Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Laminar-Turbulent Transition in Raman Fiber Lasers: A First Passage Statistics Based Analysis


Loss of coherence with increasing excitation amplitudes and spatial size modulation is a fundamental problem in designing Raman fiber lasers. While it is known that ramping up laser pump power increases the amplitude of stochastic excitations, such higher energy inputs can also lead to a transition from a linearly stable coherent laminar regime to a non-desirable disordered turbulent state. This report presents a new statistical methodology, based on first passage statistics, that classifies lasing regimes in Raman fiber lasers, thereby leading to a fast and highly accurate identification of a strong instability leading to a laminar-turbulent phase transition through a self-consistently defined order parameter. The results have been consistent across a wide range of pump power values, heralding a breakthrough in the non-invasive analysis of fiber laser dynamics.


Raman fiber lasers are fascinating nonlinear optical systems which are conspicuous in their ability to give rise to a diverse range of operational regimes through an articulate control of system parameters1,2,3,4,5,6,7,8,9,10. These lasers utilize the stimulated Raman scattering mechanism occurring in the silica based fiber medium for gain. The small value of the Raman gain coefficient leads to the requirement of watt-order pump power levels and kilometer-order lengths of fiber for realizing the lasing threshold. The long cavity lengths give rise to very high mode densities, which together with the watt-order intracavity powers, lead to extremely complex nonlinear interactions between the modes. This has been explained within the framework of weak wave turbulence6. A complex interplay of nonlinearity and dispersion brings about a diverse palette of nonlinear dynamics, that include the generation of extreme events7, formation of optical condensates8, and a more recently discovered laser analog of a laminar to turbulent transition1.

It has been reported2 that unlike conventional modes of energy cascading, there is a certain phase locking between the inertial and dissipation energy scales at which energy cascades are respectively pumped in and out (dissipated) of the system, for which reason these two scales can not be non-equivocally unentangled. Effectively this amounts to an uncertainty in the inertial interval and hence in the categorization of the laminar-turbulent transition regime. While there are obvious theoretical and experimental challenges associated with such studies in relation to wave turbulent dynamics1, an understanding of the underlying nonlinear dynamics of such many degree-of-freedom systems can lead to the realization of a new generation of high power, coherent laser systems. The parameter space is however extremely large, and identification of crossovers to different operational regimes - and more importantly, their predetermination is a problem of interest, especially from an application perspective.

In this study, we propose the use of an alternative time domain methodology for the classification of lasing regimes known in the statistical physics literature as persistence probability density function (PPDF)11,12, drawn from the rich legacy of first passage statistics which essentially quantifies how long a random walker persists in any particular state before changing over to the other state(s). The effectiveness of the methodology is highlighted by using it to accurately identify the laminar-turbulent transition point in a Raman fiber laser. In the process, it is shown how the PPDF can be used for the definition of a suitable non-dimensional order parameter, which can then be used as a unique identifier of the lasing regime.


First passage statistics and the persistence probability distribution function

The concept underlying first passage extremal statistics, also popularly referred to as persistence statistics, relies on a specific characterization of temporal statistics based on an analysis of two-point autocorrelation functions11. The topic has been studied in grand details in the context of statistical mechanics12, as also in a variety of interdisciplinary applications, like fractional dynamics13,14, network characterization15, immunological synapse modelling16,17,18 and gene data sequencing19. The key question here is the role that extremal statistic plays in the measurement of a statistical quantity. As an example, one may consider large amplitude fluctuations in a stochastic time series data set and estimate how such extremal events influence a mean statistical property e.g. the relevant temporal probability density function. Once characterized as a cumulative distribution function, the resulting power-law like statistics12 could be utilized to define an order parameter to represent phase crossovers or transitions20, and generically phase boundaries.

