2004ApJ...611L.121Y

We have analyzed Rossi X-Ray Timing Explorer pointed observations of the outbursts of black hole and neutron star soft X-ray transients in which an initial low/hard state, or ``island'' state, followed by a transition to a softer state was observed. In three sources–the black hole transient XTE J1550-564, the neutron star transient Aquila X-1, and the quasi-persistent neutron star low-mass X-ray binary 4U 1705-44–two such outbursts were found. We find that the flux of the soft X-ray peak, which lags the hard X-ray peak by a few days to several weeks, scales with the flux of the hard X-ray peak. We conclude that we are able to predict the soft X-ray outburst peak flux based on the ``preceding'' hard X-ray peak flux, implying an early setup of the outbursts. We also find that the X-ray luminosity corresponding to the peak of the hard X-ray flux, which corresponds to the X-ray luminosity of the start of the hard-to-soft state transition, varies by a factor of about 2. If the accretion geometry early in the outburst rise is composed of two flows (e.g., a hot sub-Keplerian halo flow and a Keplerian disk flow or an outflow and a Keplerian disk flow), the correlation indicates that the two flows are initially related, probably owing to processes in the outer part of the accretion disk. We discuss constraints on a single-flow model and a disk-jet model from these observations.