Nonlinear Talbot Effect: Recurrence in Nonlinear Systems

• Nonlinear Talbot effect is a self-imaging phenomenon where periodic wavefields in nonlinear media recur through intrinsic eigenmode dynamics or engineered nonlinear sources.
• In Kerr media, rogue-wave eigenmodes lead to full image recurrences at the Talbot length and π-phase shifted images at half the Talbot length, suppressing fractional patterns.
• The effect extends to temporal soliton crystals and dispersive revival in periodic-boundary PDEs, offering broad applications in imaging, optical switching, and photonic structuring.

Searching arXiv for relevant papers on the nonlinear Talbot effect and closely related variants. Search query: "nonlinear Talbot effect rogue waves Kerr medium temporal Talbot soliton crystals arXiv" The nonlinear Talbot effect denotes a family of self-imaging and revival phenomena in periodic wavefields when either the propagation law is nonlinear or the periodic field itself is generated or modulated by a nonlinear or coherently prepared medium. In one established optical sense, it is the recurrence of periodic rogue-wave or breather patterns in a cubic Kerr medium governed by the nonlinear Schrödinger equation (NLSE), with revivals at a nonlinear Talbot length and at its half-length with a $\pi$ phase shift, but without the fractional images characteristic of classical Fresnel Talbot carpets (Zhang et al., 2014). In broader usage, the term also covers second-harmonic Talbot self-imaging in $\chi^{(2)}$ media, electromagnetically induced Talbot imaging in atomic lattices, temporal Talbot recurrences in nonlinear fibers, and periodic-boundary dispersive recurrences for nonlinear Schrödinger-type equations on the torus (Liu et al., 2013, Zhang et al., 2016, Zajnulina et al., 2024, Erdogan et al., 2013).

1. Definition and scope of the term

The surveyed literature uses the expression “nonlinear Talbot effect” in more than one sense. The narrowest definition is the intrinsic self-imaging of a periodic nonlinear eigenmode during propagation in a nonlinear medium. This is the meaning emphasized for periodic rogue-wave breathers in a bulk Kerr medium: the field is itself an eigenmode of the NLSE, and the recurrence is not reducible to linear Fresnel diffraction of a passive grating (Zhang et al., 2014). The same intrinsic meaning appears in the two-dimensional Kerr extension, where a periodic transverse pattern on a finite background propagates in a bulk 3D nonlinear medium and recurs only at the Talbot length and half-Talbot length (Zhang et al., 2014).

A second usage attaches “nonlinear” to the source or grating rather than to the propagation operator. In second-harmonic Talbot imaging, the periodic structure is the nonlinear polarization $$P^{(2)}(2\omega,\mathbf{r})=\epsilon_0\chi^{(2)}(\mathbf{r})E^2(\omega,\mathbf{r})$$, and the self-imaged field is the generated second harmonic (Liu et al., 2013). In electromagnetically induced Talbot imaging, the grating is an optically induced atomic lattice created by coherent preparation of an EIT medium, with both linear and Kerr contributions to the probe susceptibility and refractive-index modulation (Zhang et al., 2016).

A third usage belongs to dispersive PDE theory on periodic domains. There, the Talbot effect refers to the rational-time revival and irrational-time fractalization of solutions of periodic Schrödinger-type equations. In the cubic NLS and in the Manakov system on the torus, the nonlinear Talbot effect describes persistence of the rational-versus-irrational dichotomy into the nonlinear regime, rather than near-field diffraction behind a physical grating (Erdogan et al., 2013, Yin et al., 2023).

Across these meanings, the linear benchmark remains the classical Talbot relation $z_T = 2d^2/\lambda$ for a periodic structure of period $d$ illuminated at wavelength $\lambda$. What changes is the mechanism enforcing recurrence: quadratic phase rephasing of independent harmonics in the linear case, nonlinear eigenmode periodicity or nonlinear mode coupling in the intrinsic nonlinear case, and nonlinear source engineering or coherent atomic susceptibility in the broader usage.

