betaFeedback

WorksPhysics and Astronomy

Breathers across the Focusing-to-Defocusing Transition: From Diffraction-Like Dynamics to Talbot Carpets

Abstract

We experimentally investigate fundamental unstable breather wave groups propagating across the transition from focusing to defocusing regimes. In hydrodynamics, this transition is realized through a rapid reduction in water depth over a short spatial interval. We demonstrate that the time-periodic Akhmediev breather develops a Talbot-like recurrence pattern in the defocusing regime, although only its initial stages are experimentally accessible. In contrast, the Peregrine breather exhibits qualitatively different behavior, characterized by oscillatory envelope dynamics. These oscillations can be understood as a purely linear phenomenon arising within the nonlinear Schrödinger framework when dispersion dominates over nonlinearity, as further confirmed by numerical simulations over extended propagation distances. The experimental observations agree closely with finite-depth nonlinear Schrödinger predictions across regimes with substantially different relative strengths of dispersion and nonlinearity. Our findings reveal a temporal analogue of diffraction and provide a foundation for exploring focusing-to-defocusing transitions in other nonlinear dispersive media.

Introduction

Transitions between different propagation regimes are ubiquitous in physical systems. Nonlinear hydrodynamic waves are particularly sensitive to variations in water depth, which play a critical role in coastal hydrodynamics by governing processes such as wave shoaling and run-up [1, 2, 3, 4]; nonlinear wave interactions radically modify wave statistics [5, 6, 7, 8, 9, 10], while variations in group-velocity dispersion and the Kerr nonlinear coefficient strongly influence the dynamics of solitary waves in nonlinear optics [11, 12]. Thus far, only a few theoretical and experimental studies have investigated the evolution of modulated wave packets such as solitons and breathers in non-uniform media [13, 14, 15, 16, 17].

A key step towards understanding the effects of regime transitions on nonlinear waves can be obtained by studying breathers, beating wave groups describing the nonlinear stage of modulational instability (MI) of waves propagating in a uniform medium and strong candidates for modeling rogue waves [18, 19, 20]. Within the nonlinear Schrödinger equation (NLSE) framework [21, 22], two fundamental solutions occurring in the focusing regime, i.e., the Akhmediev breathers (AB) [23] and the doubly-localized Peregrine breather (PB) [24], have been extensively studied with a striking agreement between theory and observations [25, 26, 27, 28] and have been shown to be the fundamental building blocks to trigger and control complex dynamical phenomena in a variety of nonlinear dispersive systems [29, 30, 31, 32, 33].

Here, we report an experimental study analyzing the propagation of modulationally unstable wave groups in the form of ABs and PBs over a rapid depth variation in a large-scale water wave facility, which implies a transition from the focusing to the defocusing regime, i.e., a sign change of the nonlinearity in the hydrodynamic NLSE [34, 35]. Our data allow us to explore the long-term behavior of these two types of fundamental wave groups before and after the transition. The experiments show that ABs exhibit the initial stages of Talbot-like pattern formation upon transitioning into a nearly linear regime, while PBs transform into an oscillating pattern that spreads while propagating. This can be interpreted as a time-analogue of diffraction, since linear effects dominate over nonlinear ones. This latter phenomenon is sensitive to the phase of the wave packet at the point where nonlinearity ceases to be effective. These findings are expected to stimulate further experimental, numerical, and theoretical work, not only in water waves, but also in optics, plasma, transmission lines, and Bose-Einstein condensates, to elucidate which linear and nonlinear effects are triggered by rapid dispersion and nonlinearity variations in the medium.

Results

We have conducted numerous experiments on AB and PB transitions across a range of breather and carrier wave parameters, but here we present only four representative cases, two for each family of breather boundary conditions, to illustrate the underlying physical phenomena investigated in this study. The experiments have been performed in the large-scale facility of the Tainan Hydraulics Laboratory (THL), Taiwan, with dimensions 200×2×2 m3200 \times2 \times2\ \mathrm{m}^3, suitable for studying coastal processes and tracking the long-term evolution of hydrodynamic waves. The long-crested waves are generated by a hydraulic piston-type wave maker, suitable for generating waves in finite water depth. An artificial concrete bathymetry was installed to emulate a rapid depth transition from h1=1.2 mh_1 = 1.2\ \mathrm{m} to h2=0.2 mh_2 = 0.2\ \mathrm{m} for the AB wave packet (respectively, h1=1.3 mh_1 = 1.3\ \mathrm{m} to h2=0.3 mh_2 = 0.3\ \mathrm{m} for the PB wave packet) through a linear transition over a distance of 20 m20\ \mathrm{m}, resulting in a 1/20 slope. Moreover, 60 resistive wave gauges have been deployed along the facility to track the wave packet progression in the different depth zones. Details of the experimental setup are summarized and shown in Figure 1. This bathymetric transition can be considered rapid, not relative to the carrier wavelength (of the order of one meter), but rather to the characteristic length scale of nonlinear evolution described by the FDNLSE, which exceeds 20 m for the typical initial carrier wave parameters considered here.

Large-scale unidirectional water wave tank schematic

Figure 1. Large-scale unidirectional water wave tank schematic, depicting the piston-type wave maker, the wave gauge locations, the bathymetry characteristic change, and the wave absorption installations. The initial depth is h1=1.2 mh_1 = 1.2\ \mathrm{m} for AB initial conditions (respectively, h1=1.3 mh_1 = 1.3\ \mathrm{m} for the PB) and becomes h2=0.2 mh_2 = 0.2\ \mathrm{m} (respectively, h2=0.3 mh_2 = 0.3\ \mathrm{m}) after the 1/20 slope transition.

