Bright Sirens: Gravitational-Wave Cosmology

• Bright sirens are gravitational-wave events with directly measured redshifts from electromagnetic counterparts, providing a clear probe of the distance–redshift relation.
• They yield self-calibrated luminosity distances that bypass the cosmic distance ladder, enabling precise measurements of the Hubble constant and dark-energy dynamics.
• Observational challenges such as event rarity, counterpart detection limits, and lensing effects underscore the need for advanced detectors and comprehensive follow-up strategies.

Bright sirens are gravitational-wave standard sirens for which a source redshift is obtained through electromagnetic information, typically by identifying an electromagnetic counterpart and the host galaxy, so that the gravitational-wave measurement of luminosity distance can be paired directly with a redshift to constrain the distance–redshift relation and late-time cosmology. In current usage, the term most commonly refers to binary neutron star and neutron star–black-hole mergers with counterparts such as kilonovae, short gamma-ray bursts, or gamma-ray burst afterglows, but the same logic also appears in forecasts for massive and supermassive black hole binaries when host identification or an already known candidate host/redshift is available (Menote et al., 21 Oct 2025). Because the distance scale is encoded in the gravitational-wave waveform itself, bright sirens are self-calibrated and independent of the cosmic distance ladder; this makes them central to measurements of $H_0$, dark-energy parameters, and modified gravitational-wave propagation (Matos, 2024).

1. Definition and taxonomic scope

A bright siren is a gravitational-wave event for which the host galaxy can be identified directly and its redshift measured individually. In the compact-binary literature, this usually means a binary neutron star or neutron-star–black-hole merger with an identified electromagnetic counterpart, most importantly a kilonova or a short gamma-ray burst (Matos, 2024). In studies of future spectroscopic infrastructure, the defining feature is not the transient type itself but the fact that an electromagnetic counterpart allows direct host-galaxy identification and therefore direct redshift measurement, in contrast to statistical host assignment in dark-siren analyses (Borghi et al., 20 Dec 2025).

The contrast with dark sirens is methodological rather than ontological. Bright sirens provide a direct distance–redshift pair for each event. Dark sirens instead infer redshift statistically by associating the gravitational-wave localization volume with galaxies in a catalog. Several papers emphasize that bright sirens are the cleanest route to standard-siren cosmology, whereas dark sirens are more numerous but require catalog completeness and host-probability modeling (Matos, 2024). One proceeding states that the current ratio is roughly one bright siren per hundred dark sirens, capturing the basic tradeoff between directness and event rate (Matos, 2024).

The source classes discussed under the bright-siren label are broader than low-mass mergers alone. Ground-based forecasts focus on binary neutron star and neutron-star–black-hole systems (Menote et al., 21 Oct 2025), space-based forecasts consider massive black hole binaries in gas-rich environments as candidate bright sirens for LISA-like missions (Afroz et al., 2024), and pulsar-timing-array forecasts treat individually resolved inspiraling supermassive black hole binaries with an identified electromagnetic counterpart or an already known candidate host/redshift as bright sirens in the nanohertz regime (Wang et al., 2022). A plausible implication is that “bright siren” is best understood as a redshift-acquisition category that cuts across detector bands and source populations, rather than as a synonym for any single astrophysical channel.

2. Measurement principle and distance–redshift inference

The physical basis of bright-siren cosmology is that the gravitational-wave signal determines an absolute luminosity distance, while electromagnetic information supplies the redshift. In compact-binary forecasts this is often written through the standard luminosity-distance relation,

$$d_L(z)=(1+z)\,d_M(z), \qquad d_M(z)=\int_0^z \frac{c\,dz'}{H(z')},$$

or, for specific background models, through flat $w$CDM and $w_0w_a$CDM parameterizations (Menote et al., 21 Oct 2025). In PTA work, the same logic is expressed in terms of the inspiral-wave amplitude and the FLRW distance–redshift relation in flat $w$CDM, with cosmological inference proceeding from the pair $(d_{\rm L},z)$ (Wang et al., 2022).

The “bright” designation matters because redshift is not otherwise obtained cleanly from the gravitational-wave waveform. Multiple papers stress the mass–redshift degeneracy: the waveform measures redshifted masses, so an external redshift is needed unless an alternative strategy such as tidal-information inference is used (Canevarolo et al., 2023). Bright sirens circumvent this by providing a spectroscopic host redshift through counterpart identification (Borghi et al., 20 Dec 2025).

