Active Galactic Nuclei
Reverberation Mapping: Timing Light Echoes to Weigh Black Holes
Point a telescope at a distant quasar for a hundred nights and you'll catch its glare flickering — and then, a week or two later, its broad hydrogen emission lines flicker in exactly the same pattern. That delay, sometimes just a few light-days, is a light echo: photons from the black hole's inner accretion disk racing outward, striking clouds of gas, and lighting them up on a stopwatch you can read from Earth.
Reverberation mapping exploits this echo to measure the size of the broad-line region (BLR) around a supermassive black hole — and from that size plus the gas's orbital speed, to weigh the black hole itself. It is the only technique that resolves the gravitational sphere of influence of black holes millions to billions of light-years away without ever spatially resolving them, turning a time delay into a mass in solar units.
- TypeBlack-hole mass measurement technique
- RegimeActive galactic nuclei / quasars
- ProposedBlandford & McKee 1982
- Typical BLR scaleLight-days to light-months (few to ~100 ld)
- Key equationM = f·c·τ·(ΔV)²/G
- Observed in~200+ AGN (NGC 5548, Mrk 817, NGC 4151...)
Interactive visualization
Press play, or step through manually. The visualization is yours to drive — try it before reading on.
Watch the 60-second explainer
A condensed visual walkthrough — narrated, captioned, under a minute.
What It Is and Why the Echo Works
An active galactic nucleus (AGN) is a supermassive black hole devouring gas through a hot accretion disk that radiates a blaze of ultraviolet and optical continuum light. Surrounding it, a few light-days to light-months out, orbits the broad-line region: dense clouds of gas moving fast enough (thousands of km/s) that their emission lines are Doppler-broadened into wide humps.
The trick is that the continuum source is compact and variable — it brightens and dims on timescales of days. The BLR gas is photoionized by that continuum, so when the central source flares, the broad lines respond with a delay equal to the light-travel time across the region. Measure that delay τ and you have the BLR radius, R = c·τ, without ever resolving it on the sky. This is why the method is sometimes called echo mapping: you are timing a reverberation, exactly as a bat maps a cave with sound echoes, but with light and gravity instead.
From Lag to Mass: The Virial Derivation
The BLR clouds orbit in the black hole's gravity, so to good approximation they are virialized: the gas velocity balances gravity, GM/R ≈ V². Rearranging gives the black-hole mass directly:
M = f · c·τ · (ΔV)² / G
- c·τ is the BLR radius R from the measured continuum-to-line lag τ.
- ΔV is the line-of-sight velocity from the broad-line width — either the FWHM or the line dispersion σ_line.
- f is the dimensionless virial factor, order unity, that absorbs the unknown geometry, inclination, and kinematics of the BLR (a thin disk seen face-on gives very different projected velocities than a spherical shell).
The quantity c·τ·(ΔV)²/G is the virial product, purely observational. All the astrophysical ignorance is bundled into f, typically calibrated so that reverberation masses match the M-sigma relation of quiescent galaxies, giving a population average of roughly f ≈ 4-5 (for σ_line) with object-to-object scatter of 0.3-0.4 dex.
Characteristic Numbers and a Worked Example
Take NGC 5548, the most-monitored Seyfert galaxy, watched almost every year since Bradley Peterson's campaigns began in 1989. Its Hβ lag is roughly τ ≈ 6-20 light-days depending on the epoch (the BLR breathes in and out as luminosity changes), so R ≈ 6 ld ≈ 1.5 × 10^14 m. The broad Hβ line has ΔV ≈ 5000 km/s.
Plugging in: M ≈ f · (c·τ)·(ΔV)²/G ≈ 1 × (1.5×10^14 m)(5×10^6 m/s)² / (6.67×10^-11) ≈ 1 × 10^38 kg ≈ 5 × 10^7 M_sun. Modern analyses put NGC 5548 near 5-7 × 10^7 solar masses.
- Lag range across AGN: ~1 light-day (low-luminosity Seyferts) to >100 light-days (luminous quasars).
- Mass range mapped: ~10^6 to ~10^9 M_sun.
- Line used: Hβ (optical, low-z), Mg II 2798 Å and C IV 1549 Å (UV, high-z quasars).
- Campaign length: months to years, cadence of days.
How It Is Observed and the R-L Relation
An observer collects a densely sampled light curve of the continuum and a parallel time series of the broad-line flux, then cross-correlates them. The peak or centroid of the cross-correlation function (the ICCF method, or Bayesian tools like JAVELIN and CREAM) gives the lag τ. Doing this line-by-line across velocity — velocity-resolved reverberation mapping — even reveals whether the gas is inflowing, outflowing, or in a rotating disk.
The landmark payoff is the radius-luminosity (R-L) relation: because a more luminous AGN ionizes gas out to larger radii, R scales as R ∝ L^0.5 from simple photoionization physics. Bentz et al. (2013), correcting for host-galaxy starlight with HST imaging, measured a slope of α = 0.533 ± 0.035 for Hβ, beautifully close to the predicted 0.5. This relation is the workhorse of the field: it lets you skip the multi-year campaign and estimate a BLR radius — hence a single-epoch black-hole mass — from one spectrum's luminosity plus line width, the method used to weigh hundreds of thousands of SDSS quasars.