Assuming a rapid, linear bathymetric transition from an initial depth h1h_1 to a final depth h2h_2, the wavenumber kk along the wave tank can be determined from the linear surface gravity wave dispersion relation, ω2=gktanh⁡(kh)\omega^2 = gk \tanh(kh), where ω=2πf\omega= 2\pi f is the angular frequency and gg denotes gravitational acceleration. In order to model the evolution of the envelope along the tank, we employ the finite-depth NLSE (FDNLSE), where the coefficients depend explicitly on the product κ=kh\kappa= kh and shoaling is neglected (see Methods). Importantly, at κ=1.363\kappa= 1.363, the nonlinearity vanishes, allowing for a transition from the focusing to defocusing regime [36]. The boundary conditions, i.e., the respective wave packets as generated by the wave maker, are determined by the parametrization of exact FDNLSE solutions and the carrier wave parameters, e.g., wave frequency and wave steepness, so that the maximal theoretical focusing is expected to occur near the junction between focusing and defocusing regimes, that is, when κ=1.363\kappa= 1.363 is satisfied on the slope. Note that the wave envelope is extracted from the measured surface elevation data using the Hilbert transform, enabling a direct comparison between the observed wave evolution and theoretical predictions based on the FDNLSE [37, 38].

Figure 2 illustrates the evolution of the two AB wave envelopes propagating in the wave flume with the rapid change in water depth over the variable bathymetry. The first describes the case of classical modulation instability dynamics at its maximal growth rate, i.e., following Ref. [38], a = 0.25, while the second is for a larger perturbation period, close to the center of the MI frequency range, which corresponds to a = 0.37, i.e., the limit where higher-order modulation instability appears [39, 40].

Evolution of two AB envelopes normalized by the initial carrier amplitude as propagating over the variable bathymetry in the wave flume

Figure 2. Evolution of two AB envelopes normalized by the initial carrier amplitude as propagating over the variable bathymetry in the wave flume. Left: Experiments. Right: Corresponding FDNLSE simulations. (a) and (b) AB modeling the case of maximal growth rate for α=0.25\alpha= 0.25 for the carrier steepness ε=0.1\varepsilon= 0.1 and frequency f=1f = 1 Hz. (c) and (d) AB modeling the case of higher-order MI for α=0.37\alpha= 0.37 and adopting the same carrier parameters as in the previous case. The vertical dashed line denotes the location of the transition at κ=1.363\kappa= 1.363 or γ=0\gamma= 0.

The carrier wave parameters are provided in the figure caption. Note that the AB envelopes are parameterized as in [38] and are recalled in Methods. From these measurements, we draw three conclusions: first, we can clearly observe that unstable waves keep their coherence as well as their focusing trend on the slope; second, the rogue-like wave patterns do not immediately disappear when entering the defocusing regime; third, when propagating further in the tank starting from the initial κ1=4.80\kappa_1 = 4.80 into a shallower regime satisfying κ2=1.02\kappa_2 = 1.02, the wave packets first broaden, then break into a characteristic periodic pattern. We will discuss below the long-term evolution beyond experimental constraints. Both measurements are in very good agreement with the FDNLSE simulations, accounting for the same bathymetry change; the dissipation coefficient has been determined from the measured wave attenuation and is set to D=3.8×10−3 m−1\mathcal{D} = 3.8 \times10^{-3}\ \mathrm{m}^{-1}. The phenomenology in the defocusing regime exhibits for both values of a a pattern of finer and finer fringes with the successive appearance of secondary refocusing events. This hints at a phenomenology similar to Talbot carpets [41] in diffraction theory, where the fringes generated by a diffraction grating interact constructively or destructively, ideally producing a self-image of the grating itself at well-predicted lengths. However, despite the considerable length of the wave facility, only the initial stages of these dynamics could be observed experimentally. Their subsequent evolution is therefore investigated through numerical simulations extending beyond the experimental domain under identical physical parameters, as discussed below.

Next, we focus on the rapid depth-transition dynamics of the PB, which features distinct and interesting dynamics. The experimental results of two runs are shown in Figure 3. We present two slightly different initial steepness values, as reported in the caption. While the initial focusing process in the deeper part of the tank is similar to the time-periodic AB observations, the evolution of the PB differs after transitioning into the shallow area. In these experiments, the depth transition is from κ1=3.62\kappa_1 = 3.62 to κ2=1.05\kappa_2 = 1.05. We clearly observe that an oscillating and diverging pattern emerges. The PB is characterized by two symmetric dips close to the rogue event: similar fission wave dynamics has been suggested in a numerical work in which the PB is propagating in the presence of shear currents of slowly varying vorticity [42]. In simulations, Figure 3(b,d), where the dissipation coefficient is set to D=3.6×10−3 m−1\mathcal{D} = 3.6 \times10^{-3}\ \mathrm{m}^{-1}, it is apparent instead that the two dips tend to spread during propagation. Between these two depressions, the PB peak broadens to an almost uniform background, gently decaying during evolution as a result of overall dissipation. It is worth noting how the pattern develops in propagation on the two temporal sides of the envelope: an oscillating pattern alternating peaks and dips is clearly visible. A closer inspection shows that this pattern does not diverge with a regular speed, as could be expected from a dispersive shock wave (DSW) fan modeled by the Korteweg-de Vries-type or Whitham evolution equations [43, 44, 45, 46]. This evolution can therefore be interpreted as a linear diffraction-like phenomenon, as wave damping and the resulting weakening of nonlinear effects

Evolution of two PB envelopes normalized by the initial carrier amplitude as propagating over the variable bathymetry in the wave flume

Figure 3. Evolution of two PB envelopes normalized by the initial carrier amplitude as propagating over the variable bathymetry in the wave flume. The carrier frequency is f=0.83f = 0.83 Hz. Left: Experiments. Right: Corresponding FDNLSE simulations. (a) and (b) steepness ε=0.09\varepsilon= 0.09. (c) and (d) steepness ε=0.11\varepsilon= 0.11. The vertical lines share the same meaning as in Figure 2.

prevent the onset of the DSW regime within the experimentally accessible propagation distance. This will be further justified by an approximate linear theory presented in the Methods section.

The corresponding spatial evolution of the normalized spectra of these cases is shown in Figure 4. Figure 4(a,Figure 4b) show the evolution of the AB spectra. As it is well understood, the evolution in the deeper region is characterized by the cascaded generation of phase-locked frequency components: apart from the carrier, two pairs of sidebands are clearly identified. In Figure 4(b), they are more closely packed, as expected from the larger value of the breather parameter α\alpha. As soon as the wave propagates in the defocusing regime, this process is almost halted. We observe weak oscillations in space that account for a quasi-periodic conversion between the carrier and the sideband pairs, as occurs outside of MI sidebands [47]. In contrast, the PB exhibits qualitatively different behavior, with an initial nearly symmetric spectral broadening in the focusing regime followed by additional asymmetric broadening. Within the limitations of spectral reconstruction from noisy time traces, this looks more apparent for the larger steepness, so it may be of nonlinear origin. The spectrum narrows by the end of the tank because of wave damping. This also hints at a possible interpretation of the dips as the transient appearance of a gray soliton after the depth change: the quantitative clarification of their nature will be addressed in a future work.

