What is a pulsar for laypeople

Gravitational waves from the home computer

The Einstein @ Home project enables everyone to search for gravitational waves on their own PC, laptop or smartphone and thus become an explorer themselves. Bruce Allen, Director at the Max Planck Institute for Gravitational Physics in Hanover, founded this Citizen Science project. In the meantime, the software also detects pulsars in the big data. Researchers from the Max Planck Institute for Radio Astronomy in Bonn are also involved in this search.

Text: Thomas Bührke

The discovery of a gravitational wave on September 14, 2015 with the LIGO detectors in the USA is considered a scientific sensation. One of the last predictions of Albert Einstein's general theory of relativity was thus confirmed. Scientists from the Max Planck Institute for Gravitational Physics in Golm and Hanover played a key role in the discovery (MaxPlanckResearch 1/2016, page 78 ff.).

The first detected gravitational wave was unexpectedly strong, its signal could be seen even with the naked eye in the data stream that the supercomputer Atlas in Hanover was constantly analyzing. There a researcher from the Max Planck Institute was the first to notice the signal. But there was another way. Since the beginning of the LIGO measurements, the scientists have been looking for weak periodic gravitational waves, such as those probably emitted by rapidly rotating neutron stars. Atlas is the world's largest data analysis cluster for this purpose, but this task even reaches its limits.

The designers of the gravitational wave detectors were aware of this problem from the start, and so two of them came up with an idea. “It was August 19, 1999,” Bruce Allen remembers very clearly. At the California Institute of Technology (Caltech) he met his colleague Stuart Anderson for dinner. Allen had read an article about the SETI @ Home project in the Los Angeles Times. The search for signals from an extraterrestrial intelligence in the data of large radio telescopes confronts the researchers with the same problems as Allen and colleagues: How can one find periodic signals in the huge mass of data?

A network with gigantic computing capacity

SETI @ Home is based on a decentralized analysis of the data, which is distributed in packets to thousands of private computers. This is how it works: You log on to your home PC and receive software that scans the data whenever the screen saver starts. The result will be sent back automatically. In this way, the search for signals can be distributed over a network with enormous computing capacity. SETI @ Home was very well received from the start. However: the search for aliens has so far been in vain.

"I discussed with Stuart the possibility of having the gravitational wave data of the two LIGO instruments searched in the same way," says Bruce Allen. "But then we thought: Everyone is interested in aliens, but who cares about gravitational waves?" With that, the idea died; for the time being anyway. Four years later the ball got rolling. Allen received a call from a SETI @ Home pioneer looking for actions for the upcoming international Einstein year 2005.

Everyone immediately remembered the conversation in the Caltech canteen - and suddenly saw an opportunity for the idea discussed at the time. The researcher immediately submitted an application to the National Science Foundation for funding of two million dollars over three years, in which he included the University of Berkeley and the Max Planck Institute. But the comparatively small amount was not approved. It was now June 2004, the Einstein year not far away.

Without further ado, Bruce Allen decided to develop the corresponding software on their own together with his employees. The project was given the finishing touches by David Anderson from the University of Berkeley, who had also written the software for SETI @ Home. "We carried them over to our project, which was a huge step forward," says Allen. The scientists succeeded in completing a first version by February 2005, which they presented at a press conference at the annual meeting of the American Association for the Advancement of Science.

The media took up the project with enthusiasm. It was named Einstein @ Home and was tailor-made for the Einstein year. The news spread quickly: According to Allen, around 20,000 participants had registered within a few days. This in turn also called on the National Science Foundation, which now pledged financial support without further ado.

The Max Planck Institute for Gravitational Physics, to which Allen was appointed director in 2007, was part of Einstein @ Home from the start. To date, several hundred thousand people around the world have taken part in the project, with some jumping off again and again, so that around 40,000 hobby researchers - sometimes with several devices at the same time - are always active at a certain point in time.

Signals are chopped up into small packets

There are around 100 other projects in which data is viewed in distributed computing. The spectrum ranges from drug development against malaria and molecular simulations of proteins to the search for the largest known prime number. Einstein @ Home is one of the largest among them. Today it achieves a total computing power of 1.7 petaflops per second, i.e. 1.7 quadrillion computing steps. This makes this computer network one of the 60 most powerful supercomputers in the world. “Our Atlas cluster plays a central role in the network,” explains Bruce Allen. It processes the signals coming from the LIGO detectors and chops them up into small packages.

They are chosen so that each participating computer does not receive more than one megabyte of data per hour. Atlas only needs one percent of its output for this administrative activity. The data scanned by the PCs and laptops is sent back to Atlas and processed for the scientists, for example in the form of diagrams. If a suspicious point in the data stream has been reported, Atlas takes a close look at it.

