The 800,000 Notifications: How the Vera Rubin Observatory is Breaking Astronomy

On the evening of February 24, 2026, a telescope situated atop a mountain in northern Chile silently accomplished something that had never been accomplished by any device in human history. In just two minutes, it observed 800,000 things that had changed since it had last looked at the sky and informed everyone about them. Not a synopsis. Not an example. Eight hundred thousand distinct cosmic alerts were sent out like phone notifications, each pointing to something that had moved, brightened, dimmed, or appeared where nothing had previously been.

At that number, it’s difficult not to stop. Tens of millions of objects have been found by optical observatories over the course of centuries of astronomy. Once fully operational, the Vera Rubin Observatory anticipates matching or surpassing that number in a single year.

Located on Chile’s Cerro Pachón, a high-altitude ridge swept by dry, stable air, the observatory offers one of the cleanest views of the southern sky on the planet. The LSST Camera, the largest digital camera ever built at 3,200 megapixels, is the focal point of the facility, which is a squat, industrial structure rather than the romantic dome of popular imagination. During the night, it swings to a different area of the sky every 40 seconds, snaps a picture, and then sends the data northward for a few seconds via fiber-optic transmission to the U.S. Data Facility at SLAC National Accelerator Laboratory in California. There, it is automatically compared to a previously taken template image of the same area. An alert is set off by anything that deviates.

From shutter to scientist’s screen, the entire process takes less than two minutes.
The size of the mirror or even the number of megapixels isn’t what really sets this telescope apart from all others. It’s the rhythm. Long exposures, meticulous planning, and years of waiting for telescope time have all historically made astronomy a patient discipline. That’s not how Rubin operates. For ten years, it functions more like a surveillance system for the cosmos, monitoring every area of the southern sky that is accessible on a rotating basis, night after night. Rubin is the continuous camera for the Legacy Survey of Space and Time (LSST), which is essentially a ten-year time-lapse of the universe.

Supernovae, variable stars whose brightness pulses in and out on regular cycles, active galactic nuclei—the glowing cores of far-off galaxies where supermassive black holes are actively consuming matter—and a variety of asteroids threading through the solar system were among the first batch of alerts received that February night. There will be new discoveries of some of those asteroids. Others will be well-known objects whose locations Rubin has updated with new accuracy. Eventually, a few may prove worthwhile to watch for more personal reasons. It is anticipated that Rubin will significantly enhance humanity’s list of potentially dangerous near-Earth objects, providing planetary defense scientists with a far more comprehensive dataset.

Speaking with the scientists involved gives the impression that the sheer amount of data is both the goal and the difficulty. According to Eric Bellm, who oversaw the team developing Rubin’s alert production pipeline, real-time discovery across ten terabytes of images every night requires years of technological innovation, not only in image processing but also in the databases and orchestration systems that support the entire system. Most engineers have never encountered a software issue involving processing that much sky so quickly.

A network of automated broker systems—software platforms with machine learning algorithms that filter, sort, and classify incoming alerts before directing them to the research teams or observatories most suited to follow up—is the answer to navigating the deluge of alerts. Rubin’s output is currently handled by seven official broker systems, each with a distinct area of expertise. Some concentrate on the early detection of supernovae. Others focus on objects in the solar system. The idea is that, rather than being overwhelmed by the torrent, a human scientist or a ground-based telescope with a limited field of view should be able to receive a curated, pertinent stream from it.

The fact that the alerts are public is noteworthy. The same data used by professional astronomers is available to anybody with a connection to one of the broker platforms. Integrations that would enable non-specialists to categorize cosmic events and make significant contributions to discovery are already being planned by citizen science platforms such as Zooniverse. It’s still unclear how much of the final science will come from that community, but it’s no longer science fiction that a high school student in Nairobi might identify an unusual variable star for expert investigation.

In the excitement surrounding supernovae and asteroids, Rosaria Bonito, a researcher at the Italian National Institute for Astrophysics, has highlighted one effect of ongoing observation that is often forgotten: young stars. Young stars are unpredictable. They brighten and dim for unknown reasons, flare unexpectedly when falling matter crashes onto their surfaces, and these events can be fleeting enough to elude even careful observers. Since Rubin catches everything, he will catch them. That adds up to a dataset on exceptional youth over a ten-year period that no other method could generate.
The larger scientific goals are noteworthy. Deep, extensive, and long-term surveys are essential for understanding dark energy, dark matter, and the universe’s large-scale structure. All three will be included in Rubin’s survey. It’s possible that the dataset will eventually uncover phenomena that no one has yet considered looking for, which is what scientists say when they can’t quite contain their excitement but also don’t want to overpromise.

In a sense, the question of what will happen after the initial 800,000 alerts is equally fascinating. The actual LSST has not yet begun. The February detections were not the start of the survey; rather, they were a significant step toward its completion. The alert rate is anticipated to increase to seven million per night once Rubin reaches its operational stride later this year. The scale at which the broker systems will be tested is unprecedented. A truly new issue will confront the astronomy community: not a lack of data, but nearly the opposite. It’s difficult not to feel that the field is at the beginning of something for which it hasn’t fully prepared as we watch this develop over the next ten years; historically, this is precisely when the greatest discoveries tend to occur.