Post-Keplerian Parameters: How Binary Pulsars Weigh Neutron Stars and Test Gravity

What Post-Keplerian Parameters Are

In Newtonian gravity a two-body orbit is a fixed ellipse fully described by five Keplerian parameters — orbital period P_b, eccentricity e, projected semi-major axis x = a·sin(i)/c, and the angle and epoch of periastron (ω, T_0). Real relativistic binaries do not obey this perfectly. The orbit precesses, the clock runs slow, and the orbit slowly shrinks. These departures are captured by post-Keplerian parameters: a small set of additional observables layered on top of the Keplerian ones.

The great advantage is that the PK framework is theory-independent. The Keplerian and post-Keplerian quantities are simply what you fit to the pulse arrival times — no theory of gravity is assumed in the measurement itself. Only afterward do you ask: what does a given theory (general relativity, or a scalar-tensor alternative) predict for each PK parameter in terms of the two stellar masses? The five classic PK parameters are:

• ω̇ — the rate of periastron advance
• γ — the Einstein delay (time-dilation/redshift amplitude)
• Ṗ_b — the orbital period decay
• r and s — the range and shape of the Shapiro delay

The Mechanism: One Theory, Two Unknowns, Many Equations

The physical basis is that in general relativity each PK parameter is a known function of only the two component masses m_1 (pulsar) and m_2 (companion) plus the already-measured Keplerian quantities. For periastron advance:

ω̇ = 3 (2π/P_b)^(5/3) (T_sun M)^(2/3) / (1 − e²)

where M = m_1 + m_2 is the total mass and T_sun = G·M_sun/c³ ≈ 4.925 μs is the solar mass in time units. Similarly, the orbital decay from gravitational-wave emission follows the quadrupole formula:

Ṗ_b = −(192π/5)(2π/P_b)^(5/3) (T_sun^(5/3) m_1 m_2 / M^(1/3)) · f(e)

where f(e) = (1 + 73e²/24 + 37e⁴/96)/(1 − e²)^(7/2) enhances the loss for eccentric orbits. The Einstein delay γ and Shapiro terms r, s are likewise fixed functions of the masses and geometry.

Here is the crux: two masses are two unknowns. Measure any two PK parameters and you solve for both masses. Measure a third, and GR must predict it correctly — a consistency check. Each PK parameter draws a curve in the (m_1, m_2) mass plane, and in a valid theory all curves cross at a single point.

Characteristic Numbers and a Worked Example

Consider the Hulse-Taylor pulsar, PSR B1913+16: P_b = 7.7519 hours, e = 0.617, and pulsar spin period 59 ms. Its measured periastron advance is ω̇ = 4.226598°/yr — roughly 35,000 times the famous 43 arcsec-per-century anomaly of Mercury, and accumulating in a day what Mercury takes a century to do.

• From ω̇ alone the total mass is pinned to M = m_1 + m_2 ≈ 2.828 M_sun.
• Adding the Einstein delay γ ≈ 4.3 ms separates the masses: m_1 ≈ 1.438 M_sun, m_2 ≈ 1.390 M_sun.
• General relativity then predicts the orbital decay: Ṗ_b ≈ −2.40 × 10^-12 s/s, i.e. the 7.75-hour orbit shortens by about 76 microseconds per year.

The observed decay matches the GR prediction to a ratio of 0.997 ± 0.002 after correcting for the galactic acceleration term — the first quantitative proof that gravitational waves exist. In the Double Pulsar the same energy loss makes the two stars close by roughly 7 mm per day.

How They Are Observed: Pulsar Timing

PK parameters are extracted by pulsar timing. A millisecond or recycled pulsar is a natural atomic clock, emitting a pulse each rotation with a stability rivaling laboratory standards. Radio telescopes record pulse times-of-arrival (TOAs), and a timing model predicts each TOA by accounting for the pulsar spin, its slow-down, the Earth's motion, interstellar dispersion, and — crucially — the binary orbit.

