|Two years of INTEGRAL monitoring of GRS 1915+105 Part 1: multiwavelength coverage with INTEGRAL, RXTE, and the Ryle radio Telescope|
|J. Rodriguez, D.C. Hannikainen, S.E. Shaw, G. Pooley, S. Corbel, M. Tagger, I.F. Mirabel, T. Belloni, C. Cabanac, M. Cadolle Bel, J. Chenevez, P. Kretschmar, H.J. Lehto, A. Paizis, P. Varniere, O. Vilhu|
We report the results of monitoring observations of the Galactic microquasar GRS 1915+105 performed simultaneously with INTEGRAL and RXTE from 3 up to ~300 keV, and the Ryle Telescope at 15 GHz. We present the results of the whole INTEGRAL campaign, report the sources that are detected and their fluxes and identify the classes of variability in which GRS 1915+105 is found. The long and continuous INTEGRALexposures enable us to see several direct transitions between different classes of variability. We focus on the connection between the behavior of GRS 1915+105 at X-ray energies and at radio wavelengths. The data are studied in a model independent manner through the source light curves, its hardness ratio, and color color diagrams. During a period of steady ``hard\'\' X-ray state (the so-called class chi) we observe a steady radio flux. This is interpreted as the signature of the presence of a compact jet. We then turn to 3 particular observations during which we observe several types of soft X-ray dips and spikes cycles, followed by radio flares. During these observations GRS 1915+105 is in the so-called nu, lambda, and beta classes of variability. The observation of ejections during class lambda are the first ever reported. Our model independent approach of the high energy data allows us to generalize the fact that a (non-major) discrete ejection always occurs, in GRS 1915+105, as a response to an X-ray sequence composed of a spectrally hard X-ray dip (more pronounced at soft X-rays) terminated by an X-ray spike marking the disappearance of the hard X-ray emission above 18~keV. This model independent approach also permits us to identify the trigger of the ejection as this X-ray spike. In addition, a possible correlation between the amplitude of the radio flare and the duration of the X-ray dip is found in our data. In this case the X-ray dips prior to ejections could be seen as the time during which the source accumulates energy and material that is ejected later. The fact that these results do not rely on any spectral modelling enhances their robustness.