Full text loading...
(a) Conceptual illustration of a narrowband extreme event in SC generation. (b) Experimental measurement of red-filtered SC output power vs average power of input pulse train. Right-skewed heavy-tailed statistics exist below threshold; left-skewed (reverse) heavy-tailed statistics appear above threshold. (c) Experimental observation of reverse-heavy-tailed statistics at the input power level labeled by the arrow in B. Gaussian is fit to the main portion of the distribution. (d) Filtered SC pulse train measured in real time.
Reverse-heavy-tailed distribution from simulation of 2000 independent events (red-filtered energy). (, ) Gaussian is fit to the main portion of the distribution; a portion of the extreme tail is shown in log scale. Inset: Spectra of one anomalous (green) and one normal event (blue). Pump (input) spectrum also shown (black).
(a) A scanned time-frequency filter (Gaussian profile, ; ; center position specified by x and y axes) is applied to the input noise of two simulated anomalous events; the NLSE is then solved. [(b) and (c)] Redshifted energy at the fiber output for each filter position and two separate anomalous (frustrated broadening) events.
Windowed Fourier transforms of simulated fiber outputs for different initial conditions. (a) Anomalous event . (b) Same event after removal of a noise component as shown in Fig. 3. (c) Noiseless Gaussian interfering seed pulses. (d) Noiseless Gaussian seed pulse.
Propagation in the fiber of (a) an anomalous and (b) a normal event above the soliton fission threshold. Color scale illustrates the field magnitude (bright: Large magnitude). Arrows in (a) point to three presolitonic features. (c) Statistical analysis of the number of separate peaks above an arbitrary reference level in normal events (blue) and frustrated events (green, black outline) at the end of the fiber.
Article metrics loading...