Despite years of effort, the search for gravitational waves has remained unsuccessful. This is a bit frustrating, but resourceful researchers can even draw astrophysical conclusions from this null signal. In terms of neutron stars - around 20 kilometers of remains from exploded suns.

These neutron stars have extreme properties: The matter in them is so compressed that a teaspoon of them on earth weighs as much as a million long-distance trains. They also rotate very quickly around their own axis. This is a good prerequisite for radiating gravitational waves, whereby the frequency of such a wave corresponds to the rotational frequency of the body. But neutron stars only emit these space-time waves when they are not perfectly symmetrical.

However, neutron stars are probably among the roundest bodies in the universe and are therefore bad transmitters. From the fact that no periodic signal has yet been found from them, something can be said about their symmetry. The intensity of a gravitational wave received on earth decreases with increasing distance from the neutron star, and the LIGO detectors are most sensitive in the frequency range from a few dozen to a few hundred Hertz. Therefore, only statistical statements can be made about the shape of neutron stars.

According to this, within a radius of about 1000 light years there is no neutron star with a rotational frequency of 100 Hertz or more whose surface deviates from the spherical shape by more than ten centimeters. An extremely remarkable result. "In the field of electromagnetic waves, we have so far detected several thousand neutron stars - out of a total of perhaps 100 million that exist in our Milky Way", says Maria Alessandra Papa from the Max Planck Institute in Hanover. "In the future, gravitational wave astronomy will offer a whole new way of getting more information about this invisible population."

Bundles of radio sweep the earth like spotlights

These findings are important to astrophysicists. Even so, the enthusiasm of even the greatest Einstein @ Home enthusiast wanes when no signal goes online for years. This worried Bruce Allen, so he looked for another area of ​​application. He found this after hearing a radio astronomer talk about the search for pulsars in late 2007. Behind these objects are neutron stars that emit two bundled radio beams along the magnetic field axis in opposite directions into space. If the axis of rotation and the axis of the magnetic field are inclined towards each other, the two radio bundles sweep through space like the headlights of a lighthouse. If they happen to cross the earth, the telescopes receive a periodic signal with the rotation frequency of the pulsar.

Bruce Allen immediately realized that Einstein @ Home should apply to this area. In particular for double systems in which a neutron star and a companion orbit each other, Einstein @ Home should make a significant contribution to the discovery of such systems. "With their analysis methods, the radio astronomers can only find pairs whose orbital period lasts more than about an hour," says Allen. "But we would also have to be able to track closer pairs down to a period of ten minutes."

The analysis of the measurement data from radio telescopes is similar to that from gravitational wave detectors, but this extension required considerable effort. Allen's doctoral student at the time, Benjamin Knispel, found the task exciting and set about it. It became his doctoral thesis. “The software had to be rewritten considerably,” recalls Knispel. "The data from radio telescopes differ in many ways from that of the LIGO detectors."

The greatest challenge is that the physicists do not know whether a pulsar signal is hidden in the respective data set. And if so, at what frequency. If a pulsar is in a double system, another difficulty arises: If it moves on its path towards us, the pulses arrive in shorter succession; if he runs away from us, the pulse rate becomes slower. The pulse frequency changes periodically with the duration of the pulsar's revolution. "This blind search for signals whose parameters are not known at all is very time-consuming," explains Knispel: "We want to use our limited computing capacity optimally, as if one wanted to get the greatest profit out of a certain bet in a casino." Einstein @ Home is ideally suited for this blind search because it analyzes small data packets with great computing power particularly efficiently. Because of the many decentralized private computers, you can get them almost free of charge.

Successful search for unknown beeper

Einstein @ Home has also been looking for radio pulsars since March 2009. The data comes from the PALFA (Pulsar Surveys with the Arecibo L-Feed Array) project, which is running on the 305-meter antenna of the Arecibo observatory. It only took about a year to be discovered. The PCs of two participants had detected a conspicuous signal in the same data set. A follow-up analysis with Atlas confirmed the find. Now the professionals got involved. In July 2010, astronomers used the radio telescope in Green Bank (USA) to search for the previously unknown beeper. And they were successful: It was a pulsar that swirls around its axis 41 times per second.

The astronomers aimed further radio telescopes at the newly discovered celestial body, including the Effelsberg 100-meter antenna of the Max Planck Institute for Radio Astronomy. These follow-up observations revealed that the pulsar is a loner some 17,000 light-years away, with a magnetic field around 20 billion times stronger than that of Earth.