GW170608Observation of a 19 Solar-mass Binary Black Hole Coalescence

Two laser interferometers — one in Livingston, Louisiana and one in Hanford, Washington — each stretching four kilometers along two arms at a right angle, register a disturbance smaller than a thousandth the width of a proton. Two black holes, totaling just nineteen solar masses, have collided roughly a billion light-years away. The number is the point. Every binary black hole merger that LIGO had observed before this one was heavier. This was something new at the low end of the mass scale, and it would open a comparison that astronomers had never been able to make before. The signal was first spotted in data from the LIGO Livingston Observatory as a loud single-detector event, with a signal-to-noise ratio of about nine, flagged by a low-latency matched-filter search. A matched-filter search works by sliding theoretical waveform templates — predictions of what gravitational waves from merging black holes should look like — across the incoming data stream and listening for agreement. At the time, the Hanford detector was running stably but undergoing a routine angular-control procedure, so its data weren't being analyzed automatically. Researchers went back and ran a targeted matched-filter search on Hanford data starting at thirty hertz, found a consistent counterpart, and combined the two detectors to get a network signal-to-noise ratio of thirteen. Two independent matched-filter pipelines both confirmed the event. One estimated the background noise rate using time-shifted data and found that a noise event ranked this highly would occur less than once in three thousand years. The other pipeline put that number at once in one hundred sixty thousand years. An unmodeled coherent search — one that doesn't use templates at all — also found the signal, but with lower significance, about once in thirty years. That's expected: template-free searches are less sensitive to lower-mass binaries, which spend more time in the detector band and match templates well. Thirteen and a half hours after the event, an alert went out to electromagnetic observing partners with a sky localization covering eight hundred sixty square degrees. Now, about that angular-control procedure happening at Hanford during the event. This matters because you need to trust the detector. The procedure — called angular coupling minimization — periodically adjusts how the four suspended cavity mirrors are controlled, to prevent tiny angular motions from leaking into the differential arm length measurement that carries the gravitational-wave signal. It works by injecting high-amplitude pitch and yaw excitations into eight angular degrees of freedom at frequencies spaced between roughly nineteen and twenty-three hertz, then measuring how strongly each degree of freedom couples into the length signal. The feed-forward gains are stepped roughly every forty-five seconds to find the global minimum coupling. Those excitations do inject excess power into the strain data in the nineteen-to-twenty-three hertz band — you can see it in the spectrogram. But Abbott and colleagues show amplitude spectral densities computed from two hundred forty seconds of data before and during the excitations, and above thirty hertz there is no visible effect on the spectrum. The nearest abrupt gain step occurred ten seconds before the merger, and inspection showed no transient excess noise outside that low-frequency band. Imposing the thirty hertz lower cutoff on the Hanford analysis was expected to cost about one percent of signal-to-noise or less. The signal was real. The calibration underpinning all of this used modulated auxiliary lasers to push the suspended test masses with photon pressure, translating known forces into known length changes. Over the frequency range twenty to one thousand twenty-four hertz, maximum one-sigma calibration uncertainties were five percent in amplitude and three degrees in phase. Those uncertainties were folded into the parameter estimation. Which brings us to what the signal actually says about the two black holes that made it. Abbott and colleagues ran a coherent Bayesian analysis: compare the recorded strain against many model waveforms, and build probability distributions for every physical parameter consistent with the data. Two independent waveform models were used and their results averaged. One model includes spin-induced orbital precession; the other assumes spins aligned with the orbital axis. The headline numbers: a primary black hole mass of twelve solar masses, with a ninety percent credible interval running from ten to nineteen, and a secondary of seven solar masses, with an interval from five to nine. Total source-frame mass near nineteen solar masses. The best-measured single quantity is the chirp mass — seven point nine solar masses, with an uncertainty of just zero point two solar masses at ninety percent confidence. The chirp mass is the combination of the two component masses that most directly controls how fast the binary's orbital frequency accelerates during inspiral: take the product of the two masses raised to the three-fifths power, divide by their sum raised to the one-fifth power. That particular combination is essentially what the gravitational-wave phase is measuring, which is why it's so tightly constrained even when the individual masses are not. The source sits at a luminosity distance of three hundred forty megaparsecs — plus or minus one hundred forty — corresponding to a redshift of about zero point zero seven. Roughly a billion light-years, give or take. Spins are harder to read. They enter the inspiral primarily through a single mass-weighted combination called the effective inspiral spin, chi-eff. Abbott and colleagues report chi-eff of negative zero point zero seven, with a ninety percent credible interval from negative zero point two three to positive zero point zero nine. That posterior disfavors large spins strongly anti-aligned with the orbital angular momentum but is otherwise consistent with small spins in either direction. Information about spin components in the orbital plane — the precessing part — is essentially absent: the relevant parameter's posterior is dominated by the prior, meaning the data add almost nothing there. The primary black hole's dimensionless spin magnitude is constrained to be less than zero point seven five at ninety percent confidence, and that limit holds across different spin priors and waveform models. The collaboration also ran standard tests of general relativity on the waveform. The approach is to check whether the inspiral phase, the merger, and the ringdown are mutually consistent and whether the measured inspiral phasing coefficients agree with general relativity's predictions. Because GW170608 is a low-mass binary, it spends more cycles in the sensitive detector band during inspiral than a heavier system would — providing more signal to test the theory against. Tests involving the merger and ringdown parameters turned out not to be informative because the merger occurs at relatively high frequencies where detector sensitivity drops. But for all tested inspiral coefficients, the general relativity-predicted values fall within the ninety percent credible intervals of the posteriors. No deviation found. The collaboration also checked for a non-zero graviton mass and obtained an upper bound comparable to previous LIGO-Virgo results. Here is where GW170608's low mass becomes scientifically distinctive beyond just being a record. The component masses of twelve and seven solar masses fall directly in the range of black holes measured in low-mass X-ray binaries — systems where a black hole is pulling material from a stellar companion, producing X-rays that allow dynamical mass measurements. Those electromagnetically measured black holes typically come in below about ten solar masses. GW170608 sits right in that neighborhood. That means, for the first time, we can meaningfully compare black holes detected through gravitational waves with black holes detected through light. They are, in some sense, the same class of object seen through two completely different windows. Abbott and colleagues are careful about what this implies for how these black holes formed. The usual expectation is that stars born in high-metallicity environments — where metals here means anything heavier than helium — lose more mass to stellar winds and leave behind lighter remnants, while low-metallicity progenitors produce heavier black holes. GW170608's low component masses don't actually require a low-metallicity origin; they're consistent with higher-metallicity formation too. On the question of whether the binary formed through isolated binary stellar evolution or through dynamical assembly in a dense stellar cluster, the mass and spin constraints from this single event are simply insufficient to decide. Distinguishing between formation channels will likely require on the order of one hundred detections. The current binary black hole merger rate estimate — roughly twelve to two hundred thirteen mergers per cubic gigaparsec per year — remains compatible with this event. What GW170608 contributes most concretely is range. It pushes the gravitationally detected black hole mass distribution downward into territory previously accessible only through electromagnetic observations. The full observing run catalog from O2 was in preparation at the time of publication, and with increasing detector sensitivity, the paper states that binary black hole detections will become routine. When they do, the accumulating population will refine mass distributions, constrain formation channels, and build what we currently don't have: a census of stellar-mass black holes across cosmic time, assembled not from light, but from the ripples that matter makes in space itself. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.