\texttt{cWB-space}: A time-frequency transient-search pipeline for LISA

Space is not silent. While we often think of the universe as a vast, quiet vacuum, it is actually filled with ripples in the fabric of space and time itself. These ripples, known as gravitational waves, are created when massive objects like black holes collide or when stars explode. For decades, scientists on Earth have been listening for these whispers using giant laser detectors. Now, a new generation of detectors is being prepared for space. The Laser Interferometer Space Antenna, or LISA, will be a constellation of three spacecraft flying in a giant triangle, millions of kilometers apart, designed to listen to a different range of cosmic sounds than those heard on Earth. However, listening from space presents a unique challenge: the sky is crowded. Unlike the sparse signals Earth detectors often catch, LISA will hear a constant hum of thousands of sources all at once, overlapping like voices in a busy room. Among this noise, there may be sudden, unexpected bursts of sound—transient events that scientists cannot predict in advance. Finding these hidden signals requires a new kind of listening strategy, one that does not rely on knowing exactly what the sound should look like before it arrives.

A team of researchers has developed a new method called cWB-space to tackle this problem. Instead of guessing the shape of a signal, this approach acts like a highly sensitive spotlight that scans the data for any sudden concentration of energy. It looks at how the power of the signal is distributed across time and frequency, searching for bright spots that stand out against the background noise. Once a candidate signal is found, the system does not just flag it; it attempts to reconstruct the actual shape of the wave using only the data collected by the detectors. Crucially, it does this without forcing the signal to fit a pre-made template. This is vital because if scientists only look for signals they already expect, they might miss something entirely new. The method also uses the unique way LISA measures light to help distinguish between a real cosmic event and a glitch in the instrument itself. Since the spacecraft measure changes in the distance between them using laser light, a real gravitational wave will affect the measurements in a specific, coordinated way across the entire constellation, whereas a mechanical disturbance inside one part of the spacecraft will look different.

To test if this system works, the researchers ran a series of rigorous simulations. They created a digital version of the LISA detector, complete with realistic noise and the expected background of thousands of overlapping sources. Into this digital environment, they injected six specific signals representing massive black holes colliding, along with other types of burst signals and various types of instrumental glitches. The goal was to see if cWB-space could find these hidden signals and tell them apart from the noise and the machine errors. The results were striking. When the team searched through the simulated data, the system identified all six of the injected black hole collisions. Not only did it find them, but it ranked them as the six most significant candidates in the entire dataset, placing them at the very top of the list. Furthermore, when the researchers looked at the reconstructed waveforms—the shapes of the waves the system pulled out of the noise—they found that these shapes matched the original injected signals with remarkable precision. The system successfully recovered the main features of the waves, even though it had no prior knowledge of what they were supposed to look like.

The method also proved effective at distinguishing real signals from instrument trouble. The researchers tested the system with artificial glitches, which are sudden spikes in the data caused by the equipment rather than the universe. In most cases, the system could tell the difference. A real gravitational wave creates a specific pattern across the different laser links of the spacecraft, while a glitch usually affects only one link or a specific part of the instrument. By comparing how the signal appeared in these different channels, the system calculated a ratio that helped identify the origin. Most of the glitches produced a high ratio, while the real black hole signals produced a very low one, allowing the system to filter out the false alarms. However, the researchers also found a limitation: if a glitch happened to affect two specific links in a perfectly synchronized way, the system could be fooled, mistaking the machine error for a real signal. This rare case highlights that while the method is powerful, it is not infallible, and further work is needed to refine how these edge cases are handled.

The final test involved a more complex scenario using a year's worth of simulated data from a public challenge known as Sangria. This dataset contained a mix of many different sources, including the six black hole collisions hidden within a year of continuous observation. The cWB-space pipeline scanned this entire year of data and again produced a ranked list of candidates. The six injected black hole mergers appeared as the top six entries on this list, confirming that the method can find specific targets even when they are buried in a massive amount of data. The researchers then reconstructed the waveforms for these top candidates and compared them to the known signals. The match was excellent, with the reconstructed waves capturing the essential details of the merger events. This success demonstrates that a template-free search can effectively find and describe signals that were not anticipated, offering a new way to explore the universe.

The implications of this work extend beyond just finding black holes. By providing a way to identify and reconstruct signals without needing a perfect model beforehand, this method opens the door to discovering entirely new types of cosmic events. It allows scientists to listen for the unexpected, whether that is a black hole spinning in a complex way, a star being torn apart, or a phenomenon that physics has not yet predicted. The ability to reconstruct the waveform directly from the data means that scientists can study the physics of these events in detail, checking if they match our current theories or if they reveal something new. While the current tests were performed on simulated data, the results suggest that when LISA launches, this pipeline will be ready to help astronomers navigate the crowded gravitational-wave sky, ensuring that no interesting signal goes unheard. The work represents a significant step forward in preparing for a future where we can listen to the universe with both precision and openness.