Skip to main content
Sandbox Physics
Pulsar timing arrays and the Hellings–Downs curve

What a pulsar timing residual measures

Sandbox Physics ·

A timing residual is the difference between when a pulse from a pulsar actually arrived and when a fitted timing model said it would arrive. The model removes everything that is specific to one pulsar and predictable: its spin and slow-down, its position and motion, its orbit and the delay added by the interstellar plasma. A slow gravitational wave survives that fit only partly, as a wandering of the residuals that lasts years. This chapter builds the residual, separates white noise from red noise, and works out how large the nanohertz signal reported by NANOGrav is in nanoseconds.

You need the idea of a sine wave, a least-squares fit and a variance. Times are in seconds, frequencies in hertz, and one nanohertz is 10−9 Hz10^{-9}\,\mathrm{Hz}.

From pulses to residuals

A pulsar's rotation phase grows almost linearly with time. Over years its spin frequency ν\nu slowly decreases, so to first approximation

ϕ(t)=ϕ0+ν t+12ν˙ t2,\phi(t)=\phi_0+\nu\,t+\tfrac12\dot\nu\,t^2 ,

and a pulse arrives whenever ϕ\phi passes a whole number. Observers record times of arrival, convert them to the centre of mass of the Solar System, and fit a timing model to them. In the NANOGrav 15-year analysis that model contains the spin period and its derivative, sky position, proper motion and parallax for every pulsar, five orbital elements and any needed extra terms for pulsars in binaries, and a piecewise-constant correction for the dispersion measure called DMX. NANOGrav Collaboration, 2023, Section 2

The residual is δt=tobs−tmodel\delta t=t_{\mathrm{obs}}-t_{\mathrm{model}}. A gravitational wave makes the pulses arrive early and late in a time series that varies at the wave's frequency, with an amplitude proportional to the wave's strain. Allen, 2022, Section I

What the fit takes away

The fit cannot tell a slow signal from a slightly wrong spin frequency or spin-down. Any part of a residual that looks like a constant, a straight line or a parabola is absorbed by ϕ0\phi_0, ν\nu and ν˙\dot\nu. Parallax and proper motion are fitted in every pulsar for the same reason: they produce delays at one cycle per half year and one cycle per year in all pulsars, which a search could otherwise mistake for a gravitational wave.

Worked example: how much of the slowest wave survives the spin fit

Sample a sine and a cosine with one full cycle over the span TT, and remove their least-squares fit by 11, tt and t2t^2. The sine keeps 39 %39\,\% of its variance and the cosine keeps 7.6 %7.6\,\%. At two cycles per span the two keep 85 %85\,\% and 94 %94\,\%, and at five cycles more than 97 %97\,\%. The lowest frequency, where a gravitational-wave background is strongest, is exactly where the spin fit removes the most.

White noise and red noise

Each arrival time carries a measurement error. NANOGrav models it with three parameters per receiver and backend: a scaling of the quoted uncertainties, an extra term added in quadrature, and a term common to all sub-bands at the same epoch. NANOGrav Collaboration, 2023, Section 2 This noise is white: independent from one epoch to the next, with equal power at all frequencies.

Red noise has more power at low frequencies. Some comes from the pulsar itself, some from the interstellar medium, and a gravitational-wave background is red too. To describe red processes, NANOGrav uses sines and cosines at the frequencies

fi=iT,T=16.03 yr,f_i=\frac{i}{T},\qquad T=16.03\ \mathrm{yr},

so f1=1.98 nHzf_1=1.98\,\mathrm{nHz}. Each sine and each cosine gets a random coefficient, and the variance of the coefficient at fif_i is the power in that bin. NANOGrav Collaboration, 2023, eqs. 1 to 3

White noise also has a variance per coefficient. Fit a unit sine to NN evenly spaced residuals with independent errors σ\sigma, and its coefficient has variance 2σ2/N2\sigma^2/N.

The power-law spectrum

A population of inspiralling supermassive black-hole binaries gives a characteristic strain hc(f)=A (f/fref)−2/3h_c(f)=A\,(f/f_{\mathrm{ref}})^{-2/3}, with fref=1 yr−1f_{\mathrm{ref}}=1\,\mathrm{yr}^{-1}. The matching variance of each Fourier coefficient is

Φi=A212π2 1T(fifref)−γfref−3,γ=133,\Phi_i=\frac{A^2}{12\pi^2}\,\frac{1}{T}\left(\frac{f_i}{f_{\mathrm{ref}}}\right)^{-\gamma}f_{\mathrm{ref}}^{-3},\qquad \gamma=\frac{13}{3},

in units of seconds squared. NANOGrav Collaboration, 2023, eqs. 6 and 8 The index 13/313/3 follows from γ=3−2α\gamma=3-2\alpha with α=−2/3\alpha=-2/3.

