The analogue-gravity bench

Why the library suggests it

Push the flow past the speed of sound in the medium and a horizon appears exactly where the two cross: upstream of it a wave still climbs out, past it the wave cannot.

The analogue horizon

Chill eight thousand rubidium atoms into a single quantum wave, then drive them over a step of laser light so that downstream they flow faster than their own speed of sound. A sound wave upstream of that crossing can still climb out. Past it, nothing can. Move the flow and watch the horizon appear, walk upstream, and finally swallow the whole channel.

Flow multiplier

Flow multiplier

1.00

Return to the measured flow

• Flow speed
• Speed of sound in the condensate

Horizon position:

-0.48 micrometres

Gradient of the escape speed at the horizon:

288 per second

Hawking temperature this implies:

0.35 nanokelvin

The same temperature in the paper's own units:

0.12

Waves that still climb out:

3 / 7

Waves the flow carries away:

4 / 7

What the channel is doing: There is a horizon. Upstream of it a sound wave still climbs out; downstream the flow outruns sound and nothing escapes, however fast it swims. This is the case the measurement reports, at a multiplier of one.

At the measured flow this returns 0.35 nanokelvin, which is the temperature the experiment measured over 7,400 repetitions, with the spectrum matching the thermal prediction and no free parameters. (8,000)

What a fluid reproduces here is the kinematics: the metric that sound waves feel, horizon and all. Deriving Einstein's equations from the same fluid — the rule that says how a metric responds to energy — is the programme's open problem, and this instrument does not touch it.

• The source sheet: the measured Hawking temperature of an analogue black hole
• The first observation, and the entanglement of the pairs
• The standard review of curved-spacetime physics in fluids

Carlos Barceló, Stefano Liberati and Matt Visser survey the whole programme and its catalogue of analogues in condensates, superfluid helium, slow light, optical fibres and water tanks, and close on the argument that matters most here — that spacetime geometry may be emergent, the long-wavelength limit of a medium, exactly as fluid mechanics is the long-wavelength limit of molecules (Analogue Gravity, 2011, with the 2026 edition of Analogue Gravity published in 2026). Jeff Steinhauer swept a step of laser light along a rubidium condensate, measured a band of correlations linking points across the horizon, a Hawking temperature near 1.2 nanokelvin, and genuine quantum entanglement between the escaping particle and its partner (Observation of quantum Hawking radiation and its entanglement in an analogue black hole, 2016). And the newest result closes the energy books: in an optical fibre horizon made of light, one pair-creating term turns pump photons into a Hawking pair, and the kick back onto the pump grows as the square of the probe power while both carry the same temperature to within two per cent (Backreaction of stimulated Hawking radiation in an optical analogue, 2026). The programme this feeds is Volovik's — the vacuum sharing the topology of superfluid helium-3, with particles and gravity emerging as low-energy excitations (The Universe in a Helium Droplet, 2003).

The experiment or build

The under-used move is the engineering one rather than the physics one. Analogue systems are almost always run to observe an effect; nobody runs them to design a metric. Take the optical-fibre horizon, where the metric is set by a pulse shape you control completely, and program a target metric — a warp-like profile from the Warp Factory sweep of section 3 — into the pump pulse. Then measure what the analogue field does inside it. The settling measurement is the transit time and the emission spectrum of a probe through a programmed analogue metric, compared against the numerical prediction for that same metric, across a family of profiles. That is metric engineering with a real feedback loop, at a cost of one fibre laser.