Forecasts typically implement the inference using a Gaussian likelihood in luminosity distance. One representative form is

$$\ln\mathcal L_{\rm GW} =-\frac12\sum_i \left[\frac{d_L(z_i;\theta_{\rm cosmo})-d_{L,i}^{\rm GW}}{\sigma_{d_L,i}}\right]^2,$$

with the combined gravitational-wave-plus-supernova likelihood taken as the product of the two independent likelihoods when Roman-like Type Ia supernovae are included (Menote et al., 21 Oct 2025). In PTA analyses, the luminosity-distance uncertainty is estimated from a Fisher matrix, and the total error budget combines instrumental, lensing, and peculiar-velocity terms (Wang et al., 2022).

This direct distance–redshift construction is why bright sirens are often described as the multi-messenger analogue of a standard candle, but the analogy is limited. The calibration enters through waveform physics rather than through an empirical luminosity relation. That distinction underlies the repeated claim that bright sirens provide a self-calibrated cosmological probe independent of the cosmic distance ladder (Menote et al., 21 Oct 2025).

3. Observational channels and detector regimes

For ground-based detectors, the dominant bright-siren channel is the compact-binary merger with a kilonova or short gamma-ray burst. In third-generation forecasts using the CosmoDC2_BCO LSST synthetic catalog, the source population consists of binary neutron star and neutron-star–black-hole mergers with electromagnetic counterparts over an LSST-like southern-sky footprint of roughly $16{,}000\,\mathrm{deg}^2$, and the detector configurations compared are LVK, ET+LVK, and CE+ET+LVK, all assumed to operate with a 70% duty cycle (Menote et al., 21 Oct 2025). A separate 2.5-generation study for Voyager and NEMO takes binary neutron star mergers with identified short gamma-ray burst counterparts as bright sirens, using a THESEUS-like mission for counterpart selection (Jin et al., 2023).

The counterpart type affects both redshift reach and event yield. One proceeding emphasizes that gamma-ray bursts can be seen to higher redshift because they are brighter, but they are detected less often because the jet must be oriented close to the line of sight, whereas kilonovae are more isotropic but visible to smaller distances (Matos, 2024). In binary neutron star forecasts comparing counterpart-based and counterpart-less approaches, kilonovae are taken to be detectable to about $z \simeq 0.5$, and only 10% of gravitational-wave events within that range are assumed to be followed up electromagnetically (Dhani et al., 2022). Studies of optical follow-up with the 2.5-meter Wide-Field Survey Telescope similarly treat kilonova discovery efficiency as the limiting factor for bright-siren yield in the 2G era (Yu et al., 2023).

In the millihertz band, bright sirens are forecast for massive black hole binaries observed by networks involving Taiji, TianQin, and LISA. For the Taiji–TianQin–LISA network, bright sirens are mergers with an identifiable electromagnetic counterpart whose redshift is measured directly from follow-up observations; expected bright-siren counts over 5 years are 49 for pop III, 18 for Q3d, and 43 for Q3nod under the adopted criteria (Jin et al., 2023). LISA-only bright-siren studies focus on massive binary black holes with assumed spectroscopic redshifts from electromagnetic counterparts and use them for model-independent tests of gravitational-wave propagation over $z=1$ to $$d_L(z)=(1+z)\,d_M(z), \qquad d_M(z)=\int_0^z \frac{c\,dz'}{H(z')},$$0 (Afroz et al., 2024).

At still lower frequencies, pulsar timing arrays admit a conceptually distinct bright-siren channel. For individually resolved inspiraling supermassive black hole binaries, a PTA-detected source becomes a bright siren if there is an identified electromagnetic counterpart or an already known candidate host/redshift. In one SKA-era forecast, the practical source pool is 154 existing supermassive black hole binary candidates with measured redshifts (Wang et al., 2022). This broadens the bright-siren category beyond merger transients and shows that host identification can arise either from contemporaneous multimessenger emission or from prior electromagnetic characterization.

4. Cosmological applications

The principal cosmological role of bright sirens is the determination of the Hubble constant and the late-time expansion history. In the low-redshift cosmography analysis of the CosmoDC2_BCO forecast, after ten years the quoted 68% uncertainties on $$d_L(z)=(1+z)\,d_M(z), \qquad d_M(z)=\int_0^z \frac{c\,dz'}{H(z')},$$1 are $$d_L(z)=(1+z)\,d_M(z), \qquad d_M(z)=\int_0^z \frac{c\,dz'}{H(z')},$$2 for LVK, $$d_L(z)=(1+z)\,d_M(z), \qquad d_M(z)=\int_0^z \frac{c\,dz'}{H(z')},$$3 for ET+LVK, and $$d_L(z)=(1+z)\,d_M(z), \qquad d_M(z)=\int_0^z \frac{c\,dz'}{H(z')},$$4 for CE+ET+LVK, corresponding to about 5.6%, 0.32%, and 0.21% precision from bright sirens alone (Menote et al., 21 Oct 2025). The same study reports that CE+ET+LVK reaches about 0.6% on $$d_L(z)=(1+z)\,d_M(z), \qquad d_M(z)=\int_0^z \frac{c\,dz'}{H(z')},$$5 within two years and about 0.2% after a decade, while LVK remains around the 6% level after a decade (Menote et al., 21 Oct 2025).

