Diego Altamirano: "Millihertz quasi-periodic oscillations and their relation with Type-I X-ray Bursts: an intimate relation?"

Diego discussed whether there is a strong link between the mHz QPOs seen in some LMXBs and the occurence of Type I X-ray bursts. Until now we haven't been able to predict X-ray bursts, the recurrence time of which depend on a lot of unknown parameters. However, mHz QPOs, sometimes drifting in frequency, disapear just before an X-ray burst. This is the first time that something in the persistent emission seems to predict an X-ray burst (though mHz QPOs are not always seen before a burst). It appears that one won't see a burst unless the frequency of the mHz QPO drops below some threshold (9mHz appears to be the "magic number" in e.g. 4U1636-53). How do we interpret this? It could possibly be due to marginally stable burning.

Are mHz QPOs a new way to study X-ray bursts and changes in spectral states? Why is 9mHz the "magic number" and why there is frequency drift remains unknown.