 c * c ( z + h * ( x , y ) ) W 0 ( x x ) V ̲ β ( x x , 0 ) d x = R 3 Γ ¯ ( x x , t ) M β exp β N 1 c * c ( z z + h * ( x x , y y ) ) W 0 ( x ) V ̲ β ( x , 0 ) d x R 3 Γ ¯ ( x x , t ) M β exp β N 1 c * c Σ W 0 ( x ) V ̲ β ( x , 0 ) d x = R 3 Γ ¯ ( x x , t ) M β exp β N 1 c * c Σ W 0 ( x ) χ D ( γ ¯ ) V ̲ β ( x , 0 ) d x + R 3 Γ ¯ ( x x , t ) M β exp β N 1 c * c Σ W 0 ( x ) χ D ( γ ¯ ) c V ̲ β ( x , 0 ) d x M β sup D ( γ ¯ ) W 0 ( x ) V ̲ β ( x , 0 ) R 3 Γ ¯ ( x x , t ) exp β N 1 c * c Σ d x + M β sup D ( γ ¯ ) c W 0 ( x ) V ̲ β ( x , 0 ) D ( γ ¯ ) c Γ ¯ ( x x , t ) exp β N 1 c * c Σ d x .

For any x D ( 2 γ ¯ ) , it follows (3.7) that

w ˆ ( x , t ) V ̲ β ( x , t ) M β sup D ( γ ¯ ) W 0 ( x ) V ̲ β ( x , 0 ) R 3 Γ ¯ ( x , t x ) exp β N 1 c * c Σ d x + M β sup D ( γ ¯ ) c W 0 ( x ) V ̲ β ( x , 0 ) B ( x , γ ¯ ) c Γ ¯ ( x , t x ) exp β N 1 c * c Σ d x = M β I sup D ( γ ¯ ) W 0 ( x ) V ̲ β ( x , 0 ) + M β I I sup D ( γ ¯ ) c W 0 ( x ) V ̲ β ( x , 0 ) ,

where

I R 3 Γ ¯ ( x x , t ) exp β N 1 c * c Σ d x , I I B ( x , γ ¯ ) c Γ ¯ ( x x , t ) exp β N 1 c * c Σ d x .

Same as the proof in [5, Lemma 3.1], we have

I 8 e N ˆ t , I I 3 e N ˆ t erfc 3 3 γ ¯ K β N 1 c * c t c t 4 t .

Then, we obtain

sup D ( 2 γ ¯ ) w ˆ ( x , t ) V ̲ β ( x , t ) 8 M β e N ˆ t sup D ( γ ¯ ) W 0 ( x ) V ̲ β ( x , 0 ) + 3 M β e N ˆ t erfc 3 3 γ ¯ K β N 1 c * c t c t 4 t sup D ( γ ¯ ) c W 0 ( x ) V ̲ β ( x , 0 ) .

w ˜ ( x , t ) = W ( x , t ) e m t and (3.6) yield that (3.4) holds.

Substituting x with x 0 and D ( γ ¯ ) with B ( x 0 , 3 R ) in (3.7), we immediately obtain (3.5). This completes the proof.□

In the following, we will show that the time-periodic pyramidal traveling front converges to the two-dimensional time-periodic V-shaped traveling front along the edges at infinity. More precisely, for a time-periodic pyramidal traveling front, as we move along any edge toward infinity, the function values on the cross-sections perpendicular to this edge converge in a certain manner to some two-dimensional time-periodic V-shaped traveling front. For this purpose, we first investigate the behavior of the two-dimensional time-periodic V-shaped traveling front on each edge. Consider the edge Γ j and the adjacent faces S j and S j + 1 forming a dihedral angle. We want to place the V-shaped traveling fronts solution U i onto this dihedral angle. For all 1 j n , let n ˆ j 1 1 + m * 2 ( m * cos θ j , m * sin θ j , 1 ) be the normal vector of the surface S j , then the direction of Γ j = S j S j + 1 is given by

n ˆ j + 1 × n ˆ j = 1 q j 2 + m * 2 p j 2 sin θ j + 1 sin θ j cos θ j cos θ j + 1 m * sin ( θ j θ j + 1 ) ,

where

p j sin ( θ j + 1 θ j ) > 0 and q j ( cos θ j + 1 cos θ j ) 2 + ( sin θ j + 1 sin θ j ) 2 > 0 , 1 j n .

Thus, the direction of traveling fronts U i is given by

n ˆ j + 1 n ˆ j n ˆ j + 1 n ˆ j × ( n ˆ j + 1 × n ˆ j ) = 1 q j q j 2 + m * 2 p j 2 m * ( sin θ j sin θ j + 1 ) p j m * ( cos θ j + 1 cos θ j ) p j q j 2 .

Let s j be the speed of U j and 2 θ ˆ j ( θ ˆ j ( 0 , π 2 ) ) be the angle between S j and S j + 1 . Then, we obtain

s j sin θ ˆ j = c * .

The angle between n ˆ j and n ˆ j + 1 n ˆ j n ˆ j + 1 n ˆ j × ( n ˆ j + 1 × n ˆ j ) equals π 2 θ ˆ j . We obtain

sin θ ˆ j = q j 2 + m * 2 p j 2 q j 1 + m * 2

and thus

s j = c q j q j 2 + m * 2 p j 2 .

The speed of U j toward the z -axis equals

q j 2 + m * 2 p j 2 q j s j = c * 1 + m * 2 = c ,

which coincides with the speed of V . Next, we use the change of variables as follows, let

R j cos θ j + 1 cos θ j q j m * ( sin θ j + 1 sin θ j ) p j q j q j 2 + m * 2 p j 2 sin θ j + 1 sin θ j q j 2 + m * 2 p j 2 sin θ j + 1 sin θ j q j m * ( cos θ j cos θ j + 1 ) p j q j q j 2 + m * 2 p j 2 cos θ j cos θ j + 1 q j 2 + m * 2 p j 2 0 q ˙ i q j 2 + m * 2 p j 2 m * p j 2 q j 2 + m * 2 p j 2

and

x y z = R j ξ η ζ , ξ η ζ = R j T x y z ,

where R j T is the transposed matrix of R j . Then

ξ = ( cos θ j + 1 cos θ j ) x + ( sin θ j + 1 sin θ j ) y q j , η = m * ( sin θ j + 1 sin θ j ) p j x + m * ( cos θ j cos θ j + 1 ) p j y + q j 2 z q j q j 2 + m * 2 p j 2 ,

ζ = ( sin θ j + 1 sin θ j ) x + ( cos θ j cos θ j + 1 ) y m * p j z q j 2 + m * 2 p j 2 .

Now, let U j be as

U j ( x , y , z , t ) V * ( ξ , η , t ; s j ) .

By [41, Lemma 4.2] and [42, Lemma 4.1], we obtain U j ( x , t ) , which is strictly monotone increasing in z for each 1 j n . Through direct computation, we obtain

[ U j ] = ( U j ) t ( U j ) x x ( U j ) y y ( U j ) z z + c ( U j ) z f ( U j ( x , y , z , t ) , t ) = ( V * ) t ( V * ) ξ ξ ( V * ) η η + s j ( V * ) η f ( V * ( ξ , η , t ; s j ) , t ) = 0 , U j ( x , y , z , t ) = V * ( ξ , η , t ; s j ) = V * ( ξ , η , t + T ; s j ) = U j ( x , y , z , t + T )

in R 3 . Thus, the function U j ( x , t ) satisfies equations (1.7) and (1.8) for each 1 j n . Set

Q j { x R 3 dist ( x , Γ ) = dist ( x , Γ j ) } for 1 j n .

Then, we have R 3 = j = 1 n Q j . Define

U ^ ( x , t ) max 1 j n U j ( x , t ) .

Since each U j is strictly monotone increasing in z , it follows that U ˆ j is also strictly monotone increasing in z .

The following lemma establishes some properties of U ^ . We first recall a result from [4, Theorem 3.2], that is

(3.8) V ( x , t ) = lim k + v ( x , t + k T ; V ̲ 0 ) , x R 3 , t 0 ,

where V ̲ 0 V ̲ ( x , 0 ) .

Lemma 3.4

The function U ^ ( x , t ) satisfies

V ̲ ( x , t ) < U ^ ( x , t ) < V ( x , t ) , x R 3 , t 0 ,

and

(3.9) lim γ ¯ + sup x D ( γ ¯ ) , t 0 U ^ ( x , t ) V ̲ ( x , t ) V ̲ β ( x , t ) = 0 , β ( 0 , 1 ) .

Proof

Theorem 2.1 implies that for each 1 j n , we have

max ϕ c * c ( z + h j ( x , y ) ) , t , ϕ c * c ( z + h j + 1 ( x , y ) ) , t < U j ( x , t ) .

Thus,

V ̲ ( x , t ) = ϕ c * c ( z + h ( x , y ) ) , t = max 1 j n ϕ c * c ( z + h j ( x , y ) ) , t < U ^ ( x , t ) .

Take max ϕ c * c ( z + h j ( x , y ) ) , 0 , ϕ c * c ( z + h j + 1 ( x , y ) ) , 0 , and ϕ c * c ( z + h ( x , y ) ) , 0 as the initial value of (1.5) and (1.6), respectively. Letting t + , by the comparison principle, we have

U j ( x , t ) < V ( x , t ) , x R 3 .

Consequently,

U ^ ( x , t ) < V ( x , t ) , x R 3 .

Then, it follows from Theorem 1.1, we obtain that (3.9) holds. This completes the proof.□

The following lemma demonstrates that v ( x , t ; v 0 ) approaches to U ^ ( x , t ) in some sense as time evolves.

Theorem 3.5

Assume that the v 0 C ( R 3 , [ 0 , 1 ] ) satisfies

lim γ ¯ + sup x D ( γ ¯ ) v 0 ( x ) V ̲ ( x , 0 ) V ̲ β ( x , 0 ) = 0

for some β 0 , 1 8 . Then, for any given ε > 0 , t [ 0 , T ] , there exists K ¯ N large enough such that

lim R + sup x R v ( x , t + k T ; v 0 ) U ^ ( x , t ) V ̲ β ( x , t ) < ε for a n y f i x e d i n t e g e r k K ¯ .

Proof

Define

I j Ω j Ω j + 1 = τ cos θ j + cos θ j + 1 sin θ j + sin θ j + 1 τ 0 for 1 j n 1 , I n Ω n Ω 1 = τ cos θ n + cos θ 1 sin θ n + sin θ 1 τ 0 .

Then, I j represents the projection of Γ j onto the x y plane, and j = 1 n I j represents the projection of Γ onto the x y plane. We divide our proof into two steps.

Step 1. For any given ε > 0 , t [ 0 , T ] , we will show that there exists a sufficiently large constant K ¯ N such that for any integer k > K ¯ , there exists a constant R ˆ > 0 satisfying

max 1 j n sup x Q j , dist ( x , Q j ) R ˆ v ( x , t + k T ; v 0 ) U j ( x , t ) V ̲ β ( x , t ) < ε .