2. Bulk Kerr self-imaging from rogue-wave eigenmodes

The canonical intrinsic nonlinear Talbot effect is formulated in the cubic 1D NLSE,

$$i\frac{\partial \psi}{\partial z}+\frac{1}{2}\frac{\partial^2\psi}{\partial x^2}+|\psi|^2\psi=0,$$

with $x$ the transverse coordinate and $z$ the propagation coordinate (Zhang et al., 2014). The relevant exact solutions are rogue-wave-type breathers on a finite background. The Akhmediev breather is periodic in $x$ and localized in $\chi^{(2)}$0; the Kuznetsov–Ma breather is periodic in $\chi^{(2)}$1 and localized in $\chi^{(2)}$2; and a Jacobi-elliptic Akhmediev solution is periodic in both $\chi^{(2)}$3 and $\chi^{(2)}$4, making it the central exact eigenmode for nonlinear Talbot recurrence (Zhang et al., 2014).

For the doubly periodic Akhmediev mode, the nonlinear Talbot length is the longitudinal period of the solution,

$\chi^{(2)}$5

with $\chi^{(2)}$6 the complete elliptic integral of the first kind. At $\chi^{(2)}$7 the full transverse pattern, including phase, exactly recurs; at $\chi^{(2)}$8 the intensity recurs with a global $\chi^{(2)}$9 phase shift, $$P^{(2)}(2\omega,\mathbf{r})=\epsilon_0\chi^{(2)}(\mathbf{r})E^2(\omega,\mathbf{r})$$0; and fractional Talbot images are absent (Zhang et al., 2014). In the same family, both the transverse period and the longitudinal period decrease monotonically with increasing $$P^{(2)}(2\omega,\mathbf{r})=\epsilon_0\chi^{(2)}(\mathbf{r})E^2(\omega,\mathbf{r})$$1, they cross at $$P^{(2)}(2\omega,\mathbf{r})=\epsilon_0\chi^{(2)}(\mathbf{r})E^2(\omega,\mathbf{r})$$2, and as $$P^{(2)}(2\omega,\mathbf{r})=\epsilon_0\chi^{(2)}(\mathbf{r})E^2(\omega,\mathbf{r})$$3 the recurrence length diverges while the solution approaches the single-parameter Akhmediev breather limit.

The same paper also launched a transversely periodic Akhmediev breather profile at $$P^{(2)}(2\omega,\mathbf{r})=\epsilon_0\chi^{(2)}(\mathbf{r})E^2(\omega,\mathbf{r})$$4 as a near-eigenmode initial condition. Numerically, the input repeated at a certain propagation length $$P^{(2)}(2\omega,\mathbf{r})=\epsilon_0\chi^{(2)}(\mathbf{r})E^2(\omega,\mathbf{r})$$5 and at $$P^{(2)}(2\omega,\mathbf{r})=\epsilon_0\chi^{(2)}(\mathbf{r})E^2(\omega,\mathbf{r})$$6 with a $$P^{(2)}(2\omega,\mathbf{r})=\epsilon_0\chi^{(2)}(\mathbf{r})E^2(\omega,\mathbf{r})$$7 phase shift, again without fractional images. The proposed mechanism was nonlinear interference among breather lobes. The authors summarized this phenomenologically by stating that the interaction is related to the transverse period and intensity of breathers, “in that the bigger the period and the higher the intensity, the shorter the TE length”; however, their Akhmediev-breather parameter scan showed a non-monotonic $$P^{(2)}(2\omega,\mathbf{r})=\epsilon_0\chi^{(2)}(\mathbf{r})E^2(\omega,\mathbf{r})$$8 with a minimum at intermediate $$P^{(2)}(2\omega,\mathbf{r})=\epsilon_0\chi^{(2)}(\mathbf{r})E^2(\omega,\mathbf{r})$$9 as $z_T = 2d^2/\lambda$0, because the transverse period diverges while the intensity saturates in the Peregrine limit (Zhang et al., 2014).

