LAUSR

We calculate the properties, occurrence rates and detection prospects of individually resolvable 'single sources' in the low frequency gravitational wave (GW) spectrum. Our simulations use the population of galaxies and massive black hole binaries from the Illustris cosmological hydrodynamic simulations, coupled to comprehensive semi-analytic models of the binary merger process. Using mock pulsar timing arrays (PTA) with, for the first time, varying red-noise models, we calculate plausible detection prospects for GW single sources and the stochastic GW background (GWB). Contrary to previous results, we find that single sources are at least as detectable as the GW background. Using mock PTA, we find that these 'foreground' sources (also 'deterministic'/'continuous') are likely to be detected with $\sim 20 \, \textrm{yr}$ total observing baselines. Detection prospects, and indeed the overall properties of single sources, are only moderately sensitive to binary evolution parameters---namely eccentricity & environmental coupling, which can lead to differences of $\sim 5 \,\textrm{yr}$ in times to detection. Red noise has a stronger effect, roughly doubling the time to detection of the foreground between a white-noise only model ($\sim 10$ - $15 \, \textrm{yr}$) and severe red noise ($\sim 20$ - $30 \, \textrm{yr}$). The effect of red noise on the GWB is even stronger, suggesting that single source detections may be more robust. We find that typical signal-to-noise ratios for the foreground peak near $f = 0.1 \, \textrm{yr}^{-1}$, and are much less sensitive to the continued addition of new pulsars to PTA.