Evolution of the respective normalized frequency spectra

Figure 4. Evolution of the respective normalized frequency spectra log⁡10[S(f)/max⁡(S(f))]\log_{10}[S(f)/\max(S(f))]: (a) the case of Figure 2(a), (b) the case of Figure 2(c), (c) the case of Figure 3(a), (d) the case of Figure 3(c). The vertical lines share the same meaning as in Figure 2.

Extending the numerical simulations to a propagation distance of x=500x = 500 m reveals these phenomena more clearly than is possible within the experimentally accessible domain. In Figure 5(a), we confirm that the AB breaks into an interfering fringe pattern, where the initial expansion is followed by a constructive interference forming a new periodic temporal train of pulses similar to the AB maximal focusing waveform, i.e., a form of recurrence or self-image of the conditions at the depth transition. The process repeats itself, and at each recurrence the maxima are shifted by a half period. This closely resembles an initialization of a Talbot carpet, which, however, becomes weaker and weaker owing to wave attenuation. Conservative FDNLSE simulations, i.e., with D=0\mathcal{D} = 0, over longer propagation distances would reveal the complete Talbot carpet, characterized by its periodic recurrences. This setup and transition phenomenon add to the possibilities for Talbot carpet generation in nonlinear dispersive media, as previously reported, for instance, in [48, 49, 50]. Notice that the simplest Talbot carpet is generated by a diffraction grating, i.e., a mask with periodically spaced slits, corresponding to a square-wave transmission. It was shown that the fringe pattern is fractal between successive recurrences [41]. The present situation is more involved, not only because of wave damping, but most importantly because the wave packet at the beginning of the Talbot pattern differs from a pure grating, both in amplitude and phase.

Comparison of FDNLSE simulations and linearized theory for AB and PB breathers

Figure 5. Comparison of FDNLSE simulation (a)-(b) showing the long-range evolution of an AB and a PB breather with the linearized theory presented in Methods, (c)-(d), normalized by the respective carrier amplitude, over three times the propagation distance in the wave flume. (a) and (c) The same AB case for a=0.25a = 0.25 as in Figure 2 (b) and (d) the same PB case for the carrier steepness ε=0.09\varepsilon= 0.09 as in Figure 3. In (c)-(d) we assume that, after the transition, only the linear dispersion β2\beta_2 is considered, and the approximate solution of the Methods section is used. The bathymetry is assumed to jump at x=50x = 50 m, while the nonlinear solutions are prepared to reach their apex at x=55x = 55 m. (c) AB, reconstructed with 9 Fourier components, which provides sufficient accuracy. (d) PB, based stationary phase approximation; notice that this is valid only for large xx and diverges near the transition. The vertical lines share the same meaning as in Figure 2.

The AB exhibits a constant background, and except at the maximal focusing position, where the envelope can be considered purely real, the background is not in phase with the oscillating part. As for the PB, we now clearly observe in Figure 5(b), that the oscillating tails do not exhibit the typical DSW behavior with a definite speed and the main dips close to the center are damped. Thus, we can interpret the dynamics after the bathymetry change as a pure diffraction-like phenomenon. As with the AB, the background and relative phase of the PB maximum also influence the dynamics, albeit less apparent than in the Talbot pattern, where the diffracting wave components recombine and relative phases matter.

Qualitatively, a rapid depth transition can be considered a step. It is thus worth comparing the simulations of Figure 5(a)-(b) with a simplified theory, where the known conditions given by the AB or PB evolution over the initial uniform bathymetry are considered as the initial conditions for the evolution over the final bathymetry region, where γ2\gamma_2 is now neglected. This is shown in Figure 5(c)-(Figure 5d). It is clear that the evolution is qualitatively the same for both examples. The main discrepancy is observed in the self-image distance, which turns out to be slightly longer than in the FDNLSE simulations. A linearized theoretical description is therefore sufficient to capture these dynamics, as detailed in Methods. This confirms the soundness of the interpretation of the rapid bathymetry change as inducing a time-analogue of diffraction in hydrodynamics.

Discussion

We have reported an experimental study investigating the dynamics of breathers transitioning from the focusing to the defocusing regime, which can be achieved in water waves through a rapid reduction in water depth. Our results confirm that the nonlinear focusing process is not compromised during the transition over the steep slope. However, analogous to diffraction from multiple or single slits, the AB and PB exhibit distinct evolution patterns in the defocusing regime. While attenuating in amplitude in the shallower section of the tank, the AB develops a Talbot carpet-like pattern, whereas the PB evolves into two unsteady amplitude dips with oscillatory tails on either side. This latter temporal analogue of diffraction is well described by a linearized theory. Nevertheless, its full development requires long propagation distances, making experimental investigations in water wave facilities with varying bathymetry challenging. Further work is required to isolate the effects of the depth transition on extreme focusing patterns and to investigate the possibility of observing DSW dynamics. The excellent agreement between the observed complex wave dynamics and FDNLSE predictions paves the way for exploring similar transitions in nonlinear wave dynamics across other dispersive media, including broadband wave fields.

Methods

Weakly Nonlinear Wave Theory in Variable Water Depth

The fundamental framework employed is the time-like finite water depth NLSE (FDNLSE) governing the evolution of wave envelope ψ\psi, as derived from the Euler equations for free surface flows via the method of multiple scales [34, 51]

i(∂ψ∂x+β1∂ψ∂t)=12β2∂2ψ∂t2+γ∣ψ∣2ψ−iDψ(1)i\left(\frac{\partial\psi}{\partial x}+\beta_1\frac{\partial\psi}{\partial t}\right)=\frac{1}{2}\beta_2\frac{\partial^2\psi}{\partial t^2}+\gamma\lvert\psi\rvert^2\psi-iD\psi \tag*{(1)}

where, from the dispersion relation ω=gkσ\omega=\sqrt{gk\sigma}, with σ≡tanh⁡κ\sigma\equiv\tanh\kappa, κ≡kh\kappa\equiv kh, kk the wavenumber, hh the water depth and gg the gravitational acceleration, we derive

β1=1cg={g2ω[σ+κ(1−σ2)]}−1\beta_1=\frac{1}{c_g}=\left\{\frac{g}{2\omega}\left[\sigma+\kappa\left(1-\sigma^2\right)\right]\right\}^{-1}
β2=∂2k∂ω2=−1cg3∂2ω∂k2=1ωcg[1−ghcg2(1−κσ)(1−σ2)]\beta_2=\frac{\partial^2 k}{\partial\omega^2}=-\frac{1}{c_g^3}\frac{\partial^2\omega}{\partial k^2}=\frac{1}{\omega c_g}\left[1-\frac{gh}{c_g^2}(1-\kappa\sigma)(1-\sigma^2)\right]

Moreover, the nonlinear coefficient reads

γ≡ωk216σ4cg{9−10σ2+9σ4−2σ2cg2gh−cg2[4cp2cg2+4cpcg(1−σ2)+ghcg2(1−σ2)2]}.\gamma\equiv\frac{\omega k^2}{16\sigma^4c_g}\left\{9-10\sigma^2+9\sigma^4-\frac{2\sigma^2c_g^2}{gh-c_g^2}\left[4\frac{c_p^2}{c_g^2}+4\frac{c_p}{c_g}(1-\sigma^2)+\frac{gh}{c_g^2}(1-\sigma^2)^2\right]\right\}.

with cg≡∂ω∂kc_g \equiv\frac{\partial\omega}{\partial k} and cp≡ωkc_p \equiv\frac{\omega}{k} denoting group and phase velocity of the carrier wave, respectively. Note that, for simplicity, both the experimental data and the corresponding theoretical predictions are presented in a reference frame moving at the local group velocity cgc_g. Moreover, although more sophisticated wave evolution models, such as that proposed in [52], could have been employed, we retain the classical NLSE because it provides a simpler framework for interpreting the results. Furthermore, the shoaling corrections are expected to have only a minor influence on the wave evolution under the present experimental conditions: given the transition from initial depth h1h_1 to final depth h2h_2, the shoaling coefficient is estimated as Ks=cg1/cg2K_s=\sqrt{c_{g1}/c_{g2}} and thus amounts here to Ks≈0.92K_s\approx0.92. That being said, a phenomenological attenuation coefficient D\mathcal{D} is introduced, as determined from the data. The latter depends on friction between the fluid and the tank’s bottom and sidewalls, as well as on the carrier wave period, and becomes increasingly important for long tank experiments [38]. The carrier amplitude aa and carrier frequency f=ω2πf=\frac{\omega}{2\pi} are first set in the experiments. Given the depth hh, all other related coefficients are computed accordingly. Apart from the damping, Eq. (1) is a standard NLSE, which can describe two distinct wave propagation regimes in uniform water depth, depending on the dimensionless depth parameter κ\kappa. In fact, if κ>1.363\kappa>1.363, then β2γ>0\beta_2\gamma>0, the NLSE is denoted as focusing, and MI, bright solitons, and breathers are admitted. If κ<1.363\kappa<1.363, then β2γ<0\beta_2\gamma<0, the FDNLSE is denoted as defocusing and admits families of dark solitons, i.e., black and gray, but MI is suppressed [22, 37, 53, 54]. We emphasize that, contrary to optical Kerr media, only the sign of the nonlinearity coefficient γ\gamma changes in the transition from the focusing to the defocusing regime (and vice versa) at κ=1.363\kappa=1.363. We refer to [55] for details on dimensional transformations of the solutions of the dimensionless NLSE in order to apply these to the water wave problem, particularly, the AB and PB solutions of interest in this study. A description of the boundary conditions to be generated and applied to the wave maker can also be found in the latter reference.

Linear Schrödinger framework

The purpose of this analysis is to demonstrate that the essential features of the observed patterns can be explained by assuming that the dynamics beyond the focusing point are predominantly governed by linear dispersive effects. Working for the time being in dimensionless units, we assume that the apex or focus point of both the Peregrine breather (PB) and the Akhmediev breather (AB) occurs at the distance x=x0x=x_0 and that the bathymetry discontinuously jumps at x=xj≠x0x=x_j\ne x_0. For x>xjx>x_j the nonlinearity is assumed to be vanishing, while the dispersion value appears only when the comparison to simulation results in dimensional units. We conjecture that a good approximation is to assume that the propagation beyond this point is dominated by linear dispersive effects. In other words, we assume that the envelope evolves according to the linear Schrödinger equation

i∂ψ∂x=βˉ22∂2ψ∂t2(2)i\frac{\partial\psi}{\partial x}=\frac{\bar{\beta}_2}{2}\frac{\partial^2\psi}{\partial t^2} \tag*{(2)}

subject to the initial condition ψ0=ψ(t,xj)\psi_0=\psi(t,x_j) at x=xjx=x_j that may or may not correspond to the apex of the PB or the AB. The variables are dimensionless, i.e., ψ→aψ\psi\to a\psi, so that the breather background before the bathymetry change ψ0=1\psi_0=1, t→T0tt\to T_0t and x→X0xx\to X_0x, where X0=[γ(1)a2]−1X_0=[\gamma^{(1)}a^2]^{-1} and T0≡[X0β2(1)]12T_0\equiv[X_0\beta_2^{(1)}]^{\frac{1}{2}}, the superscripts refer to the value of parameters before the bathymetry change and aa is the initial carrier amplitude, like above. It finally turns out that βˉ2=β2(2)/β2(1)\bar{\beta}_2=\beta_2^{(2)}/\beta_2^{(1)}.

