No Time for Dead Time: Timing analysis of bright black hole binaries with NuSTAR
Timing of high-count rate sources with the NuSTAR Small Explorer Mission requires specialized analysis techniques. NuSTAR was primarily designed for spectroscopic observations of sources with relatively low count-rates rather than for timing analysis of bright objects. The instrumental dead time per event is relatively long (~2.5 msec), and varies by a few percent event-to-event. The most obvious effect is a distortion of the white noise level in the power density spectrum (PDS) that cannot be modeled easily with the standard techniques due to the variable nature of the dead time. In this paper, we show that it is possible to exploit the presence of two completely independent focal planes and use the cross power density spectrum to obtain a good proxy of the white noise-subtracted PDS. Thereafter, one can use a Monte Carlo approach to estimate the remaining effects of dead time, namely a frequency-dependent modulation of the variance and a frequency-independent drop of the sensitivity to variability. In this way, most of the standard timing analysis can be performed, albeit with a sacrifice in signal to noise relative to what would be achieved using more standard techniques. We apply this technique to NuSTAR observations of the black hole binaries GX 339-4, Cyg X-1 and GRS 1915+105.