Without loss of generality, we can assume x Q j for some j ( 1 j n ) as x + . Since x 2 + y 2 is invariant under rotations on the x y plane, we can assume Ω j Ω j + 1 = { ( 0 , y ) y 0 } and

( cos θ j , sin θ j ) = ( A , B ) , ( cos θ j + 1 , sin θ j + 1 ) = ( A , B )

with A > 0 , B > 0 and A 2 + B 2 = 1 . In this simplified coordinate system, S j and S j + 1 are z = m * ( A x + B y ) and z = m * ( A x + B y ) , respectively. The common line Γ j of them is { x = 0 , z = m * B y } . The projection of Q j onto the x y plane is given by { y a * x , x 0 } { y b * x , x 0 } , where a > 0 and b > 0 are constants. We denote Q j simply by Q hereafter. The unit normal vector of this common line directing upward and lying on { x = 0 } is given by

1 1 + m * 2 B 2 0 m * B 1 .

Recall that 2 θ ˆ j θ ˆ j 0 , π 2 is the angle between S j and S j + 1 . We write θ ˆ j simply by θ ˆ . Then, one has

1 + m * 2 B 2 1 + m * 2 = sin θ ˆ .

The change of variables is given by

(3.10) x y z = 1 0 0 0 m * B 1 + m * 2 B 2 1 1 + m * 2 B 2 0 1 1 + m * 2 B 2 m * B 1 + m * 2 B 2 ξ η ζ .

The coefficient matrix in (3.10) is an orthogonal matrix of determinant +1. Then, we have

ξ = x , η = z + m * B y 1 + m * 2 B 2 , ζ = y m * B z 1 + m * 2 B 2 .

Set s j c 1 + m * 2 B 2 . For this change of variables, we have

V ^ * j ( ξ , η , t ) ϕ c * s j η + s j 2 c * 2 c * ξ , t = ϕ c * c ( z + m * B y + m * A x ) , t V ^ ̲ j ( x , t ) .

It is clear that V ̲ ( x , t ) = V ^ ̲ j ( x , t ) for x Q . Hereafter, we write s j , V ^ * j , V ^ ̲ j , and U j simply by s , V ^ * , V ^ ̲ , and U , respectively. Obviously,

U ( x , t ) = V * ( ξ , η , t ; s ) = V * x , z + m * B y 1 + m * 2 B 2 , t ; s

is a solution to (1.7) and (1.8). Let V ˜ ( ξ , η , t ) = V ˜ ( ξ , η , t ; V ˜ 0 ) be the solution of the following equation:

(3.11) V ˜ t V ˜ ξ ξ V ˜ η η + s V ˜ η f ( V ˜ , t ) = 0 , ( ξ , η ) R 2 , t > 0 , V ˜ ( ξ , η , 0 ) = V ˜ 0 ( ξ , η ) , ( ξ , η ) R 2 .

Then, by setting

V ^ 0 ( x ) = V ˜ 0 ( ξ , η ) ,

we have V ^ ( x , t ; V ^ 0 ) = V ˜ ( ξ , η , t ; V ˜ 0 ) satisfies

(3.12) V ^ t + [ V ^ ] = 0 , x R 3 , t > 0 , V ^ ( x , 0 ) = V ^ 0 ( x ) , x R 3 .

According to Lemma 3.4, we have

U ^ ( x , t ) < V ( x , t ) , x R 3 , t > 0 .

Following from Lemma 3.4, Theorem 1.1, and the assumption on v 0 , we obtain

lim γ ¯ + sup x D ( γ ¯ ) Q v 0 ( x , 0 ) U ( x , 0 ) V ̲ β ( x , 0 ) = 0 , β 0 , 1 8 .

By the definition of Q , we have

(3.13) dist ( x , Γ ) = dist ( x , Γ j ) = ( 1 + m * 2 B 2 ) x 2 + ( z + m * B y ) 2 1 + m * 2 B 2 , x Q .

Choose a function g C ( R ) L ( R ) and a constant M ¯ > 0 with

g ( γ ¯ ) = sup x D ( γ ¯ ) Q v 0 ( x ) U ( x , 0 ) V ̲ β ( x , 0 ) for γ ¯ 1 , sup x D ( γ ¯ ) Q v 0 ( x ) U ( x , 0 ) V ̲ β ( x , 0 ) g ( γ ¯ ) M ¯ , g ( γ ¯ ) 0 for 0 < γ ¯ < 1 , g ( γ ¯ ) = g ( γ ¯ ) for γ ¯ R .

Then, g ( γ ¯ ) is monotonically non-increasing in γ ¯ > 0 and satisfies lim γ ¯ + g ( γ ¯ ) = 0 . It follows from (3.13), we have

v 0 ( x , 0 ) U ( x , 0 ) V ̲ β ( x , 0 ) g ( dist ( x , Γ ) ) = g ( 1 + m * 2 B 2 ) x 2 + ( z + m * B y ) 2 1 + m * 2 B 2 , x Q .

We consider (3.11) for

V ˜ 0 + ( ξ , η ) min { V * ( ξ , η , 0 ; s ) + g ( ξ 2 + η 2 ) V ^ * β ( ξ , η ) , 1 } , V ˜ 0 ( ξ , η ) max { V * ( ξ , η , 0 ; s ) g ( ξ 2 + η 2 ) V ^ * β ( ξ , η ) , 0 } ,

which is equivalent to consider (3.12) for

V ^ 0 + ( x , 0 ) min U ( x , 0 ) + g ( 1 + m * 2 B 2 ) x 2 + ( z + m * B y ) 2 1 + m * 2 B 2 V ^ ̲ β ( x , 0 ) , 1 , V ^ 0 ( x , 0 ) max U ( x , 0 ) g ( 1 + m * 2 B 2 ) x 2 + ( z + m * B y ) 2 1 + m * 2 B 2 V ^ ̲ β ( x , 0 ) , 0 ,

respectively. It follows from the definition of g that

lim R + sup ξ 2 + η 2 R 2 V ˜ 0 ± ( ξ , η ) V * ( ξ , η , 0 ; s ) V ^ * β ( ξ , η , 0 ) = 0 .

By Theorem 2.1, we have

lim R + sup ξ 2 + η 2 R 2 V ˜ 0 ± ( ξ , η ) V ^ * ( ξ , η , 0 ) V ^ * β ( ξ , η , 0 ) = 0 .

Applying Theorem 2.2, we obtain

lim k + V ˜ ( ξ , η , t + k T ; V ˜ 0 ± ) V * ( ξ , η , t ; s ) V ^ * β ( ξ , η , t ) L ( R 2 ) = 0 ,

which implies

lim k + V ^ ( x , y , z , t + k T ; V ^ 0 ± ) U ( x , y , z , t ) V ^ ̲ β ( x , y , z , t ) L ( R 3 ) = 0 .

Since V ^ ̲ ( x , t ) V ̲ ( x , t ) for all x R 3 ,

lim k + V ^ ( x , y , z , t + k T ; V ^ 0 ± ) U ( x , y , z , t ) V ̲ β ( x , y , z , t ) L ( R 3 ) = 0 .

Thus, for any given ε > 0 , there exists a constant K ¯ j R large enough such that

(3.14) sup k K ¯ j V ^ ( x , y , z , t + k T ; V ^ 0 ± ) U ( x , y , z , t ) V ̲ β ( x , y , z , t ) L ( R 3 ) < ε 2 .

Set

v ± ( x , t ) v ( x , t ; v 0 ) V ^ ( x , t ; V ^ 0 ± ) .

Then, v ± ( x , t ) satisfies

t N 0 1 f ( τ v ( x , t ; v 0 ) + ( 1 τ ) V ^ ( x , t ; V ^ 0 ± ) ) d τ v ± ( x , t ) = 0 , x R 3 , t > 0 , v ± ( x , 0 ) = v 0 ( x ) V ^ 0 ± ( x ) , x R 3 .

Since V ^ ̲ ( x , t ) = V ̲ ( x , t ) for x Q , by (3.13), we obtain

v + ( x , 0 ) 0 , v ( x , 0 ) 0 , x Q .

Let v ˜ ± ( x , t ) be given by

t N 0 1 f ( τ v ( x , t ; v 0 ) + ( 1 τ ) V ^ ( x , t ; V ^ 0 ± ) ) d τ v ˜ ± ( x , t ) = 0 , v ˜ + ( x , 0 ) = max { v + ( x , 0 ) , 0 } and v ˜ ( x , 0 ) = max { v ( x , 0 ) , 0 } .

Clearly, v ˜ + ( x , 0 ) v + ( x , 0 ) and v ˜ ( x , 0 ) v ( x , 0 ) for all x R 3 . Applying the comparison principle, we have

(3.15) v ˜ + ( x , t ) v + ( x , t ) and v ˜ ( x , t ) v ( x , t ) , ( x , t ) R 3 × ( 0 , + ) .

Note that v ˜ ± ( x , 0 ) V ̲ β ( x , 0 ) is bounded for x R 3 and v ˜ ± ( x , 0 ) = 0 for x Q . Applying Lemma 3.3 to v ˜ ± ( x , t ) , we obtain

0 v ˜ ± ( x , t ) V ̲ β ( x , t ) 3 M β e N ¯ t erfc R K β N 1 c * c t c t 4 t sup R 3 v ˜ ± ( x , 0 ) V ̲ β ( x , 0 ) , t > 0

if x Q and dist ( x , Q ) > 3 R . Thus

lim R + sup x Q , dist ( x , Q ) R v ˜ ± ( x , t ) V ̲ β ( x , t ) = 0

for any fixed t > 0 . Applying this equality, (3.14) and (3.15) to

v ( x , t ; v 0 ) = v ± ( x , t ) + V ^ ( x , t ; V ^ 0 ± )

for any fixed integer k > K ¯ j , we can choose R ˆ j > 0 large enough such that

sup x Q , dist ( x , Q ) R ˆ j v ( x , t + k T ; v 0 ) U ( x , t ) V ̲ β ( x , t ) < ε .

We have now established the estimate for Q , that is, Q j for some j . Let

K ¯ max { K ¯ 1 , K ¯ 2 , , K ¯ n }

and fix integer k > K ¯ . Let R ˆ max { R ˆ 1 , R ˆ 2 , , R ˆ n } . By applying the previous argument to all 1 j n , we obtain

max 1 j n sup x Q j , dist ( x , Q j ) R ˆ v ( x , t + k T ; v 0 ) U j ( x , t ) V ̲ β ( x , t ) < ε .

Step 2. For any ε > 0 , fix k > K ¯ , t [ 0 , T ] , we will show that

lim R + max 1 j n sup x R , dist ( x , Q j ) R ˆ v ( x , t + k T ; v 0 ) U ^ ( x , t ) V ̲ β ( x , t ) < ε .

By the definition of Γ and Q j , we have

lim R + inf x R , dist ( x , Q j ) R ˆ dist ( x , Γ ) = + , 1 j n .