Introducing the shifted coordinate x^=x−xj\hat{x}=x-x_j and the Fourier transform and inverse transform

ψ~(ω,x^)=∫−∞+∞ψ(t,x^)exp⁡(iωt) dt;ψ(t,x^)=12π∫−∞+∞ψ~(ω,x^)exp⁡(−iωt) dω,(3)\begin{aligned} \tilde{\psi}(\omega,\hat{x})&=\int_{-\infty}^{+\infty}\psi(t,\hat{x})\exp(i\omega t)\,dt;\\ \psi(t,\hat{x})&=\frac{1}{2\pi}\int_{-\infty}^{+\infty}\tilde{\psi}(\omega,\hat{x})\exp(-i\omega t)\,d\omega, \tag*{(3)} \end{aligned}

the solution of Eq. (2) can be immediately expressed in Fourier space as

ψ~(ω,x^)=ψ~0(ω,x^=0)exp⁡(iβˉ2ω22x^),(4)\tilde{\psi}(\omega,\hat{x})=\tilde{\psi}_0(\omega,\hat{x}=0)\exp\left(i\frac{\bar{\beta}_2\omega^2}{2}\hat{x}\right), \tag*{(4)}

which then yields the solution for x^≥0\hat{x}\geq0 as

ψ(t,x^)=12π∫−∞+∞ψ~0(ω,0)eiβˉ2ω22x^e−iωt dω.(5)\psi(t,\hat{x})=\frac{1}{2\pi}\int_{-\infty}^{+\infty}\tilde{\psi}_0(\omega,0)e^{i\frac{\bar{\beta}_2\omega^2}{2}\hat{x}}e^{-i\omega t}\,d\omega. \tag*{(5)}

Equation (5) has the characteristic form of a diffraction integral in the Fresnel approximation.

Linear Evolution of Akhmediev Breather

As far as the AB is concerned, a slightly different approach is required. Neglecting the global phase factor, we have at the jump position x=xjx=x_j the following periodic function with temporal period 2π/ω02\pi/\omega_0, where ω0=21−2a\omega_0=2\sqrt{1-2a}, along with its formal Fourier expansion

ψAB(t,x^=0)=1+2(1−2a)cosh⁡(bx^)+ibsinh⁡(bx^)2acos⁡(ω0t)−cosh⁡(bx^)=∑n=−∞+∞u~neinω0t.(6)\begin{aligned} \psi_{AB}(t,\hat{x}=0)&=1+\frac{2(1-2a)\cosh(b\hat{x})+ib\sinh(b\hat{x})}{\sqrt{2a}\cos(\omega_0t)-\cosh(b\hat{x})}\\ &=\sum_{n=-\infty}^{+\infty}\tilde{u}_n e^{in\omega_0t}. \tag*{(6)} \end{aligned}

with x~=xj−x0\tilde{x} = x_j - x_0 and b=8a(1−2a)b = \sqrt{8a(1 - 2a)}. The Fourier coefficients u~n\tilde{u}_n are obtained by conventional techniques and read

u~0=1−C(x~)g(x~);(7)\tilde{u}_0 = 1 - \frac{C(\tilde{x})}{g(\tilde{x})}; \tag*{(7)}
u~n=−C(x~)g(x~)ρ(x~)∣n∣,(8)\tilde{u}_n = -\frac{C(\tilde{x})}{g(\tilde{x})}\rho(\tilde{x})^{|n|}, \tag*{(8)}

with C(x)≡2(1−2a)cosh⁡bx+ibsinh⁡bxC(x) \equiv2(1 - 2a)\cosh bx + ib\sinh bx, g(x)=cosh⁡2bx−2ag(x) = \sqrt{\cosh^2 bx - 2a} and ρ(x)≡cosh⁡bx−g2a\rho(x) \equiv\frac{\cosh bx-g}{\sqrt{2a}}. The term n=0n = 0 corresponds to the carrier wave, and n≠0n \ne0 for the symmetric (±n\pm n) sideband pairs. They reduce to those reported in Ref. [56] for x~=0\tilde{x} = 0. In this case, the evolution of the field is such that each discrete frequency pair ±nω0\pm n\omega_0 undergoes a phase rotation given by the term exp⁡[iβˉ22(nω0)2x^]\exp\left[i\frac{\bar{\beta}_2}{2}(n\omega_0)^2\hat{x}\right] induced by dispersion. Therefore, we obtain the evolving field from the following formula, where we exploited the parity of the coefficients of order ±n\pm n

ψ(t,x^)=u~0+2∑n=1∞u~ncos⁡(nω0t)eiβˉ2n2ω022x^(9)\psi(t,\hat{x}) = \tilde{u}_0 + 2\sum_{n=1}^{\infty}\tilde{u}_n\cos(n\omega_0t)e^{i\frac{\bar{\beta}_2n^2\omega_0^2}{2}\hat{x}} \tag*{(9)}

The result is that, in the linear limit, the AB gives a carpet-like evolution as shown in Figure 5(a,Figure 5c), where the initial pattern is recovered at distances x^m=m2π/(βω02)\hat{x}_m = m2\pi/(\beta\omega_0^2), with even m=0,2,…m = 0, 2, \ldots, whereas π\pi-shifted patterns are generated for odd m=1,3,…m = 1, 3, \ldots.

Linear Evolution of Peregrine Breather

We study the PB first, because it is a direct application of Eq. (5). Its envelope at the jump position reads

ψPB(t,x^=0)=1−41−2ix~1+4t2+4x~2,(10)\psi_{PB}(t,\hat{x}=0) = 1 - 4\frac{1 - 2i\tilde{x}}{1 + 4t^2 + 4\tilde{x}^2}, \tag*{(10)}

with x~=xj−x0\tilde{x} = x_j - x_0. The Fourier transform associated to Eq. (10) reads

ψ~0=ψ~(ω,0)=2πδ(ω)−2πeiϕAexp⁡(−∣Aω∣/2),(11)\tilde{\psi}_0 = \tilde{\psi}(\omega,0) = 2\pi\delta(\omega) - 2\pi e^{i\phi_A}\exp(-|A\omega|/2), \tag*{(11)}

with A≡1−2ix~=∣A∣eiϕAA \equiv1 - 2i\tilde{x} = |A|e^{i\phi_A}. We substitute ψ~(ω,0)\tilde{\psi}(\omega,0) from Eq. (11) into Eq. (5) to obtain