A central distinction from the classical Talbot effect follows directly from the loss of linear superposition. In the linear Talbot carpet, rational fractions of $z_T = 2d^2/\lambda$1 are produced by independent quadratic phase accumulation of many spatial harmonics. In the nonlinear Kerr case, the field evolves as a coupled nonlinear structure, so only the recurrences encoded by the intrinsic periodicity of the eigenmode survive: the primary image and the phase-inverted half-Talbot image (Zhang et al., 2014).

3. Two-dimensional and temporal nonlinear Talbot dynamics

The two-dimensional Kerr generalization uses the normalized 2D NLSE

$z_T = 2d^2/\lambda$2

for a 2D periodic transverse input propagating in a bulk 3D medium (Zhang et al., 2014). Because analytic 2D rogue-wave eigenmodes are unavailable, one construction multiplies two orthogonal 1D Akhmediev-breather profiles at $z_T = 2d^2/\lambda$3. The resulting nonlinear Talbot effect can be visualized as a 3D stack of Talbot carpets. As in the 1D Kerr case, the pattern recurs at $z_T = 2d^2/\lambda$4 and at $z_T = 2d^2/\lambda$5 with a $z_T = 2d^2/\lambda$6 phase shift, and no other fractional recurrences are observed. The same work argued that rogue-wave initial conditions are sufficient but not necessary: other periodic inputs on a finite background can also produce the effect, whereas purely periodic patterns without background do not (Zhang et al., 2014).

That 2D study also identified control parameters distinct from the 1D breather analysis. The smaller the period of the incident rogue waves, the shorter the Talbot length. Increasing the beam intensity increases the Talbot length, but above a threshold this leads to catastrophic self-focusing and destroys the effect. The authors further interpreted the recurrence as a self-Fourier transform of the initial periodic beam: the linear Talbot effect corresponds to a fractional self-Fourier transform, whereas the nonlinear Talbot effect corresponds to a regular self-Fourier transform (Zhang et al., 2014).

Temporal nonlinear Talbot dynamics in optical fibers were analyzed for a phase-modulated continuous-wave input propagated in a passive single-mode silica fiber with $z_T = 2d^2/\lambda$7 and $z_T = 2d^2/\lambda$8 (Zajnulina et al., 2024). The 2024 study identified three input-power-dependent regimes: a quasi-linear Talbot carpet at low power, an intermediate regime hosting soliton crystals rather than rogue waves, and a higher-power regime of separated Talbot solitons. The numerically determined transitions were at $z_T = 2d^2/\lambda$9 and $d$0, and Talbot-soliton beating in the nonlinear regime was proposed for pulse repetition-rate multiplication (Zajnulina et al., 2024).

A later temporal analysis developed a more detailed regime structure and interpreted the Talbot effect as the organizing skeleton of nonlinear comb dynamics (Zajnulina, 17 Aug 2025). In that framework, linear temporal Talbot beating evolves into A-type breathers, soliton crystals, and soliton gas as the hierarchy among self-phase modulation, cross-phase modulation, and four-wave mixing changes. Regular FWM among a few central comb lines supports A-type breathers; cascaded FWM among many lines supports soliton crystals; and SPM dominance breaks collective locking and produces a soliton gas. This study explicitly tied the temporal nonlinear Talbot effect to Fermi–Pasta–Ulam–Tsingou recurrence and to the beating frequencies extracted by Soliton Radiation Beat Analysis (Zajnulina, 17 Aug 2025).

4. Nonlinear-source and coherently induced Talbot effects

In second-harmonic Talbot imaging, the periodic object is not a passive mask but the nonlinear source generated in a $d$1 crystal (Liu et al., 2013). For a 2D periodically poled LiTaO$d$2 sample, the second-harmonic field at the output plane is modeled as $d$3, with $d$4 phase reversal between positive and negative domains. The primary SH Talbot length is

$d$5

and for $d$6 and $d$7 the calculated value $d$8 matched an observed primary Talbot plane at $d$9 (Liu et al., 2013). The same platform also produced periodic SH sub-diffracted spots over tens to more than one hundred micrometers from the sample. At $\lambda$0, the reported FWHM was $\lambda$1, i.e. less than $\lambda$2, without evanescent waves or sub-wavelength apertures. In this literature, “nonlinear Talbot effect” refers to the nonlinear generation of the source field, followed by Fresnel or Rayleigh–Sommerfeld propagation of the generated harmonic (Liu et al., 2013).