This aforementioned order parameter that we intend to use for the purpose of characterizing our fiber laser data is defined through a two-step process, in which at step 1 we identify a local order parameter which then leads to a global order parameter at step 2. To arrive at these order parameters, we start from the following question: what is the probability p(t1; t2) that the laser excitation intensity X(t) changes sign N times between t1 and t2 for X(t) > X0? Depending on the nature of random walk, the statistical physics parlance for the stochastic time series (fiber laser intensity) data, this probability is given by

where the non-universal exponent αP changes with the pump power P and may actually be multi-fractal in nature. The subject has a rich heritage leading to the equation above that can be found in12,21. First passage statistics has been profitably utilized in interpreting phase transition scenarios21, a line of approach that we adopt here as well.

Depending on whether the statistics is evaluated above or below the line X(t) = X0, the exponent P would generally have two different values, one above the threshold and the other below it. In our following analysis, we would call the statistics above the threshold line X(t) = X0 as being of the plus-type () whereas the ones below the threshold will be of the minus-type (). The adjoining Fig. 1 illustrates these definitions in which t+ refers to the time spent between two crossings above the line X(t) = X0(X(t) > X0), whereas t is the time spent between two crossings below the line X(t) = X0. αP is our initial estimate for a non-dimensional local order parameter at step 1. Our analysis uses αP as the effective local order parameter characterizing the phase transition since it depicts the return time from a highly excited turbulent regime to that of a relatively transient laminar profile, thereby ensuring a much higher number of statistical events characterizing it due to low energy cost compared to the take-off time distribution scenario ( statistics). αP will shortly be shown to naturally lead to the ensemble averaged global order parameter.

Figure 1: Brief Title Sentence: Cartoon diagram of Persistence mechanism.

Stochastic dynamics of the fiber laser excitation amplitude X(t) as a function of time t across the threshold X0, where X0 represents the ensemble average of X(t).


The PPDF methodology as described above is now applied to the study of the Raman fiber laser exhibiting a laminar-to-turbulent transition in intensity dynamics, as obtained in Turitsyna et al.1. The experimental apparatus uses a 770 meter long span of normal dispersion fiber (D = −44 ps/nm-km). Two highly reflective (R = 0.99) 1 nm wide super-Gaussian gratings centered at 1555 nm were spliced to the ends of the fiber to form a Fabry-Perot cavity. This was then optically pumped using a watt-order Raman fiber laser operating in the 1455 nm region. A fiber-based fused coupler (99:1 ratio) was then used to obtain a small fraction of the intracavity radiation, which was interrogated using a 50 GHz optical photodetector, attached to a 36 GHz real-time digital storage oscilloscope. The intensity dynamics of the laser was then recorded over a time span equivalent to 120 round trips, sufficient to build up laser intensity statistics. The output characteristics were measured over a range of pump powers to progressively track the phase behavior with change in power inputs. Figure 2 shows the typical dynamics of a laminar-turbulent Raman fiber laser below (blue curves) and above (orange) the transition. In the laminar regime, the spectral width of emission, as monitored using an optical spectrum analyzer (resolution 20 pm), is ~40 pm; this increases by almost an order of magnitude to ~125 pm in the turbulent regime1. The spectral width remains stable over the course of measurement, with the changes in the instantaneous laser intensity primarily arising due to turbulent four wave mixing interactions between the different longitudinal modes6. The difference in the dynamics of the operational regimes are further evidenced by the well marked difference between their intensity probability distribution functions. Note that negative excursions in the measured intensities can arise from either photodetector noise, or electrical bandwidth limitations of the photodetector-oscilloscope. Nevertheless, the intensity statistics can be reliably estimated even under such noise and bandwidth limited considerations22. In our analysis, we tracked the range between P = 0.85 W to 2.20 W in steps of 0.05 W.

Figure 2: Brief Title Sentence: Experimental intensity time series data.

(a) Experimental configuration; FBG - fibre Bragg grating, DSO - digital storage oscilloscope, OSA - optical spectrum analyzer. (b) Laminar (blue) and turbulent (orange) lasing regimes of the Raman fibre laser. Inset - Intensity PDFs of the corresponding regimes (also see Turitsyna et al.1).