ψ(t,x^)=1−eiϕA∫−∞+∞χ(ω)eiϕ(ω)dω(12)\psi(t,\hat{x}) = 1 - e^{i\phi_A}\int_{-\infty}^{+\infty}\chi(\omega)e^{i\phi(\omega)}d\omega \tag*{(12)}

where χ(ω)=e−∣Aω∣2\chi(\omega) = e^{-\frac{|A\omega|}{2}} is a rapidly decaying real function, and ϕ(ω)=(βˉ2ω22x^−ωt)\phi(\omega) = \left(\frac{\bar{\beta}_2\omega^2}{2}\hat{x} - \omega t\right) is the phase that rules the fast oscillating part of the integral. In principle, the integral in Eq. (12) could be expressed in terms of the Faddeeva function. A more transparent result, in the same spirit used for the semiclassical approximation of the Schrödinger equation, is obtained by applying the stationary phase approximation

[57]. In this framework, the dominant contribution is given by the point where the phase is stationary ϕ′(ω)=0\phi'(\omega)=0, i.e., the frequency ω=1βˉ2x^\omega=\frac{1}{\bar{\beta}_2\hat{x}}, and the final result reads

ψ(t,x^)=1−2πβˉ2x^exp⁡(iπ4sgn⁡(βˉ2x^)+iϕA)×exp⁡(−∣At2βˉ2x^∣)exp⁡(−it22βˉ2x^),for x^→∞(13)\psi(t,\hat{x})=1-\sqrt{\frac{2\pi}{\bar{\beta}_2\hat{x}}}\exp\left(\frac{i\pi}{4}\operatorname{sgn}(\bar{\beta}_2\hat{x})+i\phi_A\right) \times\exp\left(-\left|\frac{At}{2\bar{\beta}_2\hat{x}}\right|\right)\exp\left(-i\frac{t^2}{2\bar{\beta}_2\hat{x}}\right),\quad\text{for }\hat{x}\to\infty \tag*{(13)}

The pattern predicted by Eq. (13), shown in Figure 3 and Figure 5, closely resembles the observed pattern when starting from a PB.

Acknowledgements

A.A. acknowledges the financial support of the University of Ferrara (Grant FIRD 2024). A.A. and S.T. acknowledge the financial support of Italian MUR (Ministero dell’Università e della Ricerca) under grants 2020X4T57A and 2022NCTCY-NextGenerationEU, with the European Union. J.Z. acknowledges the financial support of Fundamental Research Funds for the Central Universities (Grant No. 79000039). H.C.H. acknowledges the financial support of National Science and Engineering Council of Taiwan, Republic of China (MOST 110-2221-E-110-016-MY3). L.K. and M.B. acknowledge support from the French-Taiwanese ORCHID Program of the Hubert Curien Partnership Program (PHC). A.C. acknowledges support from Okinawa Institute of Science and Technology (OIST) with subsidy funding from the Cabinet Office, Government of Japan.

Data availability

All data supporting the findings of this study are available from the corresponding author upon reasonable request.

Code availability

The codes for the analysis of this study are available from the corresponding author upon reasonable request.

Author contributions

A.C. conceived the study. H.-C.H, M.-S.L., L.K, J.Z., and A.C. performed the experiments. A.A., J.Z., and A.C. carried out the numerical simulations. A.A. and S.T. developed the Fresnel approximation for the breather evolution. All authors analyzed the data, interpreted the results, and contributed to the writing and revision of the manuscript.

References

  1. [1]Carrier, G.F., Greenspan, H.P.: Water waves of finite amplitude on a sloping beach. Journal of Fluid Mechanics 4(1), 97–109 (1958) https://doi.org/10.1017/S0022112058000331
  2. [2]Li, Y., Raichlen, F.: Non-breaking and breaking solitary wave run-up. Journal of Fluid Mechanics 456, 295–318 (2002) https://doi.org/10.1017/S0022112001007625
  3. [3]Sriram, V., Didenkulova, I., Schimmels, S., Sergeeva, A., Goseberg, N.: Long Wave Propagation, shoaling and run-up in nearshore areas. Coastal Engineering Proceedings 1(34), 20 (2014) https://doi.org/10.9753/icce.v34.currents.20
  4. [4]Li, Y., Chabchoub, A.: On the formation of coastal rogue waves in water of variable depth. Cambridge Prisms: Coastal Futures 1, 33 (2023) https://doi.org/10.1017/cft.2023.21
  5. [5]Trulsen, K., Zeng, H., Gramstad, O.: Laboratory evidence of freak waves provoked by non-uniform bathymetry. Physics of Fluids 24(9), 097101 (2012) https://doi.org/10.1063/1.4748346
  6. [6]Zhang, J., Benoit, M., Kimmoun, O., Chabchoub, A., Hsu, H.-C.: Statistics of extreme waves in coastal waters: large scale experiments and advanced numerical simulations. Fluids 4(2), 99 (2019) https://doi.org/10.3390/fluids4020099
  7. [7]Li, Y., Draycott, S., Zheng, Y., Lin, Z., Adcock, T.A., Bremer, T.S.: Why rogue waves occur atop abrupt depth transitions. Journal of Fluid Mechanics 919, 5 (2021) https://doi.org/10.1017/jfm.2021.409
  8. [8]Lawrence, C., Trulsen, K., Gramstad, O.: Statistical properties of wave kinematics in long-crested irregular waves propagating over non-uniform bathymetry. Physics of Fluids 33(4), 046601 (2021) https://doi.org/10.1063/5.0047643
  9. [9]Karmpadakis, I., Swan, C.: A new crest height distribution for nonlinear and breaking waves in varying water depths. Ocean Engineering 266, 112972 (2022) https://doi.org/10.1016/j.oceaneng.2022.112972
  10. [10]Mendes, S., He, Y., Kasparian, J., Chabchoub, A.: Energetics and stochastics of extreme waves breaking over a symmetrical breakwater. Physics of Fluids 38(6) (2026) https://doi.org/10.1063/5.0317620
  11. [11]Anastassiou, C., Shih, M.-f., Mitchell, M., Chen, Z., Segev, M.: Optically induced photovoltaic self-defocusing-to-self-focusing transition. Optics Letters 23(12), 924–926 (1998) https://doi.org/10.1364/OL.23.000924
  12. [12]Liang, G., Hong, W., Luo, T., Wang, J., Li, Y., Guo, Q., Hu, W., Christodoulides, D.N.: Transition between self-focusing and self-defocusing in a nonlocally non-linear system. Physical Review A 99(6), 063808 (2019) https://doi.org/10.1103/PhysRevA.99.063808
  13. [13]Chernikov, S.V., Mamyshev, P.V.: Femtosecond soliton propagation in fibers with slowly decreasing dispersion. Journal of the Optical Society of America B 8(8), 1633 (1991) https://doi.org/10.1364/josab.8.001633
  14. [14]Chernikov, S.V., Richardson, D.J., Dianov, E.M., Payne, D.N.: Picosecond soliton pulse compressor based on dispersion decreasing fibre. Electronics Letters 28(19), 1842–1844 (1992) https://doi.org/10.1049/el:19921175
  15. [15]Abdullaev, F., Garnier, J.: Optical solitons in random media. Progress in Optics 48(December 2017), 35–106 (2005) https://doi.org/10.1016/S0079-6638(05)48002-6
  16. [16]Armaroli, A., Gomel, A., Chabchoub, A., Brunetti, M., Kasparian, J.: Stabilization of uni-directional water wave trains over an uneven bottom. Nonlinear Dynamics 101(2), 1131–1145 (2020) https://doi.org/10.1007/s11071-020-05819-9 arXiv:2002.12923
  17. [17]Gomel, A., Chabchoub, A., Brunetti, M., Trillo, S., Kasparian, J., Armaroli, A.: Stabilization of unsteady nonlinear waves by phase-space manipulation. Physical Review Letters 126(17), 174501 (2021) https://doi.org/10.1103/PhysRevLett.126.174501
  18. [18]Onorato, M., ResidorI, S., Bortolozzo, U., Montina, A., Areсchi, F.: Rogue waves and their generating mechanisms in different physical contexts. Physics Reports 528(2), 47–89 (2013) https://doi.org/10.1016/j.physrep.2013.03.001
  19. [19]Dudley, J.M., Dias, F., Erkintalo, M., Genty, G.: Instabilities, breathers and rogue waves in optics. Nature Photonics 8(10), 755–764 (2014) https://doi.org/10.1038/nphoton.2014.220
  20. [20]Tlidi, M., Taki, M.: Rogue waves in nonlinear optics. Advances in Optics and Photonics 14(1), 87–147 (2022) https://doi.org/10.1364/AOP.438025
  21. [21]Zakharov, V.E.: Stability of periodic waves of finite amplitude on the surface of a deep fluid. Journal of Applied Mechanics and Technical Physics 9(2), 190–194 (1968) https://doi.org/10.1007/bf00913182
  22. [22]Shabat, A., Zakharov, V.: Exact theory of two-dimensional self-focusing and one-dimensional self-modulation of waves in nonlinear media. Sov. Phys. JETP 34(1), 62 (1972)
  23. [23]Akhmediev, N., Eleonskii, V., Kulagin, N.: Generation of periodic trains of picosecond pulses in an optical fiber: exact solutions. Sov. Phys. JETP 62(5), 894–899 (1985)
  24. [24]Peregrine, D.H.: Water waves, nonlinear Schrödinger equations and their solutions. The ANZIAM Journal 25(1), 16–43 (1983) https://doi.org/10.1017/s0334270000003891
  25. [25]Kibler, B., Fatome, J., Finot, C., Millot, G., Dias, F., Genty, G., Akhmediev, N., Dudley, J.M.: The Peregrine soliton in nonlinear fibre optics. Nature Physics 6(10), 790–795 (2010) https://doi.org/10.1038/nphys1740
  26. [26]Chabchoub, A., Hoffmann, N., Akhmediev, N.: Rogue wave observation in a water wave tank. Physical Review Letters 106(20), 204502 (2011) https://doi.org/10.1103/PhysRevLett.106.204502
  27. [27]Bailung, H., Sharma, S., Nakamura, Y.: Observation of Peregrine solitons in a multicomponent plasma with negative ions. Physical Review Letters 107(25), 255005 (2011) https://doi.org/10.1103/PhysRevLett.107.255005
  28. [28]Romero-Ros, A., Katsimiga, G.C., Mistakidis, S.I., Mossman, S., Biondini, G., Schmelcher, P., Engels, P., Kevrekidis, P.: Experimental realization of the peregrine soliton in repulsive two-component bose-einstein condensates. Physical Review Letters 132(3), 033402 (2024) https://doi.org/10.1103/PhysRevLett.132.033402
  29. [29]Mussot, A., Naveau, C., Conforti, M., Kudlinski, A., Copie, F., Szriftgiser, P., Trillo, S.: Fibre multi-wave mixing combs reveal the broken symmetry of Fermi–Pasta–Ulam recurrence. Nature Photonics 12(5), 303–308 (2018) https://doi.org/10.1038/s41566-018-0136-1
