Tests of General Relativity

Core Concepts

The three classic tests

Mercury's perihelion precession. In Newtonian gravity, a single planet around the Sun follows a perfect ellipse, but the other planets tug on Mercury, slowly rotating its orbit so the closest point to the Sun (the perihelion) creeps forward over time. By the late 1800s, astronomers had measured the total precession and found it was larger than Newtonian calculations predicted by about 43 arcseconds per century (a commonly taught reference value — about 1/80,000 of a full circle). GR explained the missing amount exactly: near the Sun, spacetime is curved enough that Mercury's path deviates slightly from a Newtonian ellipse. This was the first quantitative success of the theory, and it remains a textbook example of a prediction matching observation.

Deflection of starlight. Light follows the straightest possible path through curved spacetime — a geodesic. Near a massive body like the Sun, that path bends. Einstein's theory predicted starlight grazing the Sun's edge would be deflected by about 1.75 arcseconds (a commonly taught reference value), twice the deflection a Newtonian "light as particles" calculation gives. In 1919, an expedition led by Arthur Eddington photographed a total solar eclipse and measured the apparent shift of background stars near the Sun's disk; the result matched GR, and the news made Einstein world-famous. Modern radio observations of distant quasars passing behind the Sun confirm the deflection to far higher precision, and the same effect — now called gravitational lensing — is a routine tool for finding dark matter and distant galaxies.

Gravitational redshift. Light climbing out of a gravity well must spend energy to escape, and because its speed is fixed, the energy loss appears as a stretching of wavelength — a shift toward the red. In 1960 the Pound–Rebka experiment measured this effect in a 22-meter tower at Harvard using gamma rays, and today the effect is so well confirmed that it must be programmed into GPS satellite clocks. The same physics, pushed to the extreme, predicts that light emitted at the edge of a black hole never escapes at all.

Frame dragging and the geodetic effect

Einstein's equations say a rotating mass drags spacetime around with it — the Lense–Thirring effect, often called frame dragging. NASA's Gravity Probe B mission (2004–2005) tested this by putting four ultra-precise gyroscopes in orbit around Earth. If GR is right, the gyroscope axes should precess in two ways: a large effect from orbiting through curved spacetime (the geodetic effect, about 6.6 arcseconds per year) and a tiny effect from Earth's rotation dragging spacetime (about 0.039 arcseconds per year — commonly cited reference values). The measured values matched the predictions. Satellite laser-ranging data (LAGEOS missions) provided independent support.

The binary pulsar: a gravity laboratory

In 1974, Russell Hulse and Joseph Taylor discovered PSR B1913+16, a pulsar (a rapidly spinning neutron star emitting clock-like radio pulses) locked in orbit with another neutron star. By timing the pulses, they found the orbit shrinks by roughly 76 microseconds per year (a commonly cited reference value) — exactly the energy loss predicted if the system is radiating gravitational waves. This was the first indirect confirmation that gravitational waves exist and that GR correctly predicts their strength, earning Hulse and Taylor the 1993 Nobel Prize in Physics.

Gravitational waves as the newest test

On September 14, 2015, the twin LIGO detectors recorded the first direct detection of gravitational waves (event GW150914) — the merger of two black holes about 1.3 billion light-years away. The observed signal matched GR's predicted waveform so closely that it simultaneously confirmed the waves and tested GR's strong-field predictions for the first time (see Gravitational Wave Astronomy). Since then, dozens of mergers have been recorded, and GR has passed every one.

Common Confusions