Applying Lemma 3.3 to v ( x , t ; v 0 ) V ( x , t ) , we obtain

sup D ( 2 γ ¯ ) v ( x , t ; v 0 ) V ( x , t ) V ̲ β ( x , t ) 8 M β e N ¯ t sup D ( γ ¯ ) v 0 ( x ) V ( x , t ) V ̲ β ( x , t ) + 3 M β e N ¯ t erfc 3 3 γ ¯ K β N 1 c * c t c t 4 t sup D ( γ ¯ ) c v 0 ( x ) V ( x , t ) V ̲ β ( x , t )

for any γ ¯ > 0 and t > 0 . Thus, by the assumption of v 0 ( x ) , Lemma 3.4 and Theorem 1.1, we have

lim R + max 1 j n sup x R , dist ( x , Q j ) R ˆ v ( x , t + k T ; v 0 ) U ^ ( x , t ) V ̲ β ( x , t ) < ε .

Combining the estimates of Steps 1 and 2, we obtain

lim R + sup x R v ( x , t + k T ; v 0 ) U ^ ( x , t ) V ̲ β ( x , t ) < ε for any given integer k K ¯ .

This completes the proof.□

The following lemma follows directly from Theorem 3.5.

Lemma 3.6

Let V ( x , t ) be as in Theorem 1.1. It satisfies

(3.16) lim R + sup x R V ( x , t ) U ^ ( x , t ) V ̲ β ( x , t ) = 0 , β 0 , 1 8 .

Furthermore, suppose the initial value v 0 C ( R 3 , [ 0 , 1 ] ) satisfying

lim γ ¯ + sup x D ( γ ¯ ) v 0 ( x ) V ̲ ( x , 0 ) V ̲ β ( x , 0 ) = 0

for some β 0 , 1 8 . Then, for any given ε > 0 , there exists a constant K * N large enough such that

(3.17) lim R + sup x R v ( x , t + k T ; v 0 ) V ( x , t ) V ̲ β ( x , t ) < ε , k K * .

Proof

Since U ^ ( x , t ) < V ( x , t ) and by taking v 0 ( x ) = V ( x , 0 ) in Theorem 3.5, we can readily obtain (3.16). From Theorem 3.5 and equation (3.16), it follows that (3.17) holds. This completes the proof.□

In the following, we present some properties of V and V z when x sufficiently far away from the level set passing through the origin.

Lemma 3.7

lim R + sup z + h ( x , y ) R , t [ 0 , T ] V ( x , t ) V ̲ β ( x , t ) = 0 , β ( 0 , 1 ) .

Proof

It is obvious that

sup z + h ( x , y ) < R , t [ 0 , T ] V ( x , t ) V ̲ β ( x , t ) sup z + h ( x , y ) < R , t [ 0 , T ] V ( x , t ) V ̲ ( x , t ) V ̲ β ( x , t ) + sup z + h ( x , y ) < R , t [ 0 , T ] V ̲ 1 β ( x , t ) .

Since

lim γ ¯ + sup x D ( γ ¯ ) , t [ 0 , T ] V ( x , t ) V ̲ ( x , t ) V ̲ β ( x , t ) = 0 ,

we have

lim R + sup z + h ( x , y ) < R , t [ 0 , T ] V ( x , t ) V ̲ ( x , t ) V ̲ β ( x , t ) = 0 .

Additionally, by

lim R + sup z + h ( x , y ) < R , t [ 0 , T ] V ̲ 1 β ( x , t ) = 0 ,

we have proved the lemma.□

Lemma 3.8

lim R + sup z + h ( x , y ) R , t [ 0 , T ] V z ( x , t ) V ̲ β ( x , t ) = 0 , β ( 0 , 1 ) .

Proof

The proof is similar to that of [42, Lemma 4.6] and is omitted.□

Lemma 3.9

(3.18) lim R + sup z + h ( x , y ) R , t [ 0 , T ] V z ( x , t ) V ̲ β ( x , t ) = 0 , β ( 0 , 1 ) .

Furthermore, for any δ ( 0 , ε 1 ) , there exists a constant β 2 such that

(3.19) inf δ V ( x , t ) 1 δ , t [ 0 , T ] V z ( x , t ) > β 2 .

Proof

First, we prove (3.19). Since V z > 0 in R 3 × [ 0 , + ] and V is T -periodic in time, V z ( x , t ) has a positive minimum value when x is restricted to any compact subset of R 3 . Thus, it suffices to study V z as x + . Assume x i = ( x i , y i , z i ) satisfies lim i + x i = + and δ V ( x i , t ) 1 δ . It suffices to prove liminf i + ( V z ( x i , t ) ) > 0 . By Theorem 1.1 and the definition of V ̲ , we know that liminf i + ( V z ( x i , t ) ) > 0 always holds if lim sup i + dist ( x i , Γ ) + . Therefore, we only need to consider the case that lim sup i + dist ( x i , Γ ) < + . Without loss of generality, we can assume

lim i + x i = + , limsup i + dist ( x i , Γ j ) < +

for some j ( 1 j n ) . Then, we obtain

lim i + dist ( x i , Γ m ) = + , m j .

Now, (3.16) gives

lim i + V = U j in L ( B ( x i ; 2 ) ) ,

where B ( x 0 ; r ) is the ball of radius r centered at x 0 . We have

lim i + ( f ( V , t ) f ( U j , t ) ) = 0 in L p ( B ( x i ; 2 ) )

for p > 3 . Define

L ( V ) V t V x x V y y V z z + c V z .

Apply the interior L P estimate (see [19, Theorem 7.22]) to

L ( V U j ) = f ( V , t ) f ( U j , t ) in B ( x i ; 2 ) .

Then, we obtain

lim i + D ( V ( x , t ) U j ( x , t ) ) L p ( B ( x i ; 1 ) ) = 0

and thus

lim i + V z ( x i , t ) z U j ( x i , t ) = 0 .

We have

z U j ( x ) = q j q j 2 + m * 2 p j 2 η V * ( ξ , η ; s j ) .

Thus, it follows from [41, Lemma 4.2] and [42, Theorem 4.1], we obtain (3.19) hold.

Now, we proceed to prove (3.18). lim R + sup z + h ( x , y ) < R , t [ 0 , T ] V z ( x , t ) V ̲ β ( x , t ) = 0 has been proved in Lemma 3.8, we just need to prove that

lim R + sup z + h ( x , y ) R , t [ 0 , T ] V z ( x , t ) V ̲ β ( x , t ) = 0 .

Assume z + h ( x , y ) + , which implies dist ( x , Γ ) + . We obtain

lim R + sup z + h ( x , y ) R , t [ 0 , T ] V ( x , t ) 1 = 0 .

Thus, we obtain

lim R + sup { f ( V ) L ( B ( x ; 2 ) ) x = ( x , y , z ) , z + h ( x , y ) R } = 0 .

Apply the interior L P estimate (see [19, Theorem 7.22]) to L V = f ( V , t ) , we obtain

lim R + sup { D V L p ( B ( x ; 1 ) ) x = ( x , y , z ) , z + h ( x , y ) R } = 0

for any p > 3 . Thus, we obtain

lim R + sup z + h ( x , y ) R V z ( x ) = 0 .

Since

lim R + sup z + h ( x , y ) R V ̲ ( x , t ) = 1 ,

then we have

lim R + sup z + h ( x , y ) R V z ( x , t ) V ̲ β ( x , t ) = 0 , β ( 0 , 1 ) .

Thus, (3.18) holds. This completes the proof.□

The following lemma shows that the solution to (1.5) and (1.6) with initial value V ̲ approaches to V in a certain sense for a sufficiently large time t .

Lemma 3.10

Let V ( x , t ) be as in Theorem 1.1. Then

lim k + v ( x , t + k T ; V ̲ ) V ( x , t ) V ̲ β ( x , t ) L ( R 3 ) = 0 , β 0 , 1 8 .

Proof

Note that z + h ( x , y ) + implies dist ( x , Γ ) + . Then, by Theorem 1.1, we have

lim R + sup z + h ( x , y ) R V ( x , t ) V ̲ ( x , t ) V ̲ β ( x , t ) = 0 , β ( 0 , 1 ) .

Thus, for any ε > 0 , there exists a constant R ˜ > 0 large enough such that for any x R 3 with z + h ( x , y ) R ˜ , we have

sup z + h ( x , y ) R ˜ V ( x , t ) V ̲ ( x , t ) V ̲ β ( x , t ) < ε .

Since V ̲ is a subsolution of (1.5), we have

v ( x , t ; V ̲ ) V ̲ ( x , t ) V ( x , t ) ε V ̲ β ( x , t )

for x R 3 and t > 0 with z + h ( x , y ) R ˜ . Following a proof similar to that of Theorem 3.2 in [4], we can conclude

(3.20) lim k + v ( x , t + k T ; V ̲ ) V ( x , t ) C l o c 2,1 ( R 3 × [ 0 , T ] ) = 0 .

Following from Lemma 3.6, we have for any ε > 0 there exists k 1 N large enough and r 1 > 0 such that

(3.21) sup x r 1 v ( x , t ; V ̲ ) V ( x , t ) V ̲ β ( x , t ) < ε , k > k 1 .

Combining (3.20) and (3.21), we have

lim k + v ( x , t + k T ; V ̲ ) V ( x , t ) L ( R 3 × [ 0 , + ] ) = 0 .

Then, by the definition of V ̲ ( x , t ) , there exists a constant k ˆ R large enough such that for any integer k k ˆ and x R 3 with z + h ( x , y ) R ˜ ,

v ( x , t + k T ; V ̲ ) V ( x , t ) ε ϕ β c * c R ˜ , t V ( x , t ) ε V ̲ β ( x , t ) .

Combining with v ( x , t + k T ; V ̲ ) V ( x , t + k T ) = V ( x , t ) , we have

V ( x , t ) v ( x , t + k T ; V ̲ ) V ( x , t ) ε V ̲ β ( x , t ) , x R 3 , k k ˆ .

Since ε is arbitrary, this completes the proof.□

The following lemma establishes a positive lower bound for the partial derivatives of U j ( x , t ) and U ^ ( x , t ) with respect to z under certain conditions.

Lemma 3.11

Assume that δ ( 0 , ε 1 ) . Then

inf δ U j ( x , t ) 1 δ , t [ 0 , T ] z U j ( x , t ) > β 2 , 1 j n .

For any x R 3 with δ U j ( x , t ) 1 δ , it holds that

inf 0 < Θ < Θ 0 , t [ 0 , T ] U ^ ( x , y , z + Θ , t ) U ^ ( x , t ) Θ min 1 j n inf δ 2 U j ( x , t ) 1 δ 2 , t [ 0 , T ] z U j ( x , t ) > 0 ,

where Θ 0 is a positive constant depending on δ and is independent of x .