How It Differs From Its Cousins
Reverberation mapping sits in a family of black-hole scales, and it is easy to conflate them:
- vs. spatially resolved dynamics: Stellar-orbit and megamaser methods actually image orbits; RM never resolves the BLR — it substitutes time for angular resolution. That is why RM reaches quasars billions of light-years away where imaging is hopeless.
- vs. the Event Horizon Telescope: The EHT images the photon ring around M87* and Sgr A* (the shadow), probing sub-light-day scales. RM probes the BLR, thousands of gravitational radii out — a completely different, larger region.
- vs. single-epoch masses: These use the RM-calibrated R-L relation but replace the measured lag with a luminosity proxy, trading accuracy for survey speed.
- vs. continuum reverberation mapping: A newer variant times lags between wavelengths of the continuum itself to map the accretion disk's temperature structure, rather than the BLR.
All virial methods ultimately trace back to RM for their f-factor calibration, making it the foundation of the whole mass ladder for accreting black holes.
Significance, Open Questions, and Famous Cases
Reverberation mapping underpins essentially every supermassive-black-hole mass beyond the local universe, and hence our maps of black-hole growth, the M-sigma coevolution of black holes and galaxies, and even attempts to use quasars as standard candles for cosmology.
The stubborn uncertainty is the virial factor f: because we cannot yet compute the BLR's geometry and inclination from first principles, the ~0.3-0.4 dex scatter in f dominates the error budget. Dynamical BLR modeling (e.g., CARAMEL) and velocity-resolved lags aim to pin f per-object rather than as a population average. A second live debate: super-Eddington AGN (SEAMBHs) show Hβ lags markedly shorter than the standard R-L relation predicts — self-shadowing by a slim disk may shrink the effective BLR, biasing masses high if uncorrected.
- NGC 5548: the canonical RM laboratory; its 2014 AGN STORM campaign found a puzzling BLR 'holiday' where lines decoupled from the continuum.
- Landmark programs: the Lick AGN Monitoring Project, SDSS-RM, OzDES-RM (Mg II lags out to z ~ 2), and AGN STORM.
| Method | What it uses | Distance reach | Typical uncertainty |
|---|---|---|---|
| Reverberation mapping | Continuum-to-line time lag + line width (virial) | Local AGN to z ~ 2 quasars | ~0.4 dex (dominated by virial factor f) |
| Single-epoch (R-L) mass | One spectrum + R-L relation as lag proxy | Any AGN with a broad line | ~0.4-0.5 dex + R-L scatter |
| Stellar / gas dynamics | Spatially resolved star or gas orbits | Nearby quiescent galaxies (<~100 Mpc) | ~0.1-0.3 dex |
| M-sigma relation | Bulge stellar velocity dispersion | Any galaxy with a measured sigma | ~0.3-0.4 dex (scatter of relation) |
| Direct imaging (EHT) | Photon ring size around the shadow | Sgr A*, M87* only | ~10-15% for M87* |
| Megamaser disk | Keplerian H2O maser rotation | A handful of edge-on disks | ~0.01-0.1 dex (most precise) |
Frequently asked questions
What is reverberation mapping in simple terms?
It is a way to measure the size of the gas region around a supermassive black hole by timing a light echo. When the black hole's accretion disk brightens, its light travels outward and, days later, makes surrounding gas clouds glow. Measuring that delay gives the region's radius, and combining it with the gas's orbital speed yields the black hole's mass.
Why can reverberation mapping weigh black holes we can't even resolve?
Because it swaps spatial resolution for time resolution. Instead of imaging the tiny broad-line region on the sky (impossible for distant AGN), it measures the light-travel delay across it. A lag of, say, ten days directly means the region is ten light-days across, no matter how far away the galaxy is.
What is the equation for the black hole mass?
M = f·c·τ·(ΔV)²/G. Here c·τ is the broad-line region radius from the measured lag τ, ΔV is the gas velocity from the broad-line width, G is Newton's constant, and f is the dimensionless virial factor (order unity) that accounts for the unknown geometry and orientation of the gas.
What is the radius-luminosity (R-L) relation and why does it matter?
It is the empirical finding that the broad-line region radius scales with AGN luminosity as roughly R ∝ L^0.5, with a measured Hβ slope of about 0.533 (Bentz et al. 2013). It matters because it lets astronomers estimate a black hole mass from a single spectrum, skipping the years-long monitoring campaign — the basis of single-epoch masses for hundreds of thousands of quasars.
What is the biggest source of uncertainty in reverberation masses?
The virial factor f. We cannot yet calculate the broad-line region's geometry, inclination, and kinematics from first principles, so f is calibrated statistically against the M-sigma relation. Its object-to-object scatter of 0.3 to 0.4 dex (a factor of two to three) dominates the mass error budget.
Which black holes have been measured this way?
Over 200 AGN, most famously the Seyfert galaxy NGC 5548, monitored yearly since 1989, plus NGC 4151, Mrk 817, and NGC 7469. Large surveys like the Lick AGN Monitoring Project, SDSS-RM, and OzDES-RM have extended it to fainter and higher-redshift quasars out to z ~ 2 using the Mg II and C IV lines.