Bright sirens are also used to constrain dark-energy dynamics. In the joint analysis of Roman-like supernovae and bright sirens, the Figure of Merit in the $$d_L(z)=(1+z)\,d_M(z), \qquad d_M(z)=\int_0^z \frac{c\,dz'}{H(z')},$$6–$$d_L(z)=(1+z)\,d_M(z), \qquad d_M(z)=\int_0^z \frac{c\,dz'}{H(z')},$$7 plane reaches 24.9 for ET+LVK + Roman and 76.1 for CE+ET+LVK + Roman, compared in that work to 55.9 for DESI DR2 BAO + DESY5 supernovae (Menote et al., 21 Oct 2025). A separate study of future ET and CE bright sirens over five years with a 75% duty cycle considers three model classes—Barboza–Alcaniz phenomenology, hilltop quintessence, and evolving dark matter with density scaling as $$d_L(z)=(1+z)\,d_M(z), \qquad d_M(z)=\int_0^z \frac{c\,dz'}{H(z')},$$8—and concludes that bright sirens alone can yield competitive and independent constraints on the time evolution of dark energy (Afroz et al., 8 Jul 2025).

The complementarity with external probes is a recurring theme. PTA bright sirens are presented as especially valuable because they probe late-Universe distances directly, with a degeneracy direction in the $$d_L(z)=(1+z)\,d_M(z), \qquad d_M(z)=\int_0^z \frac{c\,dz'}{H(z')},$$9–$w$0 plane different from that of the CMB (Wang et al., 2022). For an SKA-era PTA with $w$1 and $w$2 ns, the combined CMB+PTA analysis gives $w$3, with corresponding relative errors on $w$4 and $w$5 of about 1.7% and about 3.1% in the best case shown (Wang et al., 2022). Likewise, the Taiji–TianQin–LISA network forecast finds $w$6 in $w$7CDM when CMB data are combined with bright sirens in the pop III scenario (Jin et al., 2023).

Bright sirens can also calibrate supernova distances rather than merely supplement them. In the Roman-like supernova joint fits, the supernova absolute magnitude $w$8 is recovered with substantially improved precision: $w$9 for LVK+Roman, $w_0w_a$0 for ET+LVK+Roman, and $w_0w_a$1 for CE+ET+LVK+Roman (Menote et al., 21 Oct 2025). This places bright sirens in a dual role as both direct distance indicators and anchors for other distance probes.

5. Tests of gravity and cosmic structure

Bright sirens are not limited to background-expansion measurements. A major line of work treats them as probes of modified gravitational-wave propagation. A compact expression used across several papers is

$w_0w_a$2

where $w_0w_a$3 is the gravitational-wave distance and $w_0w_a$4 is the electromagnetic luminosity distance (Matos, 2024). In model-independent reconstructions, this mismatch is encoded by

$w_0w_a$5

with $w_0w_a$6 in GR (Afroz et al., 2023). By combining bright sirens with BAO angular scales and the CMB sound horizon, one study reconstructs deviations in the effective Planck-mass variation as a function of redshift and reports a precision of approximately 7.9% for binary neutron stars at $w_0w_a$7 and 10% for neutron star–black-hole systems at $w_0w_a$8 with 5 years of LIGO–Virgo–KAGRA observations; CE+ET for 1 year improves this to about 1.62% for binary neutron stars and 2% for neutron star–black-hole systems at $w_0w_a$9 (Afroz et al., 2023).

In Horndeski scalar–tensor forecasts, bright sirens with gamma-ray bursts are used to constrain the Planck-mass run rate $w$0 and the tensor-speed excess $w$1. The analysis highlights the $w$2–$w$3 degeneracy, the distance–inclination degeneracy, and the role of GRB arrival-time information for constraining $w$4 (Colangeli et al., 9 Jan 2025). That study concludes that the next ten bright sirens alone will not competitively constrain cosmological gravity, but that one year of third-generation observations could detect mild departures from GR at greater than $w$5 for the fiducial non-GR injection used there (Colangeli et al., 9 Jan 2025).