Proof

The proof is similar to that of [5, Lemma 3.6] and is omitted.□

Lemma 3.12

Let v ˜ ( x , t ; v ˜ 0 ) and v ˆ ( x , t ; v ˆ 0 ) be the solution of (1.5) and (1.6) with initial values v ˜ 0 ( x ) and v ˆ 0 ( x ) , respectively. Assume that v ˜ 0 ( x ) , v ˆ 0 ( x ) C ( R 3 , [ 0 , 1 + ε 1 ] ) satisfy

sup x R 3 v ˜ 0 ( x ) v ˆ 0 ( x ) V ̲ β ( x , 0 ) < +

for some β ( 0 , 1 ) . Then, there exists a constant λ * > 0 such that

sup x R 3 v ˜ ( x , t ; v ˜ 0 ) v ˆ ( x , t ; v ˆ 0 ) V ̲ β ( x , t ) 8 M β e λ * t sup x R 3 v ˜ 0 ( x ) v ˆ 0 ( x ) V ̲ β ( x , 0 ) .

Proof

Recall that m max 1 v 2 f ( v , t ) . Let w ¯ ( x , t ) be given by

w ¯ t N w ¯ m w ¯ = 0 , x R 3 , t > 0 , w ¯ ( x , 0 ) = v ˜ 0 ( x ) v ˆ 0 ( x ) , x R 3 .

Then, we have w ¯ ( x , t ) 0 for x R 3 and t > 0 . Define u ¯ w ¯ ( v ˜ v ˆ ) . We have

u ¯ t N u ¯ 0 1 f ( ι v ˜ + ( 1 ι ) v ˆ , t ) d ι u ¯ = m 0 1 f ( ι v ˜ + ( 1 ι ) v ˆ , t ) d ι w ¯ 0

for x R 3 and t > 0 . Noting that u ¯ ( x , 0 ) 0 for all x R 3 , then by the comparison principle, we obtain

v ˜ ( x , t ) v ˆ ( x , t ) w ¯ ( x , t ) , x R 3 , t > 0 .

Similarly, we have

v ˆ ( x , t ) v ˜ ( x , t ) w ¯ ( x , t ) , x R 3 , t > 0 .

We now proceed to estimate w ¯ ( x , t ) . Obviously,

w ¯ ( x , t ) = e m t R 3 Γ ¯ ( x , t ) w ¯ ( x x , 0 ) d x , x R 3 , t > 0 ,

where Γ ¯ ( x , t ) is defined as in Lemma 3.3. Recall that

N 1 = sup ζ R , t [ 0 , T ] ϕ ζ ( ζ , t ) ϕ ( ζ , t ) and M = sup t [ 0 , T ] V ¯ β ( x , 0 ) V ̲ β ( x , t ) .

Let

M ^ sup x R 3 v ˜ 0 ( x ) v ˆ 0 ( x ) V ̲ β ( x , 0 ) = sup x R 3 w ¯ ( x , 0 ) V ̲ β ( x , 0 ) .

Then, we have

w ¯ ( x , t ) V ̲ β ( x , t ) = e m t R 3 Γ ¯ ( x , t ) w ¯ ( x x , 0 ) V ̲ β ( x , t ) d x = e m t R 3 Γ ¯ ( x , t ) V ̲ β ( x , 0 ) V ̲ β ( x , t ) V ̲ β ( x x , 0 ) V ̲ β ( x , 0 ) w ¯ ( x x , 0 ) V ̲ β ( x x , 0 ) d x M β M ^ e m t R 3 Γ ¯ ( x , t ) V ̲ β ( x x ) V ̲ β ( x ) d x = M β M ^ e m t R 3 Γ ¯ ( x , t ) ϕ β c * c ( z z + h ( x x , y y ) ) , t ϕ β c * c ( z + h ( x , y ) ) , t d x M β M ^ e m t R 3 Γ ¯ ( x , t ) ϕ β c * c ( z + z + h ( x , y ) + A ˆ x + B ˆ y ) , t ϕ β c * c ( z + h ( x , y ) ) , t d x M β M ^ e m t R 3 Γ ¯ ( x , t ) exp c * β N 1 c ( z + A ˆ x + B ˆ y ) d x M β M ^ e ( m + β N 1 c * ) t R 3 1 ( 4 π t ) 3 2 exp x 2 + y 2 + z 2 4 t × exp c * β N 1 c ( z + A ˆ x + B ˆ y ) d x 8 M β M ^ e ( m + β N 1 c * + 3 β 2 N 1 2 ) t .

Here, recall that A ˆ m * max 1 j n cos θ j and B ˆ m * max 1 j n sin θ j . Let λ * m + β N 1 c * + 3 β 2 N 1 2 . This completes the proof.□

4 Stability of pyramidal traveling fronts

In this section, we study the asymptotic stability of pyramidal traveling fronts when v 0 V ̲ . We first construct some new supersolutions and give some properties of them. At the end of article, we give the proof of Theorem 1.2. Let φ be defined as in (2.5). For α > 0 and ε > 0 , we set

V ¯ ( x , t ; β , ε , α ) = ϕ ( μ ˆ ( x ) , t ) + ε σ ( x , y ) ( ω ( ρ ˆ ( x ) ) ν ( t ) + ( 1 ω ( ρ ˆ ( x ) ) ) ϕ β ( ρ ˆ ( x ) , t ) ) ,

where

μ ˆ ( x ) = z + 1 α φ ( α x , α y ) 1 + φ ( α x , α y ) 2 , σ ( x , y ) = c 1 + φ ( α x , α y ) 2 c * , ρ ˆ ( x ) = c * c z + 1 α φ ( α x , α y ) .

Since both ϕ and ν are T -periodic in time, it follows that V ¯ is also T -periodic. Therefore, we have the following lemma, which is obtained from [4].

Lemma 4.1

[4, Lemma 3.1] For each β ( 0 , 1 ) , there exist positive constants ε 0 + ( β ) and α 0 + ( β , ε ) such that, for any 0 < ε < ε 0 + ( β ) and 0 < α < α 0 + ( β , ε ) , the function

V ¯ ( x , t ; β , ε , α ) = ϕ ( μ ˆ ( x ) , t ) + ε σ ( x , y ) ( ω ( ρ ˆ ( x ) ) ν ( t ) + ( 1 ω ( ρ ˆ ( x ) ) ) ϕ β ( ρ ˆ ( x ) , t ) )

is a supersolution of (1.7) and (1.8) on R 3 × R . Moreover, the following inequalities hold:

(4.1) lim γ ¯ + sup x D ( γ ¯ ) , t [ 0 , T ] V ¯ ( x , t ; β , ε , α ) V ̲ ( x , t ) V ̲ β ( x , t ) 2 ε , V ̲ ( x , t ) < V ¯ ( x , t ; β , ε , α ) , x R 3 , t > 0 , V ¯ z ( x , t ; β , ε , α ) > 0 , x R 3 , t > 0 .

Lemma 4.2

For any M ˜ > 0 , there exists a positive constant β 3 > 0 such that

V ¯ z ( x , t ; β , ε , α ) β 3 if M ˜ μ ˆ ( x ) M ˜ .

Proof

By the definition of V ¯ , we have

V ¯ z ( x , t ; β , ε , α ) = ϕ ξ ( μ ˆ ( x ) , t ) 1 + φ ( α x , α y ) 2 + c * c ε σ ( x , y ) [ ω ( ρ ˆ ( x ) ) ( ν ( t ) ϕ β ( ρ ˆ ( x ) , t ) ) + c * c ( 1 ω ( ρ ˆ ( x ) ) ) β ϕ β 1 ϕ ξ ( ρ ˆ ( x ) , t ) .

For any M ˜ > 0 , it follows from the properties of ϕ , ω > 0 , ν 1 for all t > 0 and Lemma 2.3 that there exists β 3 > 0 such that

V ¯ z ( x , t ; β , ε , α ) ϕ ξ ( μ ˆ ( x ) , t ) 1 + φ ( α x , α y ) 2 c * c ϕ ξ ( μ ˆ ( x ) , t ) β 3

as M ˜ μ ˆ ( x ) M ˜ . This completes the proof.□

In the following, we construct some supersolutions for (1.5).

Lemma 4.3

Let V ( x , y , z , t ) be defined by Theorem 1.1. For β ( 0 , 1 ) and

0 < ϑ < min Π ( β c * ) 48 c * K 1 , 1 , 0 < δ < ε 1 P 2 ,

there exists a positive constant ρ large enough, a positive constant κ small enough such that for any ξ R , the function

H + ( x , t ; β , ρ , δ , ξ , ϑ ) V ( x , y , z + ξ + ρ δ ( 1 e κ t ) , t ) + δ e κ t ( ω ( θ ) ν ( t ) + ( 1 ω ( θ ) ) ϕ β ( θ , t ) )

is a supersolution of (1.5), where

θ = c * c z + ξ + ρ δ ( 1 e κ t ) + φ ( ϑ x , ϑ y ) ϑ .

Proof

From (1.2) and some computations, we have

(4.2) [ H + ] = H t + H x x + H y y + H z z + + c H z + f ( H + , t ) = δ κ e κ t ω ν + ( 1 ω ) ϕ β c * c ρ δ e κ t ( ν ϕ β ) ω c * c β ρ δ e κ t ( 1 ω ) ϕ β 1 ϕ θ + ρ δ κ e κ t V z δ e κ t { ω ( θ x 2 + θ y 2 + θ z 2 ) ( ν ϕ β ) + ω ( θ x x + θ y y ) ( ν ϕ β ) 2 β ω ϕ β 1 ϕ θ ( θ x 2 + θ y 2 + θ z 2 ) + ( 1 ω ) [ β ( β 1 ) ϕ β 2 ϕ θ 2 ( θ x 2 + θ y 2 + θ z 2 ) + β ϕ β 1 ϕ θ θ ( θ x 2 + θ y 2 + θ z 2 ) + β ϕ β 1 ϕ θ ( θ x x + θ y y ) c * β ϕ β 1 ϕ θ c * ω ( ν ϕ β ) ] } + δ e κ t [ ω ν t + ( 1 ω ) β ϕ β 1 ϕ t ] + f ( V , t ) f ( H + , t ) .

To prove the lemma, we consider three cases.

Case 1. θ ( x , t ) < X for some X 1 large enough. Then, we have

[ H + ] = ρ δ κ e κ t V z δ κ e κ t ϕ β + c * c β ρ δ 2 κ e 2 κ t ϕ β 1 ϕ θ δ e κ t ϕ β β ( β 1 ) ϕ θ 2 ϕ 2 ( θ x 2 + θ y 2 + θ z 2 ) + β ϕ θ θ ϕ ( θ x 2 + θ y 2 + θ z 2 ) + β ϕ θ ϕ ( θ x x + θ y y ) c * β ϕ θ ϕ β ϕ t ϕ + f ( V , t ) f ( H + , t ) δ e κ t ϕ β κ β ( β 1 ) ϕ θ ϕ 2 ( θ x 2 + θ y 2 + θ z 2 ) β ϕ θ ϕ ( θ x x + θ y y ) + c * β ϕ θ ϕ + β ϕ t ϕ β ϕ θ θ ϕ ( θ x 2 + θ y 2 + θ z 2 ) + f ( V , t ) f ( H + , t ) δ e κ t ϕ β κ β ϕ θ ϕ 2 ϕ θ θ ϕ ( θ x 2 + θ y 2 + θ z 2 ) β 2 ϕ θ ϕ 2 ( θ x 2 + θ y 2 + θ z 2 ) + β 2 ϕ θ ϕ 2 β ϕ θ ϕ ( θ x x + θ y y ) β 2 ϕ θ ϕ 2 + c * β ϕ θ ϕ + β ϕ t ϕ + f ( V , t ) f ( H + , t ) .