• ω̇ shows up as a slow drift in where periastron sits, detectable over months to years.
• γ modulates the pulse arrival across the orbit as the pulsar samples different depths of the companion's potential and different orbital speeds.
• Ṗ_b appears as a quadratic drift in the orbital phase, growing with the square of the observation span — which is why decades of data sharpen it dramatically.
• r and s (Shapiro delay) produce a sharp, non-sinusoidal spike in arrival time near superior conjunction, when pulses graze the companion's warped spacetime; it is only measurable for near edge-on orbits (high sin i).

Instruments like Arecibo, the Green Bank Telescope, Parkes, and now MeerKAT push residuals to the sub-microsecond level, enabling ever more PK parameters — including higher-order and even 2.5-PN effects.

Related Regimes and How PK Parameters Differ

Post-Keplerian parameters occupy a distinctive corner of gravitational physics: strong-field but slow-motion. The neutron stars have enormous surface gravity and compactness (Gm/Rc² ~ 0.2), yet they orbit slowly (v/c ~ 10^-3), so the relativistic effects accumulate gently over many orbits rather than in a violent plunge.

• vs. Solar System tests: Mercury's perihelion precession and the Cassini Shapiro delay probe weak fields (Gm/Rc² ~ 10^-6). Binary pulsars test the same effects where self-gravity is a billion times stronger, catching any deviation that only appears in strong fields (e.g. dipolar gravitational radiation predicted by scalar-tensor gravity).
• vs. LIGO/Virgo mergers: Ground-based detectors observe the final seconds of a merger — strong-field and fast-motion, up to v/c ~ 0.5. Pulsar timing instead watches the long, slow inspiral millions of years earlier.
• vs. the parameterized post-Newtonian (PPN) framework: PPN describes weak-field deviations with parameters like γ_PPN and β. PK parameters are the strong-field analog, tied to a specific binary's masses.

Together these regimes bracket gravity from the weakest to the most extreme fields observable.

Significance, Famous Cases, and Open Questions

The story begins in 1974 when Russell Hulse and Joseph Taylor discovered PSR B1913+16; their measurement of its orbital decay earned the 1993 Nobel Prize in Physics and gave the first indirect proof of gravitational waves — nearly 40 years before LIGO's direct detection. The theoretical machinery to interpret PK parameters was built by Thibault Damour and Nathalie Deruelle in 1985-1986 (the 'DD' timing model).

The current champion is the Double Pulsar PSR J0737-3039A/B, discovered in 2003 — the only known system where both neutron stars are visible as pulsars. Because you also measure the mass ratio R = m_A/m_B directly from the two orbits, it yields an unusually rich set of independent tests. In 2021, Michael Kramer and collaborators reported that GR passes at the 1.3 × 10^-4 level, including for the first time signal-propagation effects like retardation and light deflection by the companion.

• Open questions: The kinematic 'galactic' correction to Ṗ_b now limits precision, requiring exquisite distances and proper motions.
• Searchers hope PK timing may one day constrain the neutron-star equation of state via spin-orbit coupling, or detect a pulsar orbiting a black hole to test the no-hair theorem.

The five classic post-Keplerian parameters and what each one measures

PK parameter	Symbol	Physical effect	Measured in Hulse-Taylor / Double Pulsar
Periastron advance	ω̇	Orbit ellipse slowly rotates (like Mercury, but ~10^5× faster)	4.23°/yr (B1913+16); 16.9°/yr (J0737-3039)
Einstein delay	γ	Time dilation + gravitational redshift of the pulsar clock	~4.3 ms amplitude (B1913+16)
Orbital decay	Ṗ_b	Orbit shrinks as gravitational waves carry off energy	-2.4 × 10^-12 s/s (B1913+16); orbit shrinks ~7 mm/day (J0737)
Shapiro range	r	Extra pulse delay set by companion mass (r = T_sun·m_c)	m_c ≈ 1.25 M_sun (Double Pulsar B)
Shapiro shape	s	Delay's dependence on orbital inclination (s = sin i)	sin i ≈ 0.99995, orbit nearly edge-on (J0737)