Another extension uses weak lensing not only as noise but as signal. For bright standard sirens with electromagnetic counterparts, the observed luminosity distance is written as

$w$6

with $w$7 determined by convergence and shear (Vaskonen, 9 Jan 2026). Modeling the full non-Gaussian magnification probability distribution, that study forecasts that $w$8 could be measured to about 10% with ET from 300 binary neutron star bright sirens and to about 30% with LISA from 12 massive black hole binary bright sirens, improving to about 8% when ET and LISA are combined (Vaskonen, 9 Jan 2026). This suggests that bright sirens can probe both geometry and structure formation when lensing is modeled as cosmological information rather than treated only as a contaminant.

6. Limitations, systematics, and enabling infrastructure

The dominant limitation of bright sirens is not the clarity of the method but the scarcity and observational difficulty of the events. Bright sirens are rare because kilonovae at third-generation detector distances are faint, reaching only about $w$9, they fade within days, and gamma-ray burst afterglows require a favorable viewing geometry (Borghi et al., 20 Dec 2025). Even where detector sensitivity is high, the rate of usable bright sirens depends strongly on counterpart detectability, duty cycle, sky localization, and follow-up strategy (Menote et al., 21 Oct 2025).

Redshift quality is a second major limitation. Forecasts for 2040s infrastructure state that photometric uncertainties degrade cosmological constraints by up to an order of magnitude compared to spectroscopic ones, with representative uncertainties

$(d_{\rm L},z)$0

for spectroscopy and

$(d_{\rm L},z)$1

for photometry (Borghi et al., 20 Dec 2025). For 100 well-localized BBH events at LVK O5-like sensitivity, spectroscopic redshifts permit percent-level $(d_{\rm L},z)$2 precision, while photometric redshifts give only 9% precision; the same study states that photometric uncertainties degrade $(d_{\rm L},z)$3 constraints by up to a factor of $(d_{\rm L},z)$4 (Borghi et al., 20 Dec 2025). This is why wide-field, high-multiplex spectroscopy is presented as essential infrastructure for bright-siren cosmology in the 2040s (Borghi et al., 20 Dec 2025).

Weak lensing is both a stochastic error source and a systematic effect. In the delensing study for supermassive black hole binary bright sirens, realistic shear-based convergence reconstruction reduces the weak-lensing error at $(d_{\rm L},z)$5 by only about a factor of two on average in the most favorable hybrid deep+wide setup, and the authors conclude that performing delensing corrections is unlikely to be worthwhile under the conditions studied (Wu et al., 2022). A separate Einstein Telescope analysis treats lensing as a source of residual bias on cosmological parameters and finds that, for a fiducial scenario with 3000 bright sirens, the lensing bias can be comparable to or greater than the expected statistical uncertainty, especially for $(d_{\rm L},z)$6, with selection effects from magnification dominating the bias budget (Canevarolo et al., 2023). A plausible implication is that future bright-siren analyses require explicit treatment of lensing selection and magnification PDFs, not merely quadrature error inflation.

Sky localization and follow-up thresholds introduce additional practical selection effects. In the CosmoDC2_BCO forecast, cuts of $(d_{\rm L},z)$7 or even $(d_{\rm L},z)$8 entail only mild penalties for third-generation networks, and sub-percent $(d_{\rm L},z)$9 precision is still achieved within about five years for ET+LVK and CE+ET+LVK (Menote et al., 21 Oct 2025). This has led to the suggestion that prioritizing events localized within $$\ln\mathcal L_{\rm GW} =-\frac12\sum_i \left[\frac{d_L(z_i;\theta_{\rm cosmo})-d_{L,i}^{\rm GW}}{\sigma_{d_L,i}}\right]^2,$$0, or more aggressively $$\ln\mathcal L_{\rm GW} =-\frac12\sum_i \left[\frac{d_L(z_i;\theta_{\rm cosmo})-d_{L,i}^{\rm GW}}{\sigma_{d_L,i}}\right]^2,$$1, is an efficient follow-up strategy (Menote et al., 21 Oct 2025). In an unusual extension of the concept, one space-based network study argues that ultra-high-precision localization by LISA–Taiji–TianQin can permit unique host identification solely from gravitational-wave signals, yielding what it calls “bright sirens without EM counterparts” when $$\ln\mathcal L_{\rm GW} =-\frac12\sum_i \left[\frac{d_L(z_i;\theta_{\rm cosmo})-d_{L,i}^{\rm GW}}{\sigma_{d_L,i}}\right]^2,$$2 in the localization volume (Zhan et al., 4 Sep 2025). This suggests that the operational boundary between bright and dark sirens may eventually be shaped as much by host-identification capability as by the presence of a transient.

Across the literature, bright sirens remain the benchmark standard-siren channel because they provide the cleanest redshift information and the most direct route from waveform amplitude to cosmology. Their current limitations are event rarity, redshift acquisition, lensing, and follow-up logistics rather than any ambiguity in the underlying inference framework (Borghi et al., 20 Dec 2025).