Figure 3 depicts three such persistence probability estimation within the aforementioned pump power value range in a log-log plot, where P refers to the value of the gradient at large t values. It is important to note that for all pump powers, the corresponding PDFs show clear gradient change e.g. at around log(t+) in Fig. 3. Our focus here is on the larger values of crossover times e.g. t to emphasize on the extremal values of excursion. The alluded αP, though, estimates only the transient persistence of the data and not a global average over the given range of pump power values. In order to evaluate the average time persistence of the dynamics in either state, laminar or turbulent, we need an ensemble averaged global order parameter that will address the global symmetry of the model, a popular approach in liquid crystal studies23,24. We define our effective (global) order parameter as

Figure 3: Brief Title Sentence: Temporal Probability Density Function (PDF) plot.

Probability Density Function (PDF) of “persistent” time series statistics in the log-log scale for pump power values P = 0.95, 1.0 and 1.05 watts: minus-type. The solid lines represent actual data while the dotted lines are the respective gradient fits (straight lines).

that quantifies the cumulative change of the gradient of the persistence distribution (measured by βP) with change in pump power P over the entire range of βP values. This is step 2 of the methodology in which a global order parameter βP is estimated that accurately quantifies the laminar-turbulent phase transition as shown in Fig. 4. In our analysis, we have used the Ehrenfest classification of characterizing phase transitions based on discontinuous jumps of the order parameter25. In effect, βP is a non-Markovian order parameter that tracks all changes in the gradients of the persistence probability density function as a function of P. The emergence of βP is key to our analysis of the system as such a statistically averaged quantity reduces the impact of fluctuations, measured by the corresponding standard deviation, in the regression (least square) fits in Fig. 3 that lead us to the values of αP, thereby stabilizing the phase transition picture overall. The error bars in Fig. 4 bear testimony to this order parameter stability aspect.

Figure 4: Brief Title Sentence: Phase transition from order parameter plot.

Order parameter βP plotted against pump power P. Inset shows comparison with experimental data; theoretical phase transition prediction closely matches the experimental observation at P ~ 1.05 watt.

Figure 4 outlines the strength of this algorithm by accurately matching the theoretical phase transition point (outset) with that observed in experiment as a step-like transition in the spectral width of the laser emission (inset). As the highlight of this study, this plot almost unerringly identifies a discontinuous jump at around P = 1.05 W, the critical laminar-turbulent phase transition point that agrees well with the independent experimental result recently published in Fig. 4 (inset)1. The errorbars in the figure (outset), calculated as standard deviations of βP as evaluated from measured data, show fluctuations within 1% of the estimated “order parameter” data, thereby confirming a very high degree of accuracy from this analysis. This is indeed remarkable given that the approach does not rely on any adjustable external parameter and relates only to the inherent stochasticity of the data leading to a self-consistent prediction of the laminar-turbulent phase transition point.


This work presents the mapping of a well established analytical structure from statistical physics in analyzing a key photonics problem. The PPDF methodology exactly identifies the laminar-turbulent phase transition point from the fiber laser intensity dynamics by estimating a self-consistent order parameter as a function of the (variable) excitation power used, independent of the nature of the sample. Indeed, the application of this technique is not limited to the case of the laminar-turbulent laser shown above, and can be used to study and classify the operational regimes of the wider class of quasi-continuous wave fiber lasers in general. The presented study extends further the scope of first passage statistics based analyzes in photonics as was reported in a recent work26,27,28,29, where statistical methods were used to aid the understanding of the onset of rogue waves in certain optical systems. As a prime example of statistical physics based analysis of photonics data, this report holds the potential to serve as a much simpler, and yet a refined powerful diagnostic tool for the analysis of complex nonlinear fiber laser dynamics.