Since lim ξ ϕ ξ ( ξ , t ) ϕ ( ξ , t ) = c * and lim ξ ϕ ξ ξ ( ξ , t ) ϕ ξ ( ξ , t ) = c * 2 uniformly in t R , there exists X 1 > 0 large enough such that

ϕ θ ϕ 2 ϕ θ θ ϕ < 1 16 Π ( β c * ) , β 2 ϕ θ ϕ 2 c * β ϕ θ ϕ < 1 2 Π ( β c * ) , β ϕ θ θ ϕ β c * ϕ θ ϕ < 1 16 Π ( β c * ) , β f ( ϕ , t ) ϕ < 1 16 Π ( β c * ) , ϕ θ ϕ < 3 2 c *

for ξ < X 1 and t [ 0 , T ] . Besides,

ϕ θ ϕ 2 ( θ x 2 + θ y 2 + θ z 2 ) ϕ θ ϕ 2 = ϕ θ ϕ 2 c * c 2 ( 1 + φ x 2 + φ y 2 ) 1 > 0 , x R 3 .

By Lemma 2.3, we have

h ( x , y ) < φ ( ϑ x , ϑ y ) h ( x , y ) + 2 π m * 0 + r 2 ρ ˜ ( r ) d r ϑ .

Then, for the reaction team f , there exists X 2 > 0 large enough such that

f ( V , t ) f ( H + , t ) = δ e κ t ϕ β f u ( V + τ δ e κ t ϕ β , t ) > δ e κ t ϕ β 1 16 Π ( β c * ) , θ < X 2 , t > 0 ,

where τ ( 0 , 1 ) .

Set X = max { 1 , X 1 , X 2 } , then for θ ( x , t ) > X , from the above estimates, we have

[ H + ] δ e κ t ϕ β κ β ϕ θ ϕ 2 ϕ θ θ ϕ ( θ x 2 + θ y 2 + θ z 2 ) β 2 ϕ θ ϕ 2 ( θ x 2 + θ y 2 + θ z 2 ) + β 2 ϕ θ ϕ 2 β ϕ θ ϕ ( θ x x + θ y y ) β 2 ϕ θ ϕ 2 + c * β ϕ θ ϕ + β ϕ t ϕ + f ( V , t ) f ( H + , t ) δ e κ t ϕ β κ + 1 16 Π ( β c * ) + 1 16 Π ( β c * ) 1 2 Π ( β c * ) + 1 16 Π ( β c * ) + 1 16 Π ( β c * ) + 1 16 Π ( β c * ) > 0 ,

provided that κ < 3 16 Π ( β c * ) .

Case 2. θ ( x , t ) > X for some X 1 large enough

(4.3) [ H + ] = δ κ e κ t ν + ρ δ κ e κ t V z + δ e κ t ν t + f ( V , t ) f ( H + , t ) = ρ δ κ e κ t V z δ κ e κ t ν + δ e κ t ( f u ( 1 , t ) ν + Λ 0 ν ) δ e κ t f u ( V + τ δ e κ t ν , t ) δ e κ t ν ( κ + Λ 0 ) + δ e κ t ν ( f u ( 1 , t ) f u ( V + τ δ e κ t ν , t ) ) .

Since δ < ε 1 P 2 , there exists X 1 > 0 large enough such that

(4.4) 1 ε 1 < V + τ δ e κ t ν < 1 + ε 1 , θ > X 1 , t > 0 .

Then, by (2.7), we have

f u ( 1 , t ) f u ( V + τ δ e κ t ν , t ) > 1 2 Λ 0 , θ > X 1 , t > 0 .

Let X = max { 1 , X 1 } , then for θ ( x , t ) > X , we have

[ H + ] δ e κ t κ + 1 2 Λ 0 P 1 0 ,

provided that 0 < κ < 1 2 Λ 0 .

Case 3. X θ ( x , t ) X . Recall that

N 1 = sup ξ R , t R ϕ ξ ( ξ , t ) ϕ ( ξ , t ) .

Furthermore, we denote

N 2 sup ξ R , t R ϕ ξ ξ ( ξ , t ) ϕ , M 0 max u [ 1 , 2 ] , t R f u ( u , t ) .

From (1.2) and the definition of ν t ( t ) , one has

( 1 ω ( η ) ) β ϕ β 1 ( η , t ) ϕ t ( η , t ) + ω ( η ) ν t ( t ) = ( 1 ω ( η ) ) β ϕ β ( η , t ) c * ϕ η ( η , t ) ϕ ( η , t ) + ϕ η η ( η , t ) ϕ ( η , t ) + f ( ϕ ( η , t ) , t ) ϕ ( η , t ) + ω ( η ) K 0 ( f u ( 1 , t ) + Λ 0 ) Φ ¯ ( t ) c * N 1 + N 2 + M 0 + K 0 ( f u ( 1 , t ) + Λ 0 ) Φ ¯ ( t ) .

Then, by the boundedness of Φ ¯ ( t ) , there exists a positive constant C 0 such that

max η R , t R ω ( η ) β ϕ β 1 ( η , t ) ϕ t ( η , t ) + ( 1 ω ( η ) ) ϕ ( t ) C 0 .

Then, for X θ ( x , t ) X , we have

[ H + ] δ κ e κ t [ ω ν + ( 1 ω ) ϕ β ] + ρ δ κ e κ t V z δ e κ t { [ ω ( θ x 2 + θ y 2 + θ z 2 ) + ω ( θ x x + θ y y ) ] ( ν ϕ β ) + ( 1 ω ) [ β ϕ β 1 ϕ θ θ ( θ x 2 + θ y 2 + θ z 2 ) + β ϕ β 1 ϕ θ ( θ x x + θ y y ) ] } δ e κ t f u ( V + τ δ e κ t ( ω ν + ( 1 ω ) ϕ β ) , t ) ( ω ν + ( 1 ω ) ϕ β ) + δ e κ t [ ω ν t + ( 1 ω ) β ϕ β 1 ϕ t ] δ e κ t ( ρ κ β 2 κ P 2 sup θ R ω P 2 m * 2 P 2 ) N 2 m * 2 N 1 M 0 P 2 C 0 0 ,

provided that

(4.5) ρ > ( κ + M 0 + sup θ R ω ) P 2 + N 2 + m * 2 ( P 2 + N 1 ) + C 0 κ β 2 .

This completes the proof.□

Lemma 4.4

Let V ¯ ( x , t ; β , ρ , δ , ε , ξ , ϑ ) define by Theorem 4.1. For β ( 0 , 1 ) and

0 < ε < min ε 0 + ( β ) , ε 1 2 P 2 ( c c * ) , 0 < ϑ < min α 0 + ( ε , β ) , Π ( β c * ) 48 c * K 1 , 0 < δ < ε 1 2 P 2 ,

there exists a positive constant ρ large enough, a positive constant κ small enough such that for any ξ R , the function

V + ( x , t ; β , ρ , δ , ε , ξ , ϑ ) = V ¯ ( x , y , z + ξ + ρ δ ( 1 e κ t ) , t ; β , ε , ϑ ) + δ e κ t ( ω ( θ ) ν ( t ) + ( 1 ω ( θ ) ) ϕ β ( θ , t ) )

is a supersolution of (1.5), where

θ = c * c z + ξ + ρ δ ( 1 e κ t ) + φ ( ϑ x , ϑ y ) ϑ .

Proof

The proof is very similar to the proof of Lemma 4.3 and we omit it.□

Fix β ( 0 , 1 8 ) . Take 0 < ε < min ε 0 + ( β ) , ε 1 2 P 2 and 0 < α < min α 0 + ( ε , β ) , Π ( β c * ) 48 c * K 1 . Since V ¯ ( x , t ; β , ε , α ) is a supersolution of (1.5) and satisfies V ¯ ( x , t ; β , ε , α ) = V ¯ ( x , t + T ; β , ε , α ) for all ( x , t ) R 3 × R . Hence, the function

(4.6) V * ( x , t ) lim k + v ( x , t + k T ; V ¯ 0 ) , ( x , t ) R 3 × R

is well-defined, where V ¯ 0 V ¯ ( x , 0 ; β , ε , α ) . Moreover, we have that v ( x , t + k T ; V ¯ 0 ) is non-increasing in k N and converges to V * under the norm C l o c 2,1 ( R 3 × R ) as k + . It is obvious that V * ( x , t ) C 2,1 ( R 3 × R ) satisfies (1.7) with V * ( x , t + T ) = V * ( x , t ) and

V ( x , t ) V * ( x , t ) V ¯ ( x , t ; β , ε , α ) , x R 3 , t > 0 .

Lemma 4.5

Let V and V * be as in (3.8) and (4.6). Then

V ( x , t ) V * ( x , t ) , x R 3 , t > 0 .

Proof

By the periodicity of V and V * , it suffices to show that

V ( x , t ) V * ( x , t ) , x R 3 , t [ 0 , T ] .

Assume that V ( x , t ) V * ( x , t ) , it then follows from the comparison principle that

V ( x , t ) < V * ( x , t ) , x R 3 , t [ 0 , T ] .

For any δ ε 1 2 P 2 , ε 1 P 2 , taking λ > 0 large enough such that

V * ( x , t ) V ( x , y , z + λ , t ) + δ ( ω ( η ˆ ) ν ( t ) + ( 1 ω ( η ˆ ) ) ϕ β ( η ˆ , t ) ) , x R 3 , t [ 0 , T ] ,

where η ˆ = c * c ( z + λ + φ ( α x ) α ) . By Lemma 4.3 and the comparison principle, we obtain

V * ( x , t ) H + ( x , t ; β , ρ , δ , λ , α ) , x R 3 , t 0 .

For k N , let k + in

V * ( x , k T + t ) V ( x , y , z + λ + ρ δ ( 1 e κ ( k T + t ) ) , k T + t ) + δ e κ ( k T + t ) ( ω ( θ ) ν ( t ) + ( 1 ω ( θ ) ) ϕ β ( θ , k T + t ) )

with t [ 0 , T ] , we have

V * ( x , t ) V ( x , y , z + λ + ρ δ , t ) , x R 3 , t [ 0 , T ] ,

where ρ > 0 is defined in Lemma 4.3. For any ε > 0 , applying (4.1) to V ¯ ( x , t ; β , ε 4 , α ) yields

lim γ ¯ + sup x D ( γ ¯ ) , t [ 0 , T ] V ¯ ( x , t ; β , ε 4 , α ) V ̲ ( x , t ) V ̲ ( x , t ) β ε 2 .

