We perform the most detailed analysis to date of the X-ray state of the Galactic black hole candidate GRSÂ 1915+105 just prior to ( 0 to 4 hours ) and during the brief ( 1 to 7 hour ) ejection of major ( superluminal ) radio flares .
A very strong model independent correlation is found between the 1.2 keV - 12 keV X-ray flux 0 to 4 hours before flare ejections with the peak optically thin 2.3 GHz emission of the flares .
This suggests a direct physical connection between the energy in the ejection and the luminosity of the accretion flow preceding the ejection .
In order to quantify this concept , we develop techniques to estimate the intrinsic ( unabsorbed ) X-ray luminosity , L _ { \mathrm { intrinsic } } , from RXTE ASM data and to implement known methods to estimate the time averaged power required to launch the radio emitting plasmoids , Q ( sometimes called jet power ) .
We find that the distribution of intrinsic luminosity from 1.2 keV - 50 keV , L _ { \mathrm { intrinsic } } ( 1.2 - 50 ) , is systematically elevated just before ejections compared to arbitrary times when there are no major ejections .
The estimated Q is strongly correlated with L _ { \mathrm { intrinsic } } ( 1.2 - 50 ) 0 to 4 hours before the ejection , the increase in L _ { \mathrm { intrinsic } } ( 1.2 - 50 ) in the hours preceding the ejection and the time averaged L _ { \mathrm { intrinsic } } ( 1.2 - 50 ) during the flare rise .
Furthermore , the total time averaged power during the ejection ( Q + the time average of L _ { \mathrm { intrinsic } } ( 1.2 - 50 ) during ejection ) is strongly correlated with L _ { \mathrm { intrinsic } } ( 1.2 - 50 ) just before launch with near equality if the distance to the source is \approx 10.5 kpc .