Description of Monte Carlo simulation
=====================================
Figure 2 of the paper [1] shows a 10,000-second-long simulated time series containing binary black hole (BBH) and binary neutron star (BNS) events. Here, we will briefly describe the Monte Carlo simulation used to produce this time series.
First, a list of binaries was produced. The following parameters are associated with each binary system:
* whether the system is a BBH or BNS event,
* tc - coalescence time,
* m1, m2 - source frame component masses,
* z - redshift,
* theta,phi - sky location,
- iota,psi - polarization angles.
The coalescence times for each binary were drawn from a Poisson distribution, with a rate \tau corresponding to the median values in Table I of Ref [1]. For BNS, \tau_BNS=13 s, for BBH \tau_BBH=223 s.
The source frame masses were drawn from the distributions specified in the paper. For BBH, the masses followed a power-law distribution in the primary mass, p(m1)~m1^a, with a=-2.35, and uniform in the secondary mass m2, subject to the constraints m1>=m2>=5 solar masses, and m1+m2<=100 solar masses. For BNS, the masses are drawn from a uniform distribution from 1 to 2 solar masses.
The redshift distribution is also as described in the paper. The probability of finding a source at a given redshift is related to the star formation rate and the distribution of time delays between binary formation and merger, for example see Eqn. 35 of Ref. [2]. We used the star formation rate of Ref. [3]. The time delay distribution follows a power law, P(t_d)~1/t, where tmin