Evolution of Compact-Binary Populations in Globular Clusters: A Boltzmann Study II. Introducing Stochasticity

TL;DR

This paper extends a Boltzmann scheme to include stochastic processes in the evolution of compact-binary populations in globular clusters, using Ito calculus to model fluctuations and comparing results with observations.

Contribution

It introduces a stochastic approach using Wiener processes and Ito calculus into the Boltzmann scheme for binary evolution in globular clusters, enhancing the modeling of dynamical stochasticity.

Findings

Average stochastic results align with continuous-limit predictions.

Scaling of X-ray binary numbers matches observed correlations.

Stochastic fluctuations vary but do not alter overall trends.

Abstract

We continue exploration of the Boltzmann scheme started in Banerjee and Ghosh (2007, henceforth Paper I) for studying the evolution of compact-binary populations of globular clusters, introducing in this paper our method of handling the stochasticity inherent in dynamical processes of binary formation, destruction and hardening in globular clusters. We describe these stochastic processes as "Wiener processes", whereupon the Boltzmann equation becomes a stochastic partial differential equation, the solution of which requires the use of "Ito calculus" (this use being the first, to our knowledge, in this subject), in addition to ordinary calculus. We focus on the evolution of (a) the number of X-ray binaries $N_{XB}$ in globular clusters, and (b) the orbital-period distribution of these binaries. We show that, although the details of the fluctuations in the above quantities differ from one "realization" to another of the stochastic processes, the general trends follow those found in the continuous-limit study of Paper I, and the average result over many such realizations is close to the continuous-limit result. We investigate the dependence of $N_{XB}$ found by these calculations on two essential globular-cluster parameters, namely, the star-star and star-binary encounter-rate parameters $\Gamma$ and $\gamma$, for which we had coined the name Verbunt parameters in Paper I. We compare our computed results with those from CHANDRA observations of Galactic globular clusters, showing that the expected scalings of $N_{XB}$ with the Verbunt parameters are in good agreement with the observed ones. We indicate what additional features can be incorporated into the scheme in future, and how more elaborate problems can be tackled.