next previous
Up: The stellar winds of


5 Conclusions

A fraction of the winds from the few dozen early-type stars in the central parsec of the GC is trapped by the potential well of Sgr A* but is not currently being accreted by the blackhole. Unlike an isolated cluster, this gas has insufficient ram pressure or thermal pressure to escape so that gas has been accumulating in the central $R_{\rm A} \sim 1\hbox{$^{\prime\prime}$ }$ of the Galaxy. Independent of the details of the flow, the estimated density of this gas is inconsistent with IR and X-ray observations unless the accumulation has been proceeding for ${\mathrel{\mathchoice {\vcenter{\offinterlineskip\halign{\hfil
$\displaystyle ... yrs. This implies the gas flows in the central ${\sim} 1\hbox{$^{\prime\prime}$ }$ are not in equilibrium and the present low mass accretion rate onto Sgr A* is a short-lived phase.

Due to the tendency to form more high mass stars when deep in a potential well, as well as the early-type stars from the present mini-starburst, the SN rate in the GC is thought to be ${\mathrel{\mathchoice {\vcenter{\offinterlineskip\halign{\hfil
$\displaystyle ... yr-1 (Tamblyn & Rieke 1993). In addition, starbursts may recur in the GC every ${\sim} 10^{7-8}$ yrs (Ozernoy 1994; Morris et al. 1999). If so, within each cycle of star formation, the sub-cycle of "binge and purge'' may repeat on a time-scale of ${\mathrel{\mathchoice {\vcenter{\offinterlineskip\halign{\hfil
$\displaystyle ... yrs. This episodic behaviour should appear in any galaxy with star formation near a central supermassive blackhole. Similar feedback mechanisms have been proposed for elliptical galaxies (Ciotti & Ostriker 2001).

In the case of the GC, one would then expect that 1) the 90 cm emission near Sgr A* to visibily brighten over the next few decades as the shell of the SNR moves completely in front of Sgr A West, 2) the emission from Sgr A* itself should increase over the next ${\mathrel{\mathchoice {\vcenter{\offinterlineskip\halign{\hfil
$\displaystyle ... yrs at all frequencies as the cluster's winds again accumulate, and 3) a SN should detonate in the next ${\mathrel{\mathchoice {\vcenter{\offinterlineskip\halign{\hfil
$\displaystyle ... yrs, leading to a brief ( ${\sim}10^3$ yrs) period of high accretion and the start of the next ``binge and purge'' cycle.

Acknowledgements
This work was supported by PPARC and has made use of NASA's Astrophysics Data System Abstract Service. I would like to thank J. M. Pittard for many useful discussions and S. Markoff and B. A. Cohen for comments on the manuscript.


next previous
Up: The stellar winds of

Copyright ESO 2001