Ethical Approval and Informed Consent

All experimental data (none involving live or biological samples) were collected in full compliance to requirements set by the Aston University Ethics Committee (

Data Availability

Data related to Figures 2, 3, 4 can be accessed from the following doi:

Additional Information

How to cite this article: Chattopadhyay, A. K. et al. Laminar-Turbulent Transition in Raman Fiber Lasers: A First Passage Statistics Based Analysis. Sci. Rep. 6, 28492; doi: 10.1038/srep28492 (2016).


  1. 1

    Turitsyna, E. G. et al. The laminar-turbulent transition in a fibre laser. Nature Photon. 7, 783–786 (2013).

    CAS  ADS  Article  Google Scholar 

  2. 2

    Turitsyn, S. K. et al. Optical wave turbulence. Advances in Wave Turbulence 83, 113–164 (2013).

    MathSciNet  Article  Google Scholar 

  3. 3

    Turitsyn, S. K. et al. Random distributed feedback fibre laser. Nature Photon. 4, 231–235 (2010).

    CAS  ADS  Article  Google Scholar 

  4. 4

    Headley, C., Mermelstein, M. & Bouteiller, J.-C. in Raman Amplifiers for Telecommunications 2, Springer Series in Optical Sciences, Vol. 90/2. (ed. Islam, M. N. ) 353–382 (Springer, New York, 2004).

    ADS  Article  Google Scholar 

  5. 5

    Agrawal, G. P. Nonlinear Fiber Optics (Academic Press, 2007).

  6. 6

    Babin, S. A., Churkin, D. V., Ismagulov, A. E., Kablukov, S. I. & Podivilov, E. V. Four-wave-mixing-induced turbulent spectral broadening in a long Raman fiber laser. J. Opt. Soc. Am. B 24(8), 1729–1738 (2007).

    CAS  ADS  Article  Google Scholar 

  7. 7

    Churkin, D. V., Gorbunov, O. A. & Smirnov, S. V. Extreme value statistics in Raman fiber lasers. Opt. Lett. 36(18), 3617–3619 (2011).

    ADS  Article  Google Scholar 

  8. 8

    Turitsyna, E. G. et al. Optical turbulence and spectral condensate in long fibre lasers. Proc. Roy. Soc. A 468(2145), 2496–2508 (2012).

    CAS  ADS  MathSciNet  Article  Google Scholar 

  9. 9

    Churkin, D. V., Smirnov, S. V. & Podivilov, E. V. Statistical properties of partially coherent CW fiber lasers. Opt. Lett. 35, 3288–32290 (2010).

    CAS  ADS  Article  Google Scholar 

  10. 10

    Churkin, D. V. et al. Stochasticity, periodicity and localized light structures in partially mode-locked fibre lasers. Nature Comms. 6, 7004 (2015).

    CAS  ADS  Article  Google Scholar 

  11. 11

    Bouchaud, J.-P. & Georges, A. Anomalous diffusion in disordered media: Statistical mechanisms, models and physical applications. Phys. Rep. 195, 127–293 (1990).

    ADS  MathSciNet  Article  Google Scholar 

  12. 12

    Majumdar, S. N. Persistence in non-equilibrium systems. Current Science 77, 370–375 (1999).

    Google Scholar 

  13. 13

    Metzler, R. & Klafter, J. The random walk’s guide to anomalous diffusion: a fractional dynamics approach. Phys. Rep. 339, 1–77 (2000).

    CAS  ADS  MathSciNet  Article  Google Scholar 

  14. 14

    Sanders, L. P., Lomholt, M. A., Lizana, L., Fogelmark, K., Metzler, R. & Ambjörnsson, T. Severe slowing-down and universality of the dynamics in disordered interacting many-body systems: ageing and ultraslow diffusion. New J. Phys. 16, 113050 (2014).

    ADS  Article  Google Scholar 

  15. 15

    Lin, Y. & Zhang, Z. Mean first-passage time for maximal-entropy random walks in complex networks. Sci. Rep. 4, 5365 (2014).

    CAS  ADS  Article  Google Scholar 

  16. 16

    Chattopadhyay, A. K. & Burroughs, N. J. Close contact fluctuations: The seeding of signalling domains in the immunological synapse. Europhys. Lett. 77, 48003 (2007).

    ADS  Article  Google Scholar 

  17. 17

    Bush, D. R. & Chattopadhyay, A. K. Contact time periods in immunological synapse. Phys. Rev. E 90, 042706 (2014).

    ADS  Article  Google Scholar 

  18. 18

    Bush, D. R. & Chattopadhyay, A. K. Temporal dynamics in an immunological synapse: Role of thermal fluctuations in signaling. Phys. Rev. E 92, 012706 (2015).

    ADS  MathSciNet  Article  Google Scholar 

  19. 19

    Chattopadhyay, A. K., Nasiev, D. & Flower, D. R. A statistical physics perspective on alignment-independent protein sequence comparison. Bioinformatics 31(15), 2469–2474 (2015).

    CAS  Article  Google Scholar 

  20. 20

    Majumdar, S. N. & Bray, A. J. Persistence with Partial Survival. Phys. Rev. Lett. 81, 2626–2629 (1998).

    CAS  ADS  Article  Google Scholar 

  21. 21

    Metzler, R., Oshanin, G. & Redner, S. First Passage Phenomena and Their Applications, Vol. 35 (eds Metzler, R. et al. ) (World Scientific, 2014).

  22. 22

    Gorbunov, O. A., Sugavanam, S. & Churkin, D. V. Revealing statistical properties of quasi-CW fibre lasers in bandwidth-limited measurements. Opt. Express 22, 28071–28076 (2014).

    CAS  ADS  Article  Google Scholar 

  23. 23

    de-Gennes, P. The Physics of Liquid Crystals (Oxford University Press, USA, 2002).

  24. 24

    Mukherjee, P. K. Phase transitions among the rotator phases of the normal alkanes: A review. Phys. Rep. 588, 1–54 (2015).

    CAS  ADS  MathSciNet  Article  Google Scholar 

  25. 25

    Zemansky, M. W., Dittman, R. H. & Chattopadhyay A. K. (adaptor). Heat and Thermodynamics (Tata McGraw-Hill, 2011).

  26. 26

    Birkholz, S., Brée, C., Demircan, A. & Steinmeyer, G. Predictability of Rogue Events. Phys. Rev. Lett. 114, 213901 (2015).

    ADS  Article  Google Scholar 

  27. 27

    Erkintalo, M. Rogue waves: Predicting the unpredictable? Nat. Photon. 9, 560–562 (2015).

    CAS  ADS  Article  Google Scholar 

  28. 28

    Aragoneses, A. et al. Unveiling Temporal Correlations Characteristic of a Phase Transition in the Output Intensity of a Fiber Laser. Phys. Rev. Lett. 116(3), 033902 (2016).

    CAS  ADS  Article  Google Scholar 

  29. 29

    Kalashnikov, V., Sergeyev, S. V., Jacobsen, G., Popov, S. & Turitsyn, S. K. Multi-scale polarisation phenomena. Light Sci. Appl. 5, el6011 (2016).

    ADS  Article  Google Scholar 

Download references


All authors acknowledge “Aston University” funding support. SS, NT, and DVC acknowledge the support by ERC (grant 267763 “Ultralaser”) and H2020 project CARDIALLY. DVC also acknowledges the Russian Ministry of Science and Education (14.584.21.0014).

Author information




A.K.C. conceived the idea and led the project. D.N. wrote the numerical code for the processing of data. S.S. and N.T. did the experiments, supervised by D.V.C. and A.K.C.; A.K.C. and S.S. jointly wrote the manuscript.

Corresponding author

Correspondence to Amit K. Chattopadhyay.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Rights and permissions

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Chattopadhyay, A., Nasiev, D., Sugavanam, S. et al. Laminar-Turbulent Transition in Raman Fiber Lasers: A First Passage Statistics Based Analysis. Sci Rep 6, 28492 (2016).

Download citation


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing