Instrumental artifacts in photonic lantern spectroastrometry and their mitigation with PLred

To see the universe in sharp detail, astronomers often rely on a technique called spectroastrometry. Imagine trying to measure the width of a distant star or the swirling gas around a black hole. These objects are so far away that even the most powerful telescopes see them as mere points of light, blurring together into a single, fuzzy dot. Spectroastrometry is a clever workaround that allows scientists to peek inside that blur. Instead of just measuring where the light is, it measures how the center of that light shifts as the color of the light changes. If different colors of light come from slightly different places within that fuzzy dot—perhaps because a disk of gas is spinning—the center of the light will wiggle back and forth across the spectrum. By tracking these tiny wiggles, astronomers can map structures that are far smaller than the telescope's normal limit, revealing the hidden architecture of the cosmos.

However, this method is incredibly sensitive to the tools used to capture the light. The instruments themselves can create false signals that look exactly like real cosmic movement, leading scientists to see patterns that aren't there. A team of researchers, led by Yoo Jung Kim at the University of California, Los Angeles, recently tackled a specific version of this problem involving a new type of instrument called a photonic lantern. This device, tested on the Subaru Telescope in Hawaii, breaks light into many separate channels to measure its position with extreme precision. While this approach avoids some old problems, it introduced new ones: the way the instrument reads the light and the way the camera records it can both create misleading signals. The team developed a new software tool, called PLred, to fix these errors, ensuring that the tiny movements they see are real and not just glitches in the machine.

The photonic lantern works differently than the traditional slits used in older telescopes. Instead of slicing a beam of light with a narrow gap, the lantern acts like a fiber-optic splitter, taking the incoming light and dividing it into nineteen separate streams. Each stream carries a piece of the spatial information, and by comparing how bright each stream is at different colors, scientists can reconstruct the shape of the object. This method is powerful, but it changes the nature of the errors. In the past, errors often came from the shape of the light beam itself being distorted. With the lantern, the errors come from how the light is measured and recorded. The researchers found that if the software extracting the data from the camera isn't perfectly accurate, it can make the light in one channel look slightly different from the light in another, even if the source is a single, unchanging point. This creates a fake signal that mimics the movement of a real cosmic object.

To solve this, the team had to look closely at the camera sensor itself, a highly sensitive device known as a qCMOS detector. They discovered that this camera behaves strangely when the light is very dim. In bright light, the camera counts photons accurately, but in the low-light conditions required to freeze the shaking of the atmosphere, the camera becomes non-linear. This means it undercounts the number of photons, and it does so differently for every single pixel on the sensor. Some pixels might count only half the light they receive, while others count almost all of it. When the researchers tried to measure the size of a star called Beta CMi, this uneven counting made the star appear larger than it actually was, creating a false impression of its structure. It was as if the camera's own internal inconsistencies were painting a picture of a bigger object than was really there.

The solution involved a careful, step-by-step process of calibration and correction. First, the team built a detailed map of how the camera responds to light at every level of brightness, using data from flat-field images taken at different exposure times. They found that by averaging many frames together, they could apply a mathematical correction to each pixel, forcing the camera to report the true number of photons even in the dimmest conditions. They also refined the way they extracted the light spectra, using a forward-modeling approach that accounted for the unique shape of the light pattern for each of the nineteen channels. By combining these corrections into a new, open-source software package called PLred, they were able to strip away the artificial signals.

The result was a clear, reliable measurement of the star Beta CMi. Before these corrections, the data suggested the star had a large, extended structure, but once the instrumental artifacts were removed, the elevated size signals across the Hα line were significantly reduced. The team demonstrated that they could achieve a precision of 50 microarcseconds, a level of accuracy that allows astronomers to resolve details on scales previously thought impossible from the ground. This work does not just fix a single observation; it establishes a new standard for how data from photonic lanterns should be handled. By understanding and correcting for the subtle quirks of the detector and the extraction process, the researchers have cleared the path for future instruments to explore the smallest, most distant corners of the universe with confidence, ensuring that what they see is the universe itself, and not the shadow of their own tools.