Worked example: the size of the reported background in nanoseconds

Take the published amplitude A=2.4×10−15A=2.4\times10^{-15}. With T=5.059×108 sT=5.059\times10^{8}\,\mathrm s and fref=3.169×10−8 Hzf_{\mathrm{ref}}=3.169\times10^{-8}\,\mathrm{Hz}, the first bin has f1/fref=1/16.03f_1/f_{\mathrm{ref}}=1/16.03 and

Φ1=5.76×10−30118.4×3.142×1022 s35.059×108 s×16.0313/3=5.03×10−13 s2.\Phi_1=\frac{5.76\times10^{-30}}{118.4}\times\frac{3.142\times10^{22}\,\mathrm{s^3}}{5.059\times10^{8}\,\mathrm s}\times16.03^{13/3}=5.03\times10^{-13}\,\mathrm{s^2}.

The coefficient spread is Φ1=709 ns\sqrt{\Phi_1}=709\,\mathrm{ns}. Each further bin is smaller by i−13/6i^{-13/6}: 158 ns158\,\mathrm{ns} at i=2i=2 and 22 ns22\,\mathrm{ns} at i=5i=5.

A sine with coefficient spread Φ\sqrt{\Phi} contributes a time-averaged residual variance Φ/2\Phi/2, and its cosine partner adds the same, so bin ii adds Φi\Phi_i to the residual variance before the timing fit.

Table 2.1 Coefficient spread of the reported background and of white noise with σ=1 μs\sigma=1\,\mu\mathrm s on N=192N=192 epochs.
Bin iifif_i (nHz)Background Φi\sqrt{\Phi_i} (ns)White noise (ns)
11.98709102
23.95158102
35.9366102
59.8822102

The background dominates the lowest two bins and falls below the white noise from the third. That is why a pulsar timing array needs many years: the signal lives in the few lowest frequencies, and each year of data moves f1f_1 lower.

Lab task

Open the pulsar-timing-array Lab and click a pulsar on the sky map. The white dots are its residuals with noise; the line is the smooth common signal alone, drawn with the signals of its nearest and farthest partners. Raise the white noise from 11 to 3 μs3\,\mu\mathrm s and watch the slow wander disappear into the scatter, while the correlation plot loses its shape.

Exercises

  1. NANOGrav limits the common-signal fit to the first 14 bins. What is f14f_{14}?
  2. By what factor is Φ2\sqrt{\Phi_2} smaller than Φ1\sqrt{\Phi_1} for γ=13/3\gamma=13/3? What would the factor be for a flat spectrum, γ=0\gamma=0?
  3. A pulsar has white noise σ=0.5 μs\sigma=0.5\,\mu\mathrm s on 400400 epochs. What is the white-noise spread of one Fourier coefficient?
  4. With the Table 2.1 numbers, at which bin index would the background equal the white noise, treating ii as continuous?
  5. Estimate the residual amplitude in the first bin from hc(f1)/(2πf1)h_c(f_1)/(2\pi f_1) and compare it with Φ1\sqrt{\Phi_1}.
Answers
  1. f14=14/T=27.7 nHzf_{14}=14/T=27.7\,\mathrm{nHz}, matching the 2 to 28 nHz range quoted at the end of Section 2 of the paper.
  2. 2−13/6=0.2232^{-13/6}=0.223. A flat spectrum gives 20=12^{0}=1: every bin has the same power, like white noise.
  3. 2/400×0.5 μs=35 ns\sqrt{2/400}\times0.5\,\mu\mathrm s=35\,\mathrm{ns}.
  4. 709 i−13/6=102709\,i^{-13/6}=102 gives i=6.956/13=2.45i=6.95^{6/13}=2.45, between the second and third bins.
  5. hc(f1)=2.4×10−15×16.032/3=1.53×10−14h_c(f_1)=2.4\times10^{-15}\times16.03^{2/3}=1.53\times10^{-14}, and hc/(2πf1)=1.2 μsh_c/(2\pi f_1)=1.2\,\mu\mathrm s. It agrees with 709 ns709\,\mathrm{ns} within a factor of two, as an order-of-magnitude estimate should.

Sources

  1. NANOGrav Collaboration, 2023, Section 2
  2. Allen, 2022, Section I

Keep reading

What this page links to directly

See this series on the content map →

Pulsar timing arrays and the Hellings–Downs curve · Reading path

From one ticking pulsar to a Galaxy-sized detector

  1. ArticleHow can a pulsar timing array tell gravitational waves from a bad clock?
  2. You are hereWhat a pulsar timing residual measures
  3. ArticleDeriving the Hellings–Downs curve
  4. ArticleThe optimal statistic and its null tests
  5. ArticleRecompute a Hellings–Downs statistic from NANOGrav 15-year data
  6. ArticleReading the NANOGrav 15-year evidence for a gravitational-wave background
Something broken, confusing, or missing on this page?Tell us
What kind of feedback?

Sent with this page’s address, its Lab ID, the site language and your device type. Please leave out personal details. Feedback is kept for up to 12 months.Privacy policy