  30. [30]Pierangeli, D., Flammini, M., Zhang, L., Marcucci, G., Agranat, A.J., Grinevich, P.G., Santini, P.M., Conti, C., DelRe, E.: Observation of fermi-pasta-ulam-tsingou recurrence and its exact dynamics. Physical Review X 8, 041017 (2018) https://doi.org/10.1103/PhysRevX.8.041017
  31. [31]Luo, D., Jin, Y., Nguyen, J.H., Malomed, B.A., Marchukov, O.V., Yurovsky, V.A., Dunjko, V., Olshanii, M., Hulet, R.: Creation and characterization of matter-wave breathers. Physical Review Letters 125(18), 183902 (2020) https://doi.org/10.1103/PhysRevLett.125.183902
  32. [32]Suret, P., Tikan, A., Bonnefoy, F., Copie, F., Ducrozet, G., Gelash, A., Prabhudesai, G., Michel, G., Cazaubiel, A., Falcon, E., et al.: Nonlinear spectral synthesis of soliton gas in deep-water surface gravity waves. Physical Review Letters 125(26), 264101 (2020) https://doi.org/10.1103/PhysRevLett.125.264101
  33. [33]Suret, P., Randoux, S., Gelash, A., Agafontsev, D., Doyon, B., El G.: Soliton gas: Theory, numerics, and experiments. Physical Review E 109(6), 061001 (2024) https://doi.org/10.1103/PhysRevE.109.061001
  34. [34]Mei, C.C., Stiassnie, M.A., Yue, D.K.-P.: Theory and Applications of Ocean Surface Waves. World Scientific, Singapore (2005)
  35. [35]Chabchoub, A., Kibler, B., Finot, C., Millot, G., Onorato, M., Dudley, J.M., Babanin, A.: The nonlinear Schrödinger equation and the propagation of weakly nonlinear waves in optical fibers and on the water surface. Annals of Physics 361, 490–500 (2015) https://doi.org/10.1016/j.aop.2015.07.003
  36. [36]Carretero-González, R., Frantzeskakis, D.J., Kevrekidis, P.G.: Nonlinear Waves & Hamiltonian Systems: from One to Many Degrees of Freedom, from Discrete to Continuum. Oxford University Press, ??? (2024)
  37. [37]Osborne, A.R.: Nonlinear Ocean Waves and the Inverse Scattering Transform. Elsevier, Amsterdam (2010)
  38. [38]Kimmoun, O., Hsu, H., Branger, H., Li, M., Chen, Y., Kharif, C., Onorato, M., Kelleher, E.J., Kibler, B., Akhmediev, N., et al.: Modulation instability and phase-shifted Fermi-Pasta-Ulam recurrence. Scientific Reports 6(1), 28516 (2016) https://doi.org/10.1038/srep28516
  39. [39]Erkintalo, M., Hammani, K., Kibler, B., Finot, C., Akhmediev, N., Dudley, J.M., Genty, G.: Higher-order modulation instability in nonlinear fiber optics. Physical Review Letters 107(25), 253901 (2011) https://doi.org/10.1103/physrevlett.107.253901
  40. [40]Kimmoun, O., Hsu, H., Kibler, B., Chabchoub, A.: Nonconservative higher-order hydrodynamic modulation instability. Physical Review E 96(2), 022219 (2017) https://doi.org/10.1103/physrevlett.107.253901
  41. [41]Berry, M.V., Klein, S.: Integer, fractional and fractal Talbot effects. Journal of Modern Optics 43(10), 2139–2164 (1996) https://doi.org/10.1080/095003496154761
  42. [42]Hsu, H., Abid, M., Chen, Y.-Y., Kharif, C.: Transformation of the Peregrine breather into gray solitons on a vertically sheared current. Frontiers in Physics 9, 631993 (2021) https://doi.org/10.3389/fphy.2021.631993
  43. [43]Zabusky, N.J., Kruskal, M.D.: Interaction of ”solitons” in a collisionless plasma and the recurrence of initial states. Physical Review Letters 15(6), 240 (1965) https://doi.org/10.1103/PhysRevLett.15.240
  44. [44]Fatome, J., Finot, C., Millot, G., Armaroli, A., Trillo, S.: Observation of optical undular bores in multiple four-wave mixing. Physical Review X 4(2), 021022 (2014) https://doi.org/10.1103/PhysRevX.4.021022
  45. [45]Trillo, S., Deng, G., Biondini, G., Klein, M., Clauss, G., Chabchoub, A., Onorato, M.: Experimental observation and theoretical description of multisoliton fission in shallow water. Physical Review Letters 117(14), 144102 (2016) https://doi.org/10.1103/PhysRevLett.117.144102
  46. [46]El, G., Hoefer, M.: Dispersive shock waves and modulation theory. Physica D: Nonlinear Phenomena 333, 11–65 (2016) https://doi.org/10.1016/j.physd.2016.04.006
  47. [47]Vanderhaegen, G., Naveau, C., Szirki̇tser, P., Kudlinski, A., Conforti, M., Mussot, A., Onorato, M., Trillo, S., Chabchoub, A., Akhmediev, N.N.: ”Extraordinary” modulation instability in optics and hydrodynamics. Proceedings of the National Academy of Sciences 118, 2019348118 (2021) https://doi.org/10.1073/pnas.2019348118
  48. [48]Bakman, A., Fishman, S., Fink, M., Fort, E., Wildeman, S.: Observation of the Talbot effect with water waves. American Journal of Physics 87(1), 38–43 (2019) https://doi.org/10.1119/1.5081051
  49. [49]Schiek, R.: Excitation of nonlinear beams: from the linear Talbot effect through modulation instability to Akhmediev breathers. Optics Express 29(10), 15830–15851 (2021) https://doi.org/10.1364/OE.425626
  50. [50]Rozenman, G.G., Schleich, W.P., Shemer, L., Arie, A.: Periodic wave trains in nonlinear media: Talbot revivals, Akhmediev breathers, and asymmetry breaking. Physical Review Letters 128(21), 214101 (2022) https://doi.org/10.1103/PhysRevLett.128.214101
  51. [51]Hasimoto, H., Ono, H.: Nonlinear modulation of gravity waves. Journal of the Physical Society of Japan 33(3), 805–811 (1972) https://doi.org/10.1143/JPSJ.33.805
  52. [52]Djordjević, V.D., Redekopp, L.G.: On the development of packets of surface gravity waves moving over an uneven bottom. Zeitschrift für angewandte Mathematik und Physik ZAMP 29(6), 950–962 (1978) https://doi.org/10.1007/BF01590816
  53. [53]Akhmediev, N.N., Ankiewicz, A.: Solitons: Nonlinear Pulses and Beams. Springer, London (1997)
  54. [54]Kharif, C., Pelinovsky, E., Slunyaev, A.: Rogue Waves in the Ocean. Springer, Berlin (2008)
  55. [55]Chabchoub, A., Grimshaw, R.H.: The hydrodynamic nonlinear Schrödinger equation: Space and time. Fluids 1(3), 23 (2016) https://doi.org/10.3390/fluids1030023
  56. [56]Bendahmane, A., Mussot, A., Kudlinski, A., Szriftgiser, P., Conforti, M., Wabnitz, S., Trillo, S.: Optimal frequency conversion in the nonlinear stage of modulation instability. Optics Express 23(24), 30861 (2015) https://doi.org/10.1364/oe.23.030861
  57. [57]Miller, P.D.: Applied Asymptotic Analysis. Graduate Studies in Mathematics. American Mathematical Society, Providence, RI (2006)DOI

Paper details

vlnc:W_7k45g7cz
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 12

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents