Tests of Einstein's general theory of relativity have mostly been carried out in weak gravitational fields where the space-time curvature effects are first-order deviations from Newton's theory. Binary pulsars provide a means of probing the strong gravitational field around a neutron star, but strong-field effects may be best tested in systems containing black holes. Here we report such a test in a close binary system of two candidate black holes in the quasar OJ 287. This quasar shows quasi-periodic optical outbursts at 12-year intervals, with two outburst peaks per interval. The latest outburst occurred in September 2007, within a day of the time predicted by the binary black-hole model and general relativity. The observations confirm the binary nature of the system and also provide evidence for the loss of orbital energy in agreement (within 10 per cent) with the emission of gravitational waves from the system. In the absence of gravitational wave emission the outburst would have happened 20 days later.

OJ287 is a quasi-periodic quasar with roughly 12 year optical cycles. It displays prominent outbursts that are predictable in a binary black hole model. The model predicted a major optical outburst in 2015 December. We found that the outburst did occur within the expected time range, peaking on 2015 December 5 at magnitude 12.9 in the optical R-band. Based on Swift/XRT satellite measurements and optical polarization data, we find that it included a major thermal component. Its timing provides an accurate estimate for the spin of the primary black hole, 0.313 0.01 c =  . The present outburst also confirms the established general relativistic properties of the system such as the loss of orbital energy to gravitational radiation at the 2% accuracy level, and it opens up the possibility of testing the black hole no-hair theorem with 10% accuracy during the present decade.

Results from regular monitoring of relativistic compact binaries like PSR 1913+16 are consistent with the dominant (quadrupole) order emission of gravitational waves (GWs). We show that observations associated with the binary black hole (BBH) central engine of blazar OJ287 demand the inclusion of gravitational radiation reaction effects beyond the quadrupolar order. It turns out that even the effects of certain hereditary contributions to GW emission are required to predict impact flare timings of OJ287. We develop an approach that incorporates this effect into the BBH model for OJ287. This allows us to demonstrate an excellent agreement between the observed impact flare timings and those predicted from ten orbital cycles of the BBH central engine model. The deduced rate of orbital period decay is nine orders of magnitude higher than the observed rate in PSR 1913+16, demonstrating again the relativistic nature of OJ287ʼs central engine. Finally, we argue that precise timing of the predicted 2019 impact flare should allow a test of the celebrated black hole "no-hair theorem" at the 10% level.

We systematically surveyed period variations of superhumps in SU UMa-type dwarf novae based on newly obtained data and past publications. In many systems, the evolution of the superhump period is found to be composed of three distinct stages: an early evolutionary stage with a longer superhump period, a middle stage with systematically varying periods, and a final stage with a shorter, stable superhump period. During the middle stage, many systems with superhump periods of less than 0.08 d show positive period derivatives. We present observational characteristics of these stages and give greatly improved statistics. Contrary to an earlier claim, we found no clear evidence for a variation of period derivatives among different superoutbursts of the same object. We present an interpretation that the lengthening of the superhump period is a result of the outward propagation of an eccentricity wave, which is limited by the radius near the tidal truncation. We interpret that late-stage superhumps are rejuvenated excitation of a 3:1 resonance when superhumps in the outer disk are effectively quenched. The general behavior of the period variation, particularly in systems with short orbital periods, appears to follow a scenario proposed in Kato, Maehara, and Monard (2008, PASJ, 60, L23). We also present an observational summary of WZ Sge-type dwarf novae. Many of them have shown long-enduring superhumps during a post-superoutburst stage having longer periods than those during the main superoutburst. The period derivatives in WZ Sge-type dwarf novae are found to be strongly correlated with the fractional superhump excess, or consequently with the mass ratio. WZ Sge-type dwarf novae with a long-lasting rebrightening or with multiple rebrightenings tend to have smaller period derivatives, and are excellent candidates for those systems around or after the period minimum of evolution of cataclysmic variables.

We present HST spectroscopy for 45 cataclysmic variables (CVs), observed with HST/COS and HST/STIS. For 36 CVs, the white dwarf is recognisable through its broad Ly α absorption profile and we measure the white dwarf effective temperatures (T eff ) by fitting the HST data assuming log g = 8.35, which corresponds to the average mass for CV white dwarfs ( 0.8 M ). Our results nearly double the number of CV white dwarfs with an accurate temperature measurement. We find that CVs above the period gap have, on average, higher temperatures ( T eff 23 000 K) and exhibit much more scatter compared to those below the gap ( T eff 15 000 K). While this behaviour broadly agrees with theoretical predictions, some discrepancies are present: (i) all our new measurements above the gap are characterized by lower temperatures (T eff 16 000-26 000 K) than predicted by the present-day CV population models (T eff 38 000-43 000 K); (ii) our results below the gap are not clustered in the predicted narrow track and exhibit in particular a relatively large spread near the period minimum, which may point to some shortcomings in the CV evolutionary models. Finally, in the standard model of CV evolution, reaching the minimum period, CVs are expected to evolve back towards longer periods with mean accretion ratesṀ 2 × 10 −11 M yr −1 , corresponding to T eff 11 500 K. We do not unambiguously identify any such system in our survey, suggesting that this major component of the predicted CV population still remains elusive to observations.