Skip to main content
Sandbox Physics

AC06 · A pulse crosses the plasma

Interstellar Radio Propagation

Why does one cosmic radio flash arrive at different times and with a rotated polarization? Free electrons delay lower radio frequencies more strongly. A magnetic field along the path also rotates linear polarization. Unequal paths through density fluctuations spread the pulse into a tail, so its timing and polarization reveal different parts of the intervening plasma.

Drag an electron screen and reverse its magnetic field. Watch a pulse disperse, grow a scattering tail and rotate its polarization, then recover the rotation from recorded detector samples.

Interactive modelInterstellar Radio Propagation
400 to 800 MHz delay389 ms389\,\mathrm{ms}
Fitted rotation measure32.5 rad m−232.5\,\mathrm{rad\,m^{-2}}
Recovered linear polarization80.0 %80.0\,\%
Prescribed tail at 600 MHz12.0 ms12.0\,\mathrm{ms}

Receiver brightness and polarization marks use the same saved synthetic I/Q/U record as these plots.

A synthetic pulse at the receiver
Align the frequency channels
Recorded linear polarization
Find the rotation measure

Gray profiles precede correction and mint profiles use the selected DM; Q is blue, U pink, and the response-aware RM fit gold, with secondary peaks from sampling aliases.

Readout and fit scope

The scene compresses distance and enlarges differential delay; blue, violet and pink mark high, middle and low frequencies. Receiver brightness uses saved intensity and its direction marks use integrated Q/U; all panels share one frozen synthetic record.

Waterfall brightness uses recorded intensity above zero with a fixed asinh stretch and a blue-white display palette. Replaying the scene preserves the acquired noise samples.

The delay and 600-MHz tail are prescribed model quantities. RM and linear fraction are recovered from time-integrated recorded Q/U, using equal Gaussian channel noise, one Faraday-thin component, and a common intrinsic polarization spectrum. The search spans minus 330 to plus 330 radians per square meter in steps of a quarter. Integration includes the saved window; its exponentially small missing tail is not rescaled. The finite-step scattering kernel gives sub-sample timing uncertainty. The display sample interval adapts to the pulse window.

Physics tutorial

Three changes to one pulse

BackgroundInterstellar space contains free electrons. The pulse drives them back and forth, and their response changes its propagation. A magnetic field adds a different response for the two circular polarization components.

Why it mattersArrival times probe an electron column. Polarization probes its magnetic weighting. A scattering tail adds information about uneven paths through the medium.

Start with the essentials

Focus question
Which changes when the field reverses: arrival time, polarization or both?
One-sentence intuition
Dispersion measures electron column; signed Faraday rotation measures the field weighted by that column.

Core mathematical model

Relative group delay

Δt=4.149 ms DMpc cm−3[(νGHz)−2−(νrefGHz)−2]\Delta t=4.149\,\mathrm{ms}\,\frac{\mathrm{DM}}{\mathrm{pc\,cm^{-3}}}\left[\left(\frac{\nu}{\mathrm{GHz}}\right)^{-2}-\left(\frac{\nu_{\rm ref}}{\mathrm{GHz}}\right)^{-2}\right]

Lower frequencies arrive later. The coefficient is rounded, and frequencies are observed frequencies.

Faraday rotation

P=Q+iU=Le2iχ,χ=χ0+RMλ2,RM=0.812∫neB∥ dlP=Q+iU=L e^{2i\chi},\qquad \chi=\chi_0+\mathrm{RM}\lambda^2,\qquad \mathrm{RM}=0.812\int n_e B_\parallel\,dl

Electron density is in inverse cubic centimeters, field in microgauss and path length in parsecs. Positive field points toward the observer; RM is in radians per square meter.

A causal scattering tail

h(t)=e−t/ττH(t),Iout(t)=∫Iin(t−u)h(u) duh(t)=\frac{e^{-t/\tau}}{\tau}H(t),\qquad I_{\rm out}(t)=\int I_{\rm in}(t-u)h(u)\,du

The normalized response redistributes fluence toward later times. Its shape assumes an isotropic thin screen.

Common difficulties

A polarization angle repeats

Typical misconceptionA straight line through the displayed angles always gives the rotation measure.

Better mental modelAn angle repeats after a half-turn. Fitting Q and U together avoids choosing an angle unwrap, though finite frequency sampling still creates competing RM peaks.

Run the experiment

  1. 01

    Change one sign

    Predict the result, then choose Reverse the field.

    What to observe: The rotation-measure peak changes sign. Dispersion follows the same electron column.
  2. 02

    Spread the arrival paths

    Choose A long scattering tail and compare the dispersed and dedispersed profiles.

    What to observe: Removing the frequency delay aligns the leading pulse, but a tail remains within each channel.
  3. 03

    Average a rotating vector

    Choose Coarse channels lose polarization. Reduce only Frequency channel width.

    What to observe: Opposing polarization vectors cancel within wide channels. Narrowing them restores the observed Q and U oscillations.

Related experiments and articles

Continue exploring

Measuring the sky

What do positions, colors, interference and radio delays reveal about the sky?

  1. LabObserving Cadence and AliasesMove exposures and compare the rhythms that survive the sampling.
  2. LabSupernova Distances and the Distance LadderMove a geometric distance anchor and follow its shared calibration error.
  3. LabStellar Parallax and Proper MotionChange the stellar distance and follow its annual motion.
  4. LabCrowded-star PhotometryMove the aperture and compare the light admitted from neighbouring stars.
  5. LabStellar Absorption LinesChange the gas temperature and velocity, then compare the absorption lines.
  6. LabGalaxy Colors and Photometric RedshiftCompare redshift, dust and stellar-population effects in the same five bands.
  7. LabStellar Interferometry and Sparse ImagingMove a telescope and compare images fitted to the same sparse signals.
  8. You are hereInterstellar Radio PropagationChange the electron column and field, then follow the arriving radio pulse.

Radio pulses through interstellar space · Reading path

Radio pulses through interstellar space

Articles on this path that use this experiment, followed by the other experiments they suggest.

  1. ArticleWhy low radio frequencies arrive late
  2. ArticleFrom electron motion to a radio delay
  3. ArticleFit the rotating polarization vector
  4. ArticleReplay a measured CHIME radio flash offline
  5. ArticleRead a Faraday spectrum without turning it into a distance map
FeedbackSomething 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