LISA pathfinder mission a glorious success


Even before the first gravitational waves were observed, plans were in place for the generation that would follow the successful LIGO detectors. The new hardware is expected to operate in space and sense gravitational waves that we have little to no chance of detecting using Earth-based observatories.

Of course, no one wants to launch a very expensive system into space without some assurance that it will work. Hence, the ESA developed a pathfinder mission that tests the technology. The latest report from the pathfinder mission is not just positive, it is what-did-I-just-snort positive.

Illuminating stretchy space

Gravitational waves are detected by sensing very tiny shifts in the distance between two mirrors, which change as a gravitational wave passes through, and the very fabric of spacetime stretches and contracts. If we can count the number of wavelengths that fit between two mirrors, we can sense the change in distance. LIGO (laser interferometer gravitational wave observatory), for instance, uses this approach to spot changes of about 10-19 meters between mirrors that are separated by four kilometers.

That sensitivity would be improved if the distance were greater. On Earth, though, the distance we can build in a straight line is limited. And, even worse, what scientists really want is sensitivity to low-frequency gravitational waves, which requires long distances and a quiet environment. Our planet is not especially quiet.

In the first gravitational-wave detection—a merger between two black holes—the signal came out of the noise and reached a peak within about half a second before disappearing. Over that time, gravitational waves with frequencies between 30 and 300Hz were detected. That frequency corresponds to the orbital period of the black holes. So, we detected this death spiral once the black holes were orbiting each other 30 odd times per second.

That pre-merger spiral had been going on for much longer than that, emitting gravitational waves with much lower frequencies. If we ever want to see these, we have to have detectors that are sensitive to waves that have a much longer wavelength. And that simply isn’t possible on Earth.

It is, however, possible in space, and that is where LISA (laser interferometer space antenna) comes in. LISA is, as the name suggests, a space-based gravitational-wave observatory. However, unlike LIGO, which could be built and incrementally improved, LISA has to work the first time.

Blazing a trail for LISA

To maximize the chance of a functional space-based gravitational-wave observatory, the European Space Agency launched the LISA pathfinder, a satellite that is designed to test key technologies required for LISA. In particular, LISA requires measuring tiny accelerations of test masses. However, these masses are sitting in the noisy environment of a spacecraft, which heats and cools and sporadically gets smacked by little rocks. One of the main mission objectives was to see just how much noise there was and what sort of technology LISA would require to operate successfully.

It must be said that the goals for the pathfinder mission were comparatively modest. LISA must be able to measure accelerations as small as three femtometers/s2 (a fm is 10-15m) at a frequency just above a millihertz. Pathfinder was expected to do no better than around 20fm/s2, just to demonstrate that the hardware was on the right track. The engineers must have partied hard into the night when they found that the pathfinder has been (and may still be) nearly at the sensitivities required for LISA—only off by a factor of two.

But the team wasn’t done there. After recovering from the hangover, the scientists returned to their data to try to understand where the rest of the noise was coming from. In their analysis, they uncovered something very odd: a systematic error that added pseudo-random noise. Normally, a systematic error is something like a constant offset—my instrument always reports 1 m/s2 more acceleration than is actually present.

But it turned out that a processing error between an analog signal and its digitization, though systematic in nature, generated an effectively random noise in the acceleration data. Once removed, the noise improved substantially.

This improvement was not the only one that turned up. The researchers also noticed that the signal steadily got cleaner with time. This steady improvement was found to be due to reduced Brownian motion.

Vacuum cleaning

This is due to how the experiment was configured. Once the spacecraft achieved orbit, test masses were suspended in the vacuum of space inside the satellite, which protected them from temperature changes and being struck by passing material. Normally, space is the best vacuum achievable and far better than the best Earth-based vacuum systems.

But the surrounding satellite turned out to be a problem. All the volatile materials in the wiring and electronics, as well as water attached to the metal walls, slowly boil off over time. As a result, the vacuum close to the masses is actually pretty poor (about two to three orders of magnitude worse than a good Earth-based vacuum system). These residual gases collide with the mass, accelerating it in random directions and generating noise. This is classic Brownian motion.

Over time, however, the volatile elements, like water, slowly escape into space. As the vacuum improved, the noise induced by Brownian motion decreased steadily.

The researchers were also able to use longer datasets to better compensate for the spacecraft’s rotation. The pathfinder satellite uses certain stars to determine its orientation and rotation, but that’s limited by how accurately the optics can determine the center of each star. By using data acquired over longer time periods, repeated observations of the stars were used to pinpoint the star’s center more accurately.

The end result of all this is that the Pathfinder satellite—which was only supposed to perform an order of magnitude worse than required for LISA—actually performs a factor of three better than required for LISA.

The engineers are probably on another bender right now.

A flatulent satellite?

All is not rosy, however. The longer datasets for averaging were acquired by removing glitches. Mathematically, the glitches are quite separate from the expected data and can be removed. But this is highly undesirable, as any gravitational wave that has the same timescale as a glitch would, at present, also be removed.

To make matters worse, no one really knows why these glitches occur. The current best suggestion is the boiling off of volatiles. One idea is that, instead of providing a steady stream of gas, sometimes the gas is trapped in a pocket that suddenly bubbles out and farts on the masses. A calculation of the required amount of gas per glitch indicates that this is not an unreasonable suggestion. Nevertheless, the data also suggests that thorough degassing is going to be critical for LISA.

This is, I think, surprising. I would expect that all the components for pathfinder were degassed about as well as possible already. In that respect, I think there will be a search for new materials that release their volatile gases smoothly and silently.

In the meantime, LISA has already been approved, and the lessons learnt from Pathfinder are probably being incorporated into the final LISA satellite designs. I look forward to finally getting to hear the Universe’s astrophysical symphony (preferably without fart noises).

Physical Review Letters, 2018, DOI: 10.1103/PhysRevLett.120.061101 (About DOIs).

More Info: