TASEP on a Ring in Sub-relaxation Time Scale


Interacting particle systems in the KPZ universality class on a ring of size L with O(L) number of particles are expected to change from KPZ dynamics to equilibrium dynamics at the so-called relaxation time scale \(t=O(L^{3/2})\). In particular the system size is expected to have little effect to the particle fluctuations in the sub-relaxation time scale \(1\ll t\ll L^{3/2}\). We prove that this is indeed the case for the totally asymmetric simple exclusion process (TASEP) with two types of initial conditions. For flat initial condition, we show that the particle fluctuations are given by the Airy\(_1\) process as in the infinite TASEP with flat initial condition. On the other hand, the TASEP on a ring with step initial condition is equivalent to the periodic TASEP with a certain shock initial condition. We compute the fluctuations explicitly both away from and near the shocks for the infinite TASEP with same initial condition, and then show that the periodic TASEP has same fluctuations in the sub-relaxation time scale.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11
Fig. 12


  1. 1.

    The paper [9] only states the result for \(\rho =1/2\), but the more general case \(\rho =1/d\), \(d=2,3,4,\ldots \), is similar. See [8] for a discrete version. We note that the more general case \(\rho =p/q\) for integer p and q, however, is still open for both infinite TASEP and TASEP on a ring.

  2. 2.

    A lattice path is a path consists of unit horizontal/vertical line segments whose endpoints are lattice points.

  3. 3.

    \(|\mathbf v|=\sqrt{\mathbf v_1^2+\mathbf v_2^2}\) denotes the norm of \(\mathbf v=(\mathbf v_1, \mathbf v_2)\).

  4. 4.

    These results are for the Poissonian version of DLPP and the geometric random variables, but the results extend to exponential random variables. But we do not need these results. Instead we use the ideas in them to prove Proposition 2.1 below which implies this statement for exponential variables.

  5. 5.

    In the periodic DLPP case, we even have \(H_{\mathbf c_i}(\mathbf q)=H_{\mathbf c_{i+1}}(\mathbf q+\mathbf v)\) and \(H_{\mathbf c_{i+1}}(\mathbf q)=H_{\mathbf c_{i+2}}(\mathbf q+\mathbf v)\) due to the periodicity. See Fig. 11.


  1. 1.

    Baik, J., Ben Arous, G., Péché, S.: Phase transition of the largest eigenvalue for nonnull complex sample covariance matrices. Ann. Probab. 33(5), 1643–1697 (2005)

    MathSciNet  Article  MATH  Google Scholar 

  2. 2.

    Baik, J., Deift, P., McLaughlin, K.T.-R., Miller, P., Zhou, X.: Optimal tail estimates for directed last passage site percolation with geometric random variables. Adv. Theor. Math. Phys. 5(6), 1207–1250 (2001)

    MathSciNet  Article  MATH  Google Scholar 

  3. 3.

    Baik, J., Ferrari, P.L., Péché, S.: Convergence of the two-point function of the stationary TASEP. In: Griebel, M., (ed.) Singular Phenomena and Scaling in Mathematical Models, pp. 91–110. Springer, New York (2014)

  4. 4.

    Baik, J., Liu, Z.: Fluctuations of TASEP on a ring in relaxation time scale. arXiv:1605.07102

  5. 5.

    Basu, R., Sidoravicius, V., Sly, A.: Last passage percolation with a defect line and the solution of the slow bond problem. arxiv:1408.3464

  6. 6.

    Ben, G., Corwin, I.: Current fluctuations for TASEP: a proof of the Prähofer-Spohn conjecture. Ann. Probab. 39(1), 104–138 (2011)

    MathSciNet  Article  MATH  Google Scholar 

  7. 7.

    Borodin, A., Ferrari, P.L.: Large time asymptotics of growth models on space-like paths I: PushASEP. Electron. J. Probab. 13(50), 1380–1418 (2008)

    MathSciNet  Article  MATH  Google Scholar 

  8. 8.

    Borodin, A., Ferrari, P.L., Prähofer, M.: Fluctuations in the discrete TASEP with periodic initial configurations and the \({\rm Airy}_1\) process. Int. Math. Res. Pap. IMRP (1), Art. ID rpm002, 47 (2007)

  9. 9.

    Borodin, A., Ferrari, P.L., Prähofer, M., Sasamoto, T.: Fluctuation properties of the TASEP with periodic initial configuration. J. Stat. Phys. 129(5–6), 1055–1080 (2007)

    ADS  MathSciNet  Article  MATH  Google Scholar 

  10. 10.

    Bothner, T., Liechty, K.: Tail decay for the distribution of the endpoint of a directed polymer. Nonlinearity 26(5), 1449–1472 (2013)

    ADS  MathSciNet  Article  MATH  Google Scholar 

  11. 11.

    Brankov, J.G., Papoyan, V.B., Poghosyan, V.S., Priezzhev, V.B.: The totally asymmetric exclusion process on a ring: exact relaxation dynamics and associated model of clustering transition. Phys. A 368(8), 471480 (2006)

    Google Scholar 

  12. 12.

    Corwin, I., Hammond, A.: Brownian Gibbs property for Airy line ensembles. Invent. Math. 195(2), 441–508 (2014)

    ADS  MathSciNet  Article  MATH  Google Scholar 

  13. 13.

    Corwin, I., Liu, Z., Wang, D.: Fluctuations of TASEP and LPP with general initial data. Ann. Appl. Probab. 26(4), 2030–2082 (2016)

  14. 14.

    Derrida, B., Lebowitz, J.L.: Exact large deviation function in the asymmetric exclusion process. Phys. Rev. Lett. 80(2), 209–213 (1998)

    ADS  MathSciNet  Article  MATH  Google Scholar 

  15. 15.

    Ferrari, P.L., Nejjar, P.: Anomalous shock fluctuations in TASEP and last passage percolation models. Probab. Theory Relat. Fields 161(1–2), 61–109 (2015)

    MathSciNet  Article  MATH  Google Scholar 

  16. 16.

    Gupta, S., Majumdar, S.N., Godrèche, C., Barma, M.: Tagged particle correlations in the asymmetric simple exclusion process: finite-size effects. Phys. Rev. E 76, 021112 (2007)

    ADS  MathSciNet  Article  Google Scholar 

  17. 17.

    Gwa, L.-H., Spohn, H.: Six-vertex model, roughened surfaces, and an asymmetric spin Hamiltonian. Phys. Rev. Lett. 68(6), 725–728 (1992)

    ADS  MathSciNet  Article  MATH  Google Scholar 

  18. 18.

    Johansson, K.: Shape fluctuations and random matrices. Commun. Math. Phys. 209(2), 437–476 (2000)

    ADS  MathSciNet  Article  MATH  Google Scholar 

  19. 19.

    Johansson, K.: Transversal fluctuations for increasing subsequences on the plane. Probab. Theory Relat. Fields 116(4), 445–456 (2000)

    MathSciNet  Article  MATH  Google Scholar 

  20. 20.

    Johansson, K.: Discrete polynuclear growth and determinantal processes. Commun. Math. Phys. 242(1–2), 277–329 (2003)

    ADS  MathSciNet  Article  MATH  Google Scholar 

  21. 21.

    Lee, D.S., Kim, D.: Universal fluctuation of the average height in the early-time regime of one-dimensional Kardar–Parisi–Zhang-type growth. J. Stat. Mech. Theory Exp. 2006(8):P08014 (2006)

  22. 22.

    Prähofer, M., Spohn, H.: Scale invariance of the PNG droplet and the Airy process. J. Stat. Phys. 108(5–6), 1071–1106 (2002). Dedicated to David Ruelle and Yasha Sinai on the occasion of their 65th birthdays

  23. 23.

    Proeme, A., Blythe, R.A., Evans, M.R.: Dynamical transition in the open-boundary totally asymmetric exclusion process. J. Phys. A 44, 035003 (2011)

    ADS  MathSciNet  Article  MATH  Google Scholar 

  24. 24.

    Prolhac, S.: Finite-time fluctuations for the totally asymmetric exclusion process. Phys. Rev. Lett. 116, 090601 (2016)

    ADS  Article  Google Scholar 

  25. 25.

    Quastel, J., Remenik, D.: Tails of the endpoint distribution of directed polymers. Ann. Inst. Henri Poincaré Probab. Stat. 51(1), 1–17 (2015)

    ADS  MathSciNet  Article  MATH  Google Scholar 

  26. 26.

    Sasamoto, T.: Spatial correlations of the 1D KPZ surface on a flat substrate. J. Phys. A 38(33), L549–L556 (2005)

    ADS  MathSciNet  Article  Google Scholar 

  27. 27.

    Tracy, C.A., Widom, H.: Level-spacing distributions and the Airy kernel. Commun. Math. Phys. 159(1), 151–174 (1994)

    ADS  MathSciNet  Article  MATH  Google Scholar 

  28. 28.

    Tracy, C.A., Widom, H.: On orthogonal and symplectic matrix ensembles. Commun. Math. Phys. 177(3), 727–754 (1996)

    ADS  MathSciNet  Article  MATH  Google Scholar 

Download references


We would like to thank Ivan Corwin and Patrik Ferrari for useful conversations and comments. The work of Jinho Baik was supported in part by NSF Grants DMS1361782.

Author information



Corresponding author

Correspondence to Zhipeng Liu.

Appendix: Density Profile of TASEP with Periodic Step Initial Condition

Appendix: Density Profile of TASEP with Periodic Step Initial Condition

In this appendix, we summarize the macroscopic picture of the periodic TASEP and the infinite TASEP with periodic step initial condition (1.9) via solving the Burger’s equation. We state the density profile, and the locations of the shock and any given particle as time t without much details since the computation is standard. Furthermore we do not study the issue of the convergence in the hydrodynamic limit to the Burger’s solution; the computations in this Appendix are used only to provide intuitive ideas and are not used in the proofs of the theorems.

We assume that \(0<\rho \le \frac{1}{2}\). Consider the Burger’s equation for the infinite TASEP

$$\begin{aligned} \frac{{\mathrm d}}{{\mathrm d}t}q(x;t) +\frac{{\mathrm d}}{{\mathrm d}x}\left( q(x,t)(1-q(x,t))\right) =0 \end{aligned}$$

with the periodic initial condition

$$\begin{aligned} q(x;0)={\left\{ \begin{array}{ll} 1,\qquad &{}-\rho \le x-[x]-1\le 0,\\ 0,\qquad &{}0< x-[x]< 1-\rho , \end{array}\right. } \end{aligned}$$

where [x] means the largest integer which is less than or equal to x. The entropy solution q(xt) represents the local density profile at location xL and time tL. Note that the solution is also periodic, \(q(x+1, t)=q(x,t)\), and hence q(xt) also represents the local density profile for the periodic TASEP with the same initial condition.

We now solve the above Burger’s equation explicitly. Due to the periodicity, we state the formula of q(xt) for x only in an interval of length 1.

For time \(t\le \frac{1}{4\rho }\), there is no shock and the solution is given by the following: For \(0\le t\le \rho \),

$$\begin{aligned} q(x;t)= {\left\{ \begin{array}{ll} 1,&{} -\rho \le x\le -t,\\ \frac{1}{2}-\frac{1}{2t}x,&{} -t<x<t,\\ 0,&{} t\le x<1-\rho . \end{array}\right. } \end{aligned}$$

For \(\rho \le t\le \frac{1}{4\rho }\),

$$\begin{aligned} q(x,t)= {\left\{ \begin{array}{ll} \displaystyle \frac{1}{2}-\frac{1}{2t}x, &{}\quad -2\sqrt{\rho t} +t\le x\le t,\\ 0, &{}\quad t<x< -2\sqrt{\rho t} +t+1. \end{array}\right. } \end{aligned}$$

The shocks are generated at time \(t=\frac{1}{4\rho }\) at the locations \(\frac{1}{4\rho }+\mathbb {Z}\). (In terms of the TASEP, the above time corresponds to the time \(\frac{1}{4\rho }L=\frac{1}{4\rho ^2}N\).) Let us denote by \(x_\mathrm{s}(t)\) the location of the shock of the Burger’s equation at time t which was initially generated at the location \(-1+\frac{1}{4\rho }\), i.e. \(x_\mathrm{s}(\frac{1}{4\rho })=-1+\frac{1}{4\rho }\). One can find that the shock location is given by

$$\begin{aligned} x_\mathrm{s}(t)=-\frac{1}{2}+(1-2\rho )t \end{aligned}$$

and the density profile is given by

$$\begin{aligned} q(x;t)=\frac{1}{2}-\frac{1}{2t}x, \quad x_\mathrm{s}(t)\le x < x_\mathrm{s}(t)+1 \end{aligned}$$

for all \(t\ge \frac{1}{4\rho }\). This shows that the density profile difference at the shock, \(\Delta q_\mathrm{s}(t):= \lim _{x\rightarrow x_\mathrm{s}(t)^+} q(x; t)-\lim _{x\rightarrow x_\mathrm{s}(t)^-} q(x; t)\), is given by \(\Delta q_\mathrm{s}(t)=\frac{1}{2t}\) at time \(t\ge \frac{1}{4\rho }\). As \(t\rightarrow \infty \), this gap tends to zero and \(q(x;t)\rightarrow \rho \) for all \(x\in \mathbb {R}\). However, the density profile is not yet “flat enough” when \(t\ll L^{1/2}\) (which corresponds to the sub-relaxation time scale \(t\ll L^{3/2}\) in TASEP). Indeed, note that that when \(t\ll L^{1/2}\), the gap satisfies \( \Delta q_\mathrm{s}(t)\gg \frac{1}{L^{1/2}}\) (and the absolute value of the slope of the density profile at continuous points is \(\gg \frac{1}{L^{1/2}}\).) In terms of the TASEP scale of time and space, \(\Delta q_\mathrm{s}(t) L \gg L^{1/2} \gg (tL)^{1/3}\) which means that the gap is greater than the KPZ height fluctuations.

Given the formula of the density profile, we can compute the expected location of the \([\alpha N]\)-th particle (the one initially located at \(-N+[\alpha N]\)) heuristically. Here \(\alpha \) is an arbitrary constant satisfying \(0<\alpha \le 1\). This particle meets a shock at the discrete (rescaled by L) times

$$\begin{aligned} \frac{\left( \sqrt{k-\alpha } +\sqrt{k+1-\alpha }\right) ^2}{ 4\rho }, \qquad k=1,2,\ldots . \end{aligned}$$

The particles location (rescaled by L) is heuristically given by

$$\begin{aligned} \begin{aligned} X_{\alpha }(t)&=(\sqrt{t}-\sqrt{(k+1-\alpha )\rho })^2-(k+1-\alpha ) \rho +k,\\&=t(1-\rho )+(\sqrt{t\rho }-\sqrt{k+1-\alpha })^2-(1-\rho )(1-\alpha )+ X_{\alpha }(0) \end{aligned} \end{aligned}$$

for time satisfying

$$\begin{aligned} \frac{\left( \sqrt{k-\alpha } +\sqrt{k+1-\alpha }\right) ^2}{ 4\rho }\le t <\frac{\left( \sqrt{k+1-\alpha } + \sqrt{k+2-\alpha }\right) ^2}{ 4\rho }. \end{aligned}$$

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Baik, J., Liu, Z. TASEP on a Ring in Sub-relaxation Time Scale. J Stat Phys 165, 1051–1085 (2016). https://doi.org/10.1007/s10955-016-1665-y

Download citation


  • TASEP on a ring
  • Sub-relaxation time scale
  • KPZ universality class