Millimeter-wave adaptive optics: Demonstrating closed-loop correction for lowest Zernike modes

The sky above a ground-based telescope is not a perfect, still window. Even on a clear night, the wind buffets the massive metal structures of modern observatories, and the sun's heat warps their surfaces in subtle, shifting ways. For telescopes that listen to the universe in millimeter and submillimeter waves—a range of light that sits between radio and infrared—these tiny distortions are catastrophic. They blur the image, turning a sharp point of light into a smeared mess. To see clearly, these giant dishes need to keep their surfaces perfectly smooth, down to the width of a human hair, despite the constant shaking and bending caused by the weather. This is the challenge of millimeter-wave adaptive optics: a system designed to measure these distortions in real time and correct them instantly, keeping the telescope's view sharp.

A team of researchers has now taken a significant step toward solving this problem by building and testing a new kind of "eye" for these telescopes. Working at the famous 45-meter telescope in Nobeyama, Japan, they developed a prototype sensor that can detect the lowest, most basic types of surface errors. By linking this sensor to a movable secondary mirror, they successfully demonstrated a closed-loop system that can sense a distortion and automatically push the mirror to cancel it out. This achievement proves that it is possible to keep the surfaces of future, even larger telescopes perfectly aligned, paving the way for clearer views of the cold, dusty regions of the cosmos.

The core of the problem lies in the sheer size of modern telescopes. As these instruments grow to capture faint signals from the early universe, their metal surfaces become more susceptible to the elements. Wind pressure and temperature changes cause the primary mirror to flex, creating waves of error across its surface. To fix this, the researchers turned to a method based on radio interferometry. Instead of trying to measure the mirror's shape with a ruler or a laser, they treated the telescope itself as a giant radio receiver. They placed small transmitters on the primary mirror at five specific points. These transmitters sent a reference signal through the telescope's optics to a receiver at the focal point. By comparing the signal received from each of the five points against a master reference, the system could calculate the exact extra distance the signal had to travel due to the mirror's distortion. This difference, known as the excess path length, revealed the shape of the wavefront error.

The team installed a five-element version of this sensor on the 45-meter telescope, operating at a frequency of 20 GHz. The five transmitters were arranged in a cross pattern, with one in the center and four extending out to the top, bottom, left, and right. This specific arrangement was chosen because it is the simplest setup capable of detecting the two most common and damaging types of distortion: tip-tilt, which is like the telescope pointing slightly off-target, and defocus, where the surface is too curved or too flat. These are the "lowest" modes of error, and they are the ones that can be corrected by moving the telescope's secondary mirror, which sits above the main dish.

To test if the sensor could actually see these distortions in the real world, the researchers conducted an experiment using the Moon. They pointed the telescope at the edge of the lunar disk. When a telescope points slightly off, the amount of the Moon's light it captures changes, a value known as the beam filling factor. If the telescope's surface is warped, the beam pointing will jitter, causing the total power of the signal to fluctuate. The researchers watched for these fluctuations while simultaneously recording the data from their new five-element sensor. They found a clear, strong connection between the two. When the sensor detected a wavefront gradient, the total power signal shifted in perfect sync. This correlation confirmed that the sensor was successfully measuring the very distortions that were causing the telescope to wobble, even as the wind blew at speeds of about 6 meters per second.

The final and most critical step was to close the loop. Detecting an error is only half the battle; the system must be able to fix it. The researchers connected the sensor to a computer controller that could move the secondary mirror. They set up a proportional-integral controller, a type of feedback system that adjusts the mirror's position based on the size and duration of the error. When they artificially introduced a disturbance equivalent to a low-order wavefront error, the system reacted immediately. The sensor measured the error, the controller calculated the necessary correction, and the secondary mirror moved to compensate. The result was a stable system where the error was suppressed, and the signal returned to normal.

Crucially, the system only corrected the errors it was designed to handle. When the researchers introduced a complex distortion that the secondary mirror could not fix, the system did not overreact or try to force a correction that was impossible. It remained stable, ignoring the uncorrectable modes. This selective stability is essential for future applications. The researchers note that this success establishes a foundation for metrology—the science of measurement—in future, massive submillimeter facilities, such as the proposed AtLAST or Large Submillimeter Telescope. By proving that a simple, five-point sensor can reliably detect and correct the most basic surface errors in real time, this work offers a practical path forward for keeping the next generation of giant telescopes perfectly focused on the universe.