An atomic version appears in an EIT-prepared rubidium vapor cell (Zhang et al., 2016). Two coupling beams interfere to form a standing-wave atomic lattice of period

$\lambda$3

which modulates the probe susceptibility through both $\lambda$4 and $\lambda$5. The resulting refractive index is written as

$\lambda$6

so the grating is both linear and Kerr-modified (Zhang et al., 2016). The observed Talbot length was about $\lambda$7, and fractional Talbot images with doubled periodicity were reported around $\lambda$8 and $\lambda$9. Because the Kerr coefficient can change sign with detuning, the induced grating can shift by half a period, and the work argued that such EIT Talbot imaging may pave a way for lensless and nondestructive imaging of ultracold atoms and molecules (Zhang et al., 2016).

These source-driven or coherently induced realizations broaden the term beyond bulk Kerr eigenmode recurrence. They preserve the Talbot idea of longitudinal self-imaging, but the nonlinear ingredient resides in second-harmonic generation or in susceptibility engineering rather than in NLSE eigenmode propagation.

5. Periodic-boundary formulations: cubic NLS and the Manakov system

In periodic-boundary dispersive PDEs, the nonlinear Talbot effect is formulated as a rational-time revival and irrational-time fractalization phenomenon. For the cubic NLS on the torus,

$$i\frac{\partial \psi}{\partial z}+\frac{1}{2}\frac{\partial^2\psi}{\partial x^2}+|\psi|^2\psi=0,$$0

with initial data of bounded variation, irrational times produce spatially continuous solutions, rational times produce bounded solutions with at most countably many discontinuities, and for data not belonging to $$i\frac{\partial \psi}{\partial z}+\frac{1}{2}\frac{\partial^2\psi}{\partial x^2}+|\psi|^2\psi=0,$$1, either the real or imaginary part of the graph has upper Minkowski dimension $$i\frac{\partial \psi}{\partial z}+\frac{1}{2}\frac{\partial^2\psi}{\partial x^2}+|\psi|^2\psi=0,$$2 for almost all times (Erdogan et al., 2013). The analysis separates the evolution into a linear Schrödinger part and a smoother nonlinear remainder, thereby showing that the Talbot-type linear dichotomy survives the cubic nonlinearity.

The coupled Manakov system on the torus,

$$i\frac{\partial \psi}{\partial z}+\frac{1}{2}\frac{\partial^2\psi}{\partial x^2}+|\psi|^2\psi=0,$$3

extends this picture to a genuinely multi-component nonlinear setting (Yin et al., 2023). For bounded-variation initial data, irrational times again yield continuous but nowhere differentiable fractal-like curves with Minkowski dimension $$i\frac{\partial \psi}{\partial z}+\frac{1}{2}\frac{\partial^2\psi}{\partial x^2}+|\psi|^2\psi=0,$$4. Numerical experiments supported persistence of Talbot-type rational-time behavior into the nonlinear regime: finite jump sets remain visible, but the profiles between jump discontinuities depend sensitively on the interplay of the two components. In particular, when the two components start from different initial profiles, the coupled evolution can induce subtle different qualitative profile between the jump discontinuities; and when one component starts smooth while the other is discontinuous, the smooth component can develop jump discontinuities through coupling (Yin et al., 2023).

This torus-based formulation is not a near-field diffraction problem, but it retains the core Talbot dichotomy. Rational times correspond to quantized revival structures, while irrational times correspond to dispersive fractalization. In this sense, the nonlinear Talbot effect names a structural property of nonlinear dispersive flow on a periodic domain rather than a specific optical platform.