That is, there exists a constant Γ 1 > 0 large enough such that

sup x D ( γ ¯ ) , t [ 0 , T ] V ¯ ( x , t ; β , ε 4 , α ) V ̲ ( x , t ) V ̲ β ( x , t ) ε 2 , γ ¯ Γ 1 .

By (1.9), there exists a constant Γ 2 > 0 large enough such that

sup x D ( γ ¯ ) , t [ 0 , T ] V ( x , t ) V ̲ ( x , t ) V ̲ β ( x , t ) ε 2 , γ ¯ Γ 2 .

Set Γ max { Γ 1 , Γ 2 } , then we have

sup x D ( γ ¯ ) , t [ 0 , T ] V ¯ ( x , t ; β , ε 4 , α ) V ( x , t ) V ̲ β ( x , t ) ε , γ ¯ Γ .

Note that V ( x , t ) V * ( x , t ) V ¯ ( x , t ; β , ε 4 , α ) , we obtain

sup x D ( γ ¯ ) , t [ 0 , T ] V * ( x , t ) V ( x , t ) V ̲ β ( x , t ) ε , γ ¯ Γ ,

which implies

lim γ ¯ + sup x D ( γ ¯ ) , t [ 0 , T ] V * ( x , t ) V ( x , t ) V ¯ β ( x , t ) = 0 .

Combining the above equation with equation (1.9), we obtain

lim γ ¯ + sup x D ( γ ¯ ) , t [ 0 , T ] V * ( x , t ) V ̲ ( x , t ) V ̲ β ( x , t ) = 0 .

Applying Theorem 3.5 to V * ( x , t ) , and noting that V * , V ̲ , and U ^ share the same time periodicity, we obtain

(4.7) lim R + sup x R V * ( x , t ) U ^ ( x , t ) V ̲ β ( x , t ) < ε .

Define

Λ inf { λ > 0 V * ( x , y , z , t ) V ( x , y , z + λ , t ) for all x R 3 and t [ 0 , T ] } ,

then we have Λ 0 and V * ( x , y , z , t ) V ( x , y , z + Λ , t ) for all x R 3 and t [ 0 , T ] . We show Λ = 0 by contradiction. Assume that Λ > 0 . Then, by the maximum principle, we have either

V * ( x , y , z , t ) V ( x , y , z + Λ , t ) , x R 3 , t [ 0 , T ]

or

V * ( x , y , z , t ) < V ( x , y , z + Λ , t ) , x R 3 , t [ 0 , T ] .

For the former case, since

lim dist ( x , Γ ) + V * ( x , y , z , t ) = ϕ ( 0 , t ) , lim dist ( x , Γ ) + V ( x , y , z + Λ , t ) = ϕ c * c Λ , t

for z = h ( x , y ) , we obtain a contradiction. Thus

V * ( x , y , z , t ) < V ( x , y , z + Λ , t ) , x R 3 , t [ 0 , T ] .

Thanks to

lim R + sup z + h ( x , y ) < R , t [ 0 , T ] V z ( x , y , z , t ) ( V ̲ ( x , y , z , t ) ) β = 0

and

lim R + sup z + h ( x , y ) > R , t [ 0 , T ] V z ( x , y , z , t ) = 0 ,

there exists R * > 0 large enough such that

(4.8) sup z + h ( x , y ) R * + Λ , t [ 0 , T ] V z ( x , y , z , t ) ( V ̲ ( x , y , z , t ) ) β < 1 3 ρ ,

(4.9) sup z + h ( x , y ) R * 2 ρ , t [ 0 , T ] V z ( x , y , z , t ) < 1 4 ρ ,

(4.10) inf z + h ( x , y ) R * 2 ρ , t [ 0 , T ] ( V ̲ ( x , y , z , t ) ) β > 1 2 .

Define

Ω { x R 3 z + h ( x , y ) R * } .

Set σ small enough such that 0 < σ < min ε 1 2 P 2 , Λ 4 ρ and e 2 N 1 ρ σ β < 3 2 . Let

δ * min ε 1 2 P 2 , 1 max 1 j n sup x Ω , t [ 0 , T ] U j x , y , z + Λ 2 , t , min 1 j n inf x Ω , t [ 0 , T ] U j ( x , t ) .

For x Ω , applying Lemma 3.11 we obtain

U ^ x , y , z + Λ 2 , t U ^ x , y , z + Λ 4 , t min Θ 0 , Λ 4 min 1 j n inf δ * 2 U j ( x , t ) 1 δ * 2 , t [ 0 , T ] z U j ( x , t ) > 0 .

Thus, we have

inf x Ω , t [ 0 , T ] ( U ^ ( x , y , z + Λ 2 ρ σ , t ) U ^ ( x , t ) ) > min Θ 0 , Λ 4 min 1 j n inf δ * 2 U j ( x , t ) 1 δ * 2 , t [ 0 , T ] z U j ( x , t ) > 0 .

Using Lemma 3.6, (3.19), and (4.7), if x Ω with x R 1 * for some large positive constant R 1 * , we have

V * ( x , t ) < V x , y , z + Λ 2 , t + σ ϕ β c * c z + Λ 2 + φ ( α x , α y ) α , t V ( x , y , z + Λ 2 ρ σ , t ) + σ ϕ β c * c z + Λ 2 ρ σ + φ ( α x , α y ) α , t .

Since Ω B ( 0 , R 1 * ) ¯ is compact, we have

V * ( x , t ) < V ( x , y , z + Λ 2 ρ σ , t )

in Ω B ( 0 , R 1 * ) ¯ for sufficiently small σ . Thus

V * ( x , t ) < V ( x , y , z + Λ 2 ρ σ , t ) + σ ϕ β c * c z + Λ 2 ρ σ + φ ( α x , α y ) α , t

for x Ω and t [ 0 , T ] . For x R 3 with z + h ( x , y ) R * and t [ 0 , T ] , it follows from (4.8) that

V ( x , y , z + Λ 2 ρ σ , t ) V ( x , y , z + Λ , t ) ϕ β c * c ( z + Λ 2 ρ σ + φ ( α x ) α ) , t V ( x , y , z + Λ 2 ρ σ , t ) V ( x , y , z + Λ , t ) ( V ̲ ( x , y , z + Λ 2 ρ σ , t ) ) β = 2 ρ σ 0 1 V z ( x , y , z + Λ 2 ρ σ τ , t ) d τ ( V ̲ ( x , y , z + Λ 2 ρ σ , t ) ) β 2 ρ σ ( V ̲ ( x , y , z + Λ , t ) ) β ( V ̲ ( x , y , z + Λ 2 ρ σ , t ) ) β 0 1 V z ( x , y , z + Λ 2 ρ σ τ , t ) ( V ̲ ( x , y , z + Λ 2 ρ σ τ , t ) ) β d τ 2 ρ σ e 2 N 1 ρ σ β 0 1 V z ( x , y , z + Λ 2 ρ σ τ , t ) ( V ̲ ( x , y , z + Λ 2 ρ σ τ , t ) ) β d τ 2 ρ σ 3 2 1 3 ρ = σ ,

which infers

V ( x , y , z + Λ 2 ρ σ , t ) V ( x , y , z + Λ , t ) σ ϕ β c * c ( z + Λ 2 ρ σ + φ ( α x ) α ) , t .

For x R 3 with z + h ( x , y ) > R * and t [ 0 , T ] , it follows from (4.9) and (4.10) that

V ( x , y , z + Λ 2 ρ σ , t ) V ( x , y , z + Λ , t ) = 2 ρ σ 0 1 V z ( x , y , z + Λ 2 ρ σ τ , t ) d τ 2 ρ σ 1 4 ρ = 1 2 σ σ ϕ β c * c ( z + h ( x , y ) ) , t σ ϕ β c * c ( z + Λ 2 ρ σ + φ ( α x ) α ) , t .

Finally, we obtain

V * ( x , y , z , t ) < V ( x , y , z + Λ 2 ρ σ , t ) + σ ϕ β c * c ( z + Λ 2 ρ σ + φ ( α x ) α ) , t

for x R 3 and t [ 0 , T ] . By Lemma 4.3 and the comparison principle, we have

V * ( x , y , z , t ) H + ( x , y , z , t ; β , α , σ , Λ 2 ρ σ ) , x R 3 , t 0 .

Let k + in

V * ( x , y , z , t ) = V * ( x , y , z , k T + t ) V ( x , y , z + Λ 2 ρ σ + ρ σ ( 1 e κ ( k T + t ) ) , k T + t ) + σ e κ ( k T + t ) ( ω ( θ ) ν ( k T + t ) + ( 1 ω ( θ ) ) ϕ β ( θ , k T + t ) ) V ( x , y , z + Λ 2 ρ σ + ρ σ ( 1 e κ ( k T + t ) ) , t ) + σ e κ ( k T + t ) ( ω ( θ ) ν ( k T + t ) + ( 1 ω ( θ ) ) ϕ β ( θ , k T + t ) )

for t [ 0 , T ] , we have

V * ( x , y , z , t ) V ( x , y , z + Λ ρ σ , t ) , x R 3 , t [ 0 , T ] ,

which contradicts the definition of Λ . Thus, Λ = 0 and the proof is thereby completed.□

We now give the proof of Theorem 1.2.

Proof of Theorem 1.2

By Lemma 3.9, we can define

A 1 = sup x R 3 , t [ 0 , T ] V z ( x , t ) V ̲ β ( x , t ) and A 2 = sup x R 3 , t [ 0 , T ] , ι [ 0 , 1 ] V ̲ β ( x , y , z + ι , t ) V ̲ β ( x , t ) .

Let ρ be defined in (4.5). For any ε > 0 , we take 0 < δ < min ε + ( β ) , ε 1 2 P 2 ( c c * ) , ε 1 2 P 2 , ε 3 ρ A 1 A 2 small enough such that

V ( x , y , z + ρ δ , t ) V ( x , y , z , t ) = ρ δ 0 1 V z ( x , y , z + τ ρ δ , t ) d τ = ρ δ V ̲ β ( x , t ) 0 1 V z ( x , y , z + τ ρ δ , t ) V ̲ β ( x , y , z + τ ρ δ , t ) V ̲ β ( x , y , z + τ ρ δ , t ) V ̲ β ( x , y , z , t ) d τ ε 3 V ̲ β ( x , t ) ,

which implies

(4.11) V ( x , y , z + ρ δ , t ) V ( x , t ) + ε 3 V ̲ β ( x , t ) , x R 3 , t 0 .

By Lemma 3.6, there exists a sufficiently large integer K 1 , such that

lim R + sup x R v ( x , t + k T ; v 0 ) V ( x , t ) V ̲ β ( x , t ) < δ 2 for any fixed k K 1 .

Thus, there exists R δ > 0 such that

(4.12) v ( x , K 1 T + t ; v 0 ) V ( x , t ) + δ 2 V ̲ β ( x , t ) , x R δ , t 0 ,

which implies

(4.13) v ( x , K 1 T + t ; v 0 ) V ¯ ( x , t ; β , δ , α ) + δ V ̲ β ( x , t ) V + ( x , t ; β , ρ , δ , δ , 0 , α ) , x R δ , t 0 .

Let 0 < α < min α + ( ε , β ) , Π ( β c * ) 48 c * K 1 be small enough such that

(4.14) ϕ ( μ ˆ ( x ) , t ) = ϕ z + 1 α φ ( α x , α y ) 1 + φ ( α x , α y ) 2 , t 1 δ 2 min x R δ V ̲ β ( x , t ) , x R δ , t 0 .

Consequently, we obtain

(4.15) V ¯ ( x , t ; β , δ , α ) 1 δ 2 V ̲ β ( x , t ) , x R δ , t 0 .

By v 0 ( x ) C ( R 3 , [ 0 , 1 ] ) and the comparison principle, we have

(4.16) 0 < v ( x , t ; v 0 ) < 1 , x R 3 , t 0 .

Taking into account (4.15) and (4.16), we obtain

(4.17) v ( x , K 1 T + t ; v 0 ) < 1 < V ¯ ( x , t ; β , δ , α ) + δ 2 V ̲ β ( x , t ) V ¯ ( x , t ; β , δ , α ) + δ V ̲ β ( x , t ) V + ( x , t ; β , ρ , δ , δ , 0 , α )

for x R δ and t 0 . Combining (4.13) and (4.17), we obtain

v ( x , K 1 T + t ; v 0 ) V ¯ ( x , t ; β , δ , α ) + δ V ̲ β ( x , t ) V + ( x , t ; β , ρ , δ , δ , 0 , α ) , x R 3 , t 0 .

By Lemma 4.4 and the comparison principle, we have

v ( x , t + t 1 + K 1 T ; v 0 ) v ( x , t 1 ; V t ) , x R 3 , t 0 , t 1 0 ,

where V t ( x ) V + ( x , t ; β , ρ , δ , δ , 0 , α ) . Since v ( x , t + k T ; V ¯ ) monotonically converges to V * ( x , t ) as k + , by (4.1) and Lemma 3.10, there exists a positive integer K 2 such that

(4.18) sup x R 3 v ( x , t + K 2 T ; V ¯ δ ( x ) ) V * ( x , y , z + ρ δ , t ) V ̲ β ( x , t ) ε 3 , k > K 2 , t [ 0 , T ] ,

where V ¯ δ ( x ) V ¯ ( x , y , z + ρ δ , 0 ; β , δ , α ) . It follows from Lemmas 3.12 and 4.4 that there exists an integer K 3 sufficiently large such that

(4.19) sup x R 3 v ( x , t + K 2 T ; V k T ( x ) ) v ( x , t + K 2 T ; V ¯ δ ( x ) ) V ̲ β ( x , t ) 8 M β sup x R 3 V + ( x , k T ; β , ρ , δ , δ , 0 , α ) V ¯ ( x , y , z + ρ δ , 0 ; β , δ , α ) V ̲ β ( x , 0 ) ε 3

for k K 3 and t [ 0 , T ] . Then, by (4.18) and (4.19), we obtain

v ( x , t + K 2 T ; V k T ( x ) ) V * ( x , y , z + ρ δ , t ) = v ( x , t + K 2 T ; V 0 ( x ) ) V * ( x , y , z + ρ δ , t ) v ( x , t + K 2 T ; V 0 ( x ) ) v ( x , t + K 2 T ; V ¯ δ ( x ) ) + v ( x , t + K 2 T ; V ¯ δ ( x ) ) V * ( x , y , z + ρ δ , t ) 2 ε 3 V ̲ β ( x , t )

for k K 3 , t [ 0 , T ] and x R 3 . Thus, we have

v ( x , t + K 1 T + K 2 T + k T ; v 0 ) v ( x , t + K 2 T ; V k t ) V * ( x , y , z + ρ δ , t ) + 2 ε 3 V ̲ β ( x , t )

for k K 3 , t [ 0 , T ] and x R 3 . Let K = K 1 + K 2 + K 3 . Lemma 4.5 implies that

v ( x , t + K T ; v 0 ) V * ( x , y , z + ρ δ , t ) + 2 ε 3 V ̲ β ( x , t ) = V ( x , y , z + ρ δ , t ) + 2 ε 3 V ̲ β ( x , t ) ,

for k K , t [ 0 , T ] and x R 3 . Combining with (4.11), we have

(4.20) v ( x , t + K T ; v 0 ) V ( x , t ) + ε V ̲ β ( x , t ) , x R 3 , k K , t [ 0 , T ] .

On the other hand, since V ̲ ( x , 0 ) v 0 ( x ) , the comparison principle yields

v ( x , t ; V ̲ ) v ( x , t ; v 0 ) , x R 3 , t 0 .

Thus, it follows from Lemma 3.10 that there exists a constant K N large enough such that

(4.21) v ( x , t + K T ; v 0 ) v ( x , t + K T ; V ̲ ) V ( x , t ) ε V ̲ β ( x , t ) , x R 3 , k K , t [ 0 , T ] .

Combining with (4.20) and (4.21), we obtain

lim k + sup x R 3 v ( x , t + k T ; v 0 ) V ( x , t ) V ̲ β ( x , t ) = 0 .

This completes the proof.□

Acknowledgements

The authors would like to thank the anonymous reviewers for their valuable comments and many useful suggestions. They also thank Mingyou Zhang for helpful discussions and suggestions.

  1. Funding information: Yuan-Hao Liu and Zhen-Hui Bu were supported by the National Natural Science Foundation of China (Nos. 12101499 and 12471163).

  2. Author contributions: All authors have accepted responsibility for the entire content of this manuscript and consented to its submission to the journal, reviewed all the results, and approved the final version of the manuscript. All authors have contributed more or less equally to the contribution of this manuscript.

  3. Conflict of interest: The authors state no conflict of interest.

  4. Data availability statement: Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.

References

[1] X.-X. Bao, W.-T. Li, and Z.-C. Wang, Time periodic traveling curved fronts in the periodic Lotka-Volterra competition-diffusion system, J. Dynam. Differential Equations 29 (2017), 981–1016, https://doi.org/10.1007/s10884-015-9512-4. Search in Google Scholar

[2] W.-J. Bo, G. Lin, and Y. Qi, Propagation dynamics of a time periodic diffusion equation with degenerate nonlinearity, Nonlinear Anal. Real World Appl. 45 (2019), 376–400, https://doi.org/10.1016/j.nonrwa.2018.07.010. Search in Google Scholar

[3] A. Bonnet and F. Hamel, Existence of nonplanar solutions of a simple model of premixed bunsen flames, SIAM J. Math. Anal. 31 (1999), 80–118, https://doi.org/10.1137/S0036141097316391. Search in Google Scholar

[4] Z.-H. Bu, C.-L. Wang, and X.-T. Zhang, Pyramidal traveling fronts of a time periodic diffusion equation with degenerate monostable nonlinearity, Electron. J. Differential Equations 2023 (2023), no. 31, 1–23, https://doi.org/10.58997/ejde.2023.31. Search in Google Scholar

[5] Z.-H. Bu, and Z.-C. Wang, Stability of pyramidal traveling fronts in the degenerate monostable and combustion equations I, Discrete Contin. Dyn. Syst. 37 (2017), 2395–2430, https://doi.org/10.3934/dcds.2017104. Search in Google Scholar

[6] Z.-H. Bu and Z.-C. Wang, Global stability of V-shaped traveling fronts in combustion and degenerate monostable equations, Discrete Contin. Dyn. Syst. 38 (2018), 2251–2286, https://doi.org/10.3934/dcds.2018093. Search in Google Scholar

[7] M. El Smaily, F. Hamel, and R. Huang, Two-dimensional curved fronts in a periodic shear flow, Nonlinear Anal. Theory Methods Appl. 74 (2011), 6469–6486, https://doi.org/10.1016/j.na.2011.06.030. Search in Google Scholar

[8] P. C. Fife and J. B. McLeod, The approach of solutions of nonlinear diffusion equations to travelling front solutions, Arch. Rational Mech. Anal. 65 (1977), 335–361, https://doi.org/10.1007/BF00250432. Search in Google Scholar

[9] R. A. Fisher, The wave of advance of advantageous genes, Ann. Eugen. 7 (1937), 355–369, https://doi.org/10.1111/j.1469-1809.1937.tb02153.x. Search in Google Scholar

[10] F. Gazzola and T. Weth, Finite time blow-up and global solutions for semilinear parabolic equations with initial data at high energy level, Differential and Integral Equations 18 (2005), 961–990, https://doi.org/10.57262/die/1356060117. Search in Google Scholar

[11] F. Hamel and N. Nadirashvili, Travelling fronts and entire solutions of the Fisher-KPP equation in RN, Arch. Rational Mech. Anal. 157 (2001), 91–163, https://doi.org/10.1007/PL00004238. Search in Google Scholar

[12] F. Hamel and R. Monneau, Solutions of semilinear elliptic equations in RN with conical & shaped level sets, Comm. Partial Differential Equations 25 (2000), 769–819, DOI: https://doi.org/10.1080/03605300008821532. 10.1080/03605300008821532Search in Google Scholar

[13] F. Hamel, R. Monneau, and J.-M. Roquejoffre, Stability of conical fronts in a model for conical flames in two space dimensions, Ann. Sci. École Norm. Sup. 37 (2004), 469–506, https://doi.org/10.1016/j.ansens.2004.03.001. Search in Google Scholar

[14] J. Huang and W. Shen, Speeds of spread and propagation for kpp models in time almost and space periodic media, SIAM J. Appl. Dyn. Syst. 8 (2009), 790–821, https://doi.org/10.1137/080723259. Search in Google Scholar

[15] R. Huang, Stability of travelling fronts of the Fisher-KPP equation in RN, Nonlinear Differ. Equ. Appl. 15 (2008), 599–622, https://doi.org/10.1007/s00030-008-7041-0. Search in Google Scholar

[16] F.-J. Jia, X. Wang, and Z.-C. Wang, Traveling curved fronts in the buffered bistable systems in R2, Commun. Nonlinear Sci. Numer. Simul. 112 (2022), 106529, https://doi.org/10.1016/j.cnsns.2022.106529. Search in Google Scholar

[17] W. Lian, J. Wang, and R.-Z. Xu, Global existence and blow up of solutions for pseudo-parabolic equation with singular potential, J. Differential Equations 269 (2020), 4914–4959, https://doi.org/10.1016/j.jde.2020.03.047. Search in Google Scholar

[18] X. Liang and X.-Q. Zhao, Asymptotic speeds of spread and traveling waves for monotone semiflows with applications, Commun. Pure Appl. Math. 60 (2007), 1–40, https://doi.org/10.1002/cpa.20154. Search in Google Scholar

[19] G. M. Lieberman, Second Order Parabolic Differential Equations, World Scientific Publishing Co., Inc., River Edge, NJ, 1996, https://doi.org/10.1142/3302. Search in Google Scholar

[20] K. Mischaikow and V. Hutson, Travelling waves for mutualist species, SIAM J. Math. Anal. 24 (1993), 987–1008, https://doi.org/10.1137/0524059. Search in Google Scholar

[21] H. Ninomiya and M. Taniguchi, Existence and global stability of traveling curved fronts in the allen-cahn equations, J. Differential Equations 213 (2005), 204–233, https://doi.org/10.1016/j.jde.2004.06.011. Search in Google Scholar

[22] H. Ninomiya and M. Taniguchi, Global stability of traveling curved fronts in the Allen-Cahn equations, Discrete Contin. Dyn. Syst. 15 (2006), 819–832, https://doi.org/10.3934/dcds.2006.15.819. Search in Google Scholar

[23] H.-T. Niu, Z.-H. Bu, and Z.-C. Wang, Global stability of curved fronts in the Belousov-Zhabotinskii reaction–diffusion system in R2, Nonlinear Anal. Real World Appl. 46 (2019), 493–524, https://doi.org/10.1016/j.nonrwa.2018.10.003. Search in Google Scholar

[24] H.-T. Niu, Z.-C. Wang, and Z.-H. Bu, Curved fronts in the belousov-zhabotinskii reaction–diffusion systems in R2, J. Differential Equations 264 (2018), 5758–5801, https://doi.org/10.1016/j.jde.2018.01.020. Search in Google Scholar

[25] L. E. Payne and D. H. Sattinger, Saddle points and instability of nonlinear hyperbolic equations, Israel J. Math. 22 (1975), 273–303, https://doi.org/10.1007/BF02761595. Search in Google Scholar

[26] W. Shen, Traveling waves in time almost periodic structures governed by bistable nonlinearities: I. stability and uniqueness, J. Differential Equations 159 (1999), 1–54, https://doi.org/10.1006/jdeq.1999.3651. Search in Google Scholar

[27] W. Shen, Traveling waves in time almost periodic structures governed by bistable nonlinearities: II. existence, J. Differential Equations 159 (1999), 55–101, https://doi.org/10.1006/jdeq.1999.3652. Search in Google Scholar

[28] W. Shen and Z. Shen, Transition fronts in time heterogeneous and random media of ignition type, J. Differential Equations 262 (2017), 454–485, https://doi.org/10.1016/j.jde.2016.09.030. Search in Google Scholar

[29] W.-J. Sheng, Time periodic traveling curved fronts of bistable reaction–diffusion equations in R3, Ann. Mat. Pura Appl. 196 (2017), 617–639, https://doi.org/10.1016/j.aml.2015.11.004. Search in Google Scholar

[30] W.-J. Sheng, W.-T. Li, and Z.-C. Wang, Periodic pyramidal traveling fronts of bistable reaction–diffusion equations with time-periodic nonlinearity, J. Differential Equations 252 (2012), 2388–2424, https://doi.org/10.1016/j.jde.2011.09.016. Search in Google Scholar

[31] M. Taniguchi, Traveling fronts of pyramidal shapes in the Allen-Cahn equations, SIAM J. Math. Anal. 39 (2007), 319–344, https://doi.org/10.1137/060661788. Search in Google Scholar

[32] M. Taniguchi, The uniqueness and asymptotic stability of pyramidal traveling fronts in the allen-cahn equations, J. Differential Equations 246 (2009), 2103–2130, https://doi.org/10.1016/j.jde.2008.06.037. Search in Google Scholar

[33] M. Taniguchi, Multi-dimensional traveling fronts in bistable reaction–diffusion equations, Discrete Contin. Dyn. Syst. 32 (2012), 1011–1046, https://doi.org/10.3934/dcds.2012.32.1011. Search in Google Scholar

[34] S.-M. Wang, Z. Feng, Z.-C. Wang, and L. Zhang, Spreading speed and periodic traveling waves of a time periodic and diffusive SI epidemic model with demographic structure, Commun. Pure Appl. Anal. 21 (2022), 2005–2034, https://doi.org/10.3934/cpaa.2021145. Search in Google Scholar

[35] X.-C. Wang and R.-Z. Xu, Global existence and finite-time blowup for a nonlocal semilinear pseudo-parabolic equation, Adv. Nonlinear Anal. 10 (2021), 261–288, https://doi.org/10.1515/anona-2020-0141. Search in Google Scholar

[36] Z.-C. Wang, Traveling curved fronts in monotone bistable systems, Discrete Contin. Dyn. Syst. 32 (2012), 2339–2374, https://doi.org/10.3934/dcds.2012.32.2339. Search in Google Scholar

[37] Z.-C. Wang and Z.-H. Bu, Nonplanar traveling fronts in reaction–diffusion equations with combustion and degenerate Fisher-KPP nonlinearities, J. Differential Equations 260 (2016), 6405–6450, https://doi.org/10.1016/j.jde.2015.12.045. Search in Google Scholar

[38] Z.-C. Wang and J. Wu, Periodic traveling curved fronts in reaction–diffusion equation with bistable time-periodic nonlinearity, J. Differential Equations 250 (2011), 3196–3229, https://doi.org/10.1016/j.jde.2011.01.017. Search in Google Scholar

[39] R.-Z. Xu, W. Lian, and Y. Niu, Global well-posedness of coupled parabolic systems, Sci. China Math. 63 (2019), 1–36, https://doi.org/10.1007/s11425-017-9280-x. Search in Google Scholar

[40] R.-Z. Xu and J. Su, Global existence and finite-time blow-up for a class of semilinear pseudo-parabolic equations, J. Funct. Anal. 264 (2013), 2732–2763, https://doi.org/10.1016/j.jfa.2013.03.010. Search in Google Scholar

[41] S. Zhang, Z.-H. Bu, and Z.-C. Wang, Periodic curved fronts in reaction–diffusion equations with ignition time-periodic nonlinearity, Discrete Contin. Dyn. Syst. Ser. B 28 (2023), 2621–2654, https://doi.org/10.3934/dcdsb.2022185. Search in Google Scholar

[42] S. Zhang, Z.-H. Bu, and Z.-C. Wang, Curved fronts in time-periodic reaction–diffusion equations with monostable nonlinearity, Discrete Contin. Dyn. Syst. Ser. S 17 (2024), 547–584, https://doi.org/10.3934/dcdss.2023043. Search in Google Scholar

Received: 2024-10-11
Revised: 2025-02-19
Accepted: 2025-02-24
Published Online: 2025-03-31

© 2025 the author(s), published by De Gruyter

This work is licensed under the Creative Commons Attribution 4.0 International License.

Articles in the same Issue

  1. Research Articles
  2. Incompressible limit for the compressible viscoelastic fluids in critical space
  3. Concentrating solutions for double critical fractional Schrödinger-Poisson system with p-Laplacian in ℝ3
  4. Intervals of bifurcation points for semilinear elliptic problems
  5. On multiplicity of solutions to nonlinear Dirac equation with local super-quadratic growth
  6. Entire radial bounded solutions for Leray-Lions equations of (p, q)-type
  7. Combinatorial pth Calabi flows for total geodesic curvatures in spherical background geometry
  8. Sharp upper bounds for the capacity in the hyperbolic and Euclidean spaces
  9. Positive solutions for asymptotically linear Schrödinger equation on hyperbolic space
  10. Existence and non-degeneracy of the normalized spike solutions to the fractional Schrödinger equations
  11. Existence results for non-coercive problems
  12. Ground states for Schrödinger-Poisson system with zero mass and the Coulomb critical exponent
  13. Geometric characterization of generalized Hajłasz-Sobolev embedding domains
  14. Subharmonic solutions of first-order Hamiltonian systems
  15. Sharp asymptotic expansions of entire large solutions to a class of k-Hessian equations with weights
  16. Stability of pyramidal traveling fronts in time-periodic reaction–diffusion equations with degenerate monostable and ignition nonlinearities
  17. Ground-state solutions for fractional Kirchhoff-Choquard equations with critical growth
  18. Existence results of variable exponent double-phase multivalued elliptic inequalities with logarithmic perturbation and convections
  19. Homoclinic solutions in periodic partial difference equations
  20. Classical solution for compressible Navier-Stokes-Korteweg equations with zero sound speed
  21. Properties of minimizers for L2-subcritical Kirchhoff energy functionals
  22. Asymptotic behavior of global mild solutions to the Keller-Segel-Navier-Stokes system in Lorentz spaces
  23. Blow-up solutions to the Hartree-Fock type Schrödinger equation with the critical rotational speed
  24. Qualitative properties of solutions for dual fractional parabolic equations involving nonlocal Monge-Ampère operator
  25. Regularity for double-phase functionals with nearly linear growth and two modulating coefficients
  26. Uniform boundedness and compactness for the commutator of an extension of Riesz transform on stratified Lie groups
  27. Normalized solutions to nonlinear Schrödinger equations with mixed fractional Laplacians
  28. Existence of positive radial solutions of general quasilinear elliptic systems
  29. Low Mach number and non-resistive limit of magnetohydrodynamic equations with large temperature variations in general bounded domains
  30. Sharp viscous shock waves for relaxation model with degeneracy
  31. Bifurcation and multiplicity results for critical problems involving the p-Grushin operator
  32. Asymptotic behavior of solutions of a free boundary model with seasonal succession and impulsive harvesting
  33. Blowing-up solutions concentrated along minimal submanifolds for some supercritical Hamiltonian systems on Riemannian manifolds
  34. Stability of rarefaction wave for relaxed compressible Navier-Stokes equations with density-dependent viscosity
  35. Singularity for the macroscopic production model with Chaplygin gas
  36. Global strong solution of compressible flow with spherically symmetric data and density-dependent viscosities
  37. Global dynamics of population-toxicant models with nonlocal dispersals
  38. α-Mean curvature flow of non-compact complete convex hypersurfaces and the evolution of level sets
  39. High-energy solutions for coupled Schrödinger systems with critical growth and lack of compactness
  40. On the structure and lifespan of smooth solutions for the two-dimensional hyperbolic geometric flow equation
  41. Well-posedness for physical vacuum free boundary problem of compressible Euler equations with time-dependent damping
  42. On the existence of solutions of infinite systems of Volterra-Hammerstein-Stieltjes integral equations
  43. Remark on the analyticity of the fractional Fokker-Planck equation
  44. Continuous dependence on initial data for damped fourth-order wave equation with strain term
  45. Unilateral problems for quasilinear operators with fractional Riesz gradients
  46. Boundedness of solutions to quasilinear elliptic systems
  47. Existence of positive solutions for critical p-Laplacian equation with critical Neumann boundary condition
  48. Non-local diffusion and pulse intervention in a faecal-oral model with moving infected fronts
Downloaded on 1.10.2025 from https://www.degruyterbrill.com/document/doi/10.1515/anona-2025-0079/html
Scroll to top button