Did GW150914 produce a rotating gravastar?

The interferometric LIGO detectors have recently measured the first direct gravitational-wave signal from what has been interpreted as the inspiral, merger and ringdown of a binary system of black holes. The signal-to-noise ratio of the measured signal is large enough to leave little doubt that it does refer to the inspiral of two massive and ultracompact objects, whose merger yields a rotating black hole. Yet, the quality of the data is such that some room is left for alternative interpretations that do not involve black holes, but other objects that, within classical general relativity, can be equally massive and compact, namely, gravastars. We here consider the hypothesis that the merging objects were indeed gravastars and explore whether the merged object could therefore be not a black hole but a rotating gravastar. After comparing the real and imaginary parts of the ringdown signal of GW150914 with the corresponding quantities for a variety of gravastars, and notwithstanding the very limited knowledge of the perturbative response of rotating gravastars, we conclude it is not possible to model the measured ringdown of GW150914 as due to a rotating gravastar.

Figure 1

Evolution of the $\ell =2=m$ axial perturbation $\psi (t)$ for a Schwarzschild black hole (black dashed line) and for three different gravastars (colored solid lines) with decreasing thickness $\delta /M=0.01$, 0.005 and 0.0025; in all cases the gravastars have the same compactness $\mu =0.48$ and the time is retarded with a tortoise coordinate $r_{*}$ [12]. Note that even for $\delta /M=0.0025$ the QNMs can be distinguished clearly over a dynamical time scale.

Figure 2

Real and imaginary parts ${\sigma }_r$ and ${\sigma }_i$, respectively, of the eigenfrequencies as computed for the $\ell =2=m$ axial QNM of nonrotating gravastars [12]. Lines of different colors refer to sequences of gravastars with constant compactness $\mu$ but varying thickness $\delta$. Thick gravastars have small eigenfrequencies which increase as the gravastars become increasingly thin. Note however that the eigenfrequencies remain distinct from that of a Schwarzschild black hole (solid red circle).

Figure 3

The same as Fig. 2 but including now also eigenfrequencies of rotating gravastars, as well as the eigenfrequencies of rotating black holes (red dashed line). Because of rotation the sequences of constant compactness no longer overlap and the shaded area bounds the space of eigenfrequencies span by rotating gravastars compatible with GW150914. This region does not overlap by the range span by the eigenfrequencies of a Kerr black hole with dimensionless spin $a=0.68_{-0.06}^{+0.05}$ (thick red solid line).

Figure 4

The same as in Fig. 3 but including also the 90% confidence regions for the GW150914 ringdown frequencies, obtained with different starting times $t_0=t_M+1,3,5,7\mathrm{ms}$ after the merger time $t_M$ [18]. The solid black line shows the 90% confidence limit for the best fit combining the inspiral-merger-ringdown (IMR). Note that the Kerr eigenfrequencies are inside the IMR region and that there is no overlap between the contours and the gravastar eigenfrequencies.