6. Mechanisms, control parameters, applications, and interpretive issues

A recurring theme across the literature is that the nonlinear Talbot effect is controlled by more parameters than the classical Talbot length alone. In periodic, partially coherent random waves propagating in a Kerr medium, the effective nonlinear Talbot length was shown to depend strongly on source coherence even at fixed source periodicity (Bai et al., 23 Jun 2026). Reducing source coherence, and therefore speckle size, shortened the Talbot length and improved the quality of recurrent Talbot images. In that study, for $$i\frac{\partial \psi}{\partial z}+\frac{1}{2}\frac{\partial^2\psi}{\partial x^2}+|\psi|^2\psi=0,$$5 the primary revival had fidelity $$i\frac{\partial \psi}{\partial z}+\frac{1}{2}\frac{\partial^2\psi}{\partial x^2}+|\psi|^2\psi=0,$$6 and the sixth-order revival $$i\frac{\partial \psi}{\partial z}+\frac{1}{2}\frac{\partial^2\psi}{\partial x^2}+|\psi|^2\psi=0,$$7, whereas for $$i\frac{\partial \psi}{\partial z}+\frac{1}{2}\frac{\partial^2\psi}{\partial x^2}+|\psi|^2\psi=0,$$8 the corresponding values were $$i\frac{\partial \psi}{\partial z}+\frac{1}{2}\frac{\partial^2\psi}{\partial x^2}+|\psi|^2\psi=0,$$9 and $x$0. The proposed explanation was coherence-mediated nonlinear mode coupling: low coherence makes local dynamics more dispersion-dominated, while high coherence shifts the system toward nonlinear-eigenmode recurrence (Bai et al., 23 Jun 2026).

The mechanism behind recurrence depends on the platform. In bulk Kerr rogue-wave propagation, recurrence is attributed to nonlinear interference of breather lobes and to the intrinsic periodicity of nonlinear eigenmodes, which is why fractional images are suppressed (Zhang et al., 2014). In 2D Kerr propagation, the recurrence can be read as a regular self-Fourier transform and is limited by catastrophic self-focusing at high intensity (Zhang et al., 2014). In temporal comb propagation, Talbot beating provides the discrete modal scaffold on which SPM, XPM, and FWM build A-type breathers, soliton crystals, or soliton gas (Zajnulina, 17 Aug 2025). In second-harmonic and EIT platforms, by contrast, the Talbot effect remains tied to Fresnel-type propagation, but the periodic source or grating is produced nonlinearly or coherently (Liu et al., 2013, Zhang et al., 2016).

This diversity has produced a persistent interpretive issue. One line of work explicitly argued that earlier reports labeled “nonlinear Talbot effect” often involved linear Talbot diffraction of a nonlinear wave, whereas the rogue-wave Kerr effect is intrinsically nonlinear because the propagation medium itself is nonlinear and the recurring pattern is an NLSE eigenmode (Zhang et al., 2014). The broader literature nonetheless continues to use the same term for second-harmonic source Talbot imaging, EIT-induced gratings, temporal fiber recurrences, and nonlinear dispersive quantization on periodic domains. A precise reading therefore requires specifying which nonlinearity is at issue: nonlinear propagation, nonlinear generation, nonlinear susceptibility, or nonlinear dispersive flow.

The applications described in the surveyed works are correspondingly diverse. Rogue-wave Kerr recurrence was proposed as potentially useful for all-optical switching, beam multiplexing, and pattern replication inside nonlinear devices (Zhang et al., 2014). The 2D Kerr effect was suggested for the production of 3D photonic crystals (Zhang et al., 2014). Second-harmonic Talbot superfocusing was discussed in connection with super-resolution imaging, nanolithography, optical data storage, and biosensing (Liu et al., 2013). The temporal nonlinear Talbot effect was linked to pulse repetition-rate multiplication (Zajnulina et al., 2024). The EIT-induced atomic version was presented as a possible route toward lensless and nondestructive imaging of ultracold atoms and molecules (Zhang et al., 2016). Taken together, these results establish the nonlinear Talbot effect not as a single phenomenon with a single formula, but as a technically unified family of nonlinear self-imaging and revival processes governed by periodicity, phase synchronization, and the specific way nonlinearity enters the wave dynamics.