Stability and fluctuations in black hole thermodynamics

I examine thermodynamic fluctuations for a Kerr-Newman black hole in an extensive, infinite environment. This problem is not strictly solvable because full equilibrium with such an environment cannot be achieved by any black hole with mass M, angular momentum J, and charge Q. However, if we consider one (or two) of M, J, or Q to vary so slowly compared with the others that we can regard it as fixed, instances of stability occur, and thermodynamic fluctuation theory could plausibly apply. I examine seven cases with one, two, or three independent fluctuating variables. No knowledge about the thermodynamic behavior of the environment is needed. The thermodynamics of the black hole is sufficient. Let the fluctuation moment for a thermodynamic quantity X be root <(Delta X)(2)>. Fluctuations at fixed M are stable for all thermodynamic states, including that of a nonrotating and uncharged environment, corresponding to average values J=Q=0. Here, the fluctuation moments for J and Q take on maximum values. That for J is proportional to M. For the Planck mass it is 0.3990h. That for Q is 3.301e, independent of M. In all cases, fluctuation moments for M, J, and Q go to zero at the limit of the physical regime, where the temperature goes to zero. With M fluctuating there are no stable cases for average J=Q=0. But, there are transitions to stability marked by infinite fluctuations. For purely M fluctuations, this coincides with a curve which Davies identified as a phase transition.