Simulating a pulsar timing array using metronomes and a microphone

To illustrate how scientists are using the ticks of an array of pulsars to look for gravitational waves, we have developed an instructional demonstration that uses metronomes and a microphone to simulate a pulsar timing array. Two metronomes play the role of two pulsars in the array; a microphone plays the role of a radio telescope on Earth; and the motion of the microphone relative to the metronomes simulates the passage of a gravitational wave near Earth.

The motion of the microphone causes the metronome pulses to arrive either slightly ahead or behind schedule due to the changing distance between the microphone and the metronomes. The changes in the pulse arrival times (called timing residuals) for the two metronomes show a "common trend" or "relationship" (also called a correlation), which depends on the location of the metronomes relative to one another. The correlations that we see in the timing residuals for the two metronomes are qualitatively similar to the correlations that we expect to see in the timing residuals for a pair of pulsars due to a passing gravitational wave. The analogy isn't perfect, but the concept of correlation is an important feature for both the metronome demonstration and a pulsar timing array.

To analyze the data generated during the demonstration, we wrote computer programs in the form of graphical user interfaces that can:

  1. record and playback the metronome pulses,
  2. estimate the pulse period (the time between successive pulses) and the pulse shape,
  3. calculate timing residuals by subtracting the expected arrival times of the pulses from the measured arrival times,
  4. calculate the correlation between a pair of measured timing residuals.

These analyses with the metronome data are very similar to the analyses that pulsar astronomers perform with real pulsar timing data. In short, by doing the demonstration we get a flavor of what pulsar timing is all about.

Want to learn more?

If you are interested in learning more about the demonstration and/or would like to do it yourself, you can download all of the relevant files, the user's guide, etc. by clicking on the links below:

Note that the analysis routines are written in python, and they require that certain packages be installed (such as pyaudio, matplotlib, numpy, scipy, etc.) prior to running.

Need to contact us?

If you run into any problems with the code, or if you have comments or suggestions that would improve the demonstration, please do not hesitate to contact us.