Every five to six minutes, another one, over five to seven Earth hours.
Half a century after Apollo astronauts departed the Moon, the faint mechanical whispers they left behind have finally been given a voice. Caltech researchers, armed with machine-learning tools unavailable to their predecessors, returned to seismic recordings gathered in 1976 and 1977 and found that the Moon's mysterious tremors were never truly mysterious — they were the predictable language of metal and dust responding to sunlight. In decoding this rhythm, science has transformed an old enigma into a practical guide for the humans who will one day return.
- Decades of unexplained lunar tremors had left a quiet gap in planetary science — regular as clockwork, yet stubbornly resistant to explanation.
- Machine-learning models applied to dormant Apollo 17 seismic archives cracked the pattern open, revealing two distinct classes of moonquake hiding within the noise.
- The most striking culprit turned out to be the abandoned lunar lander itself, popping and creaking under thermal stress every five to six minutes for hours at a stretch each lunar morning.
- Beneath that mechanical chatter, the lunar regolith was conducting its own quieter trembling, driven by the extreme temperature swings of the Moon's day-night cycle.
- The findings land with immediate consequence: future Artemis astronauts will operate on a surface whose tremors are now predictable enough to serve as a natural timer for surface operations.
Fifty years after the last Apollo astronauts left the Moon, the seismic recordings they left behind have finally yielded their secret. Four geophones planted near the Apollo 17 landing site captured tremors arriving with metronomic regularity every lunar morning and afternoon — a puzzle that sat largely unsolved in archives since 1977. A Caltech team decided to revisit that old data with modern machine-learning models, and what had seemed mysterious dissolved into something both simple and elegant.
The signals divided cleanly into two types. The first were sharp, high-amplitude jolts originating from the Apollo 17 lunar module itself, a few hundred meters away. As sunlight struck the metal structure at sunrise and sunset, thermal expansion sent a pop through the seismometers every five to six minutes, for up to seven hours at a stretch — a mechanical complaint playing out in the vacuum with perfect regularity. The second type was subtler: the lunar regolith responding naturally to the Moon's extreme temperature swings, its tremor duration tracking directly with surface heat.
The discovery carries weight beyond the satisfaction of closing a half-century-old question. Published in the Advancing Earth and Space Sciences Journal, the findings offer a practical gift to NASA's Artemis program: a lunar surface whose tremors are now predictable enough to inform spacecraft design and help mission planners schedule surface operations. The quakes are far too faint to be felt underfoot, but understanding their origin turns them from a curiosity into a resource — a natural rhythm that future explorers will learn to read before they ever set foot on the surface.
Fifty years after astronauts left the Moon, scientists have finally figured out what's been rattling the lunar surface all along. The answer was hiding in plain sight—in seismic recordings that sat largely unexamined since 1977, waiting for the right tools to decode them.
Four geophones planted near the Apollo 17 landing site captured something peculiar: tremors that arrived with metronomic precision, every lunar morning and afternoon, like clockwork. For decades, researchers knew these moonquakes existed but couldn't explain them. Now a team from Caltech has reanalyzed that old data using machine-learning models, and the mystery has dissolved into something both mundane and elegant: the Moon's surface shakes because the Sun heats it, and the spacecraft left behind by astronauts expands and contracts in response.
The seismic signals fell into two distinct categories. The first—sharp, high-amplitude jolts—originated from the Apollo 17 lunar module itself, located a few hundred meters from the instrument array. When sunlight hit the lander at lunar sunrise and sunset, the metal structure began to heat rapidly. Every five to six minutes, for five to seven Earth hours straight, another pop would register on the seismometers. "Every lunar morning when the sun hits the lander, it starts popping off," explained Allen Husker, a co-author of the study. "Every five to six minutes, another one, over five to seven Earth hours. They were incredibly regular and repeating." These weren't tremors in any traditional sense—they were the sound of a spacecraft expanding under thermal stress, a mechanical complaint that played out in the vacuum with perfect predictability.
The second category of moonquakes proved more subtle. These emergent events arose from the lunar regolith itself, the dusty surface material responding naturally to the temperature swings that accompany the Moon's extreme day-night cycle. The duration of these tremors correlated directly with how hot the ground became, suggesting they might stem from changes in how the regolith scattered seismic waves, or from more energetic daytime events occurring at greater distances.
What makes this discovery significant extends beyond the satisfaction of solving a half-century-old puzzle. The researchers published their findings in the Advancing Earth and Space Sciences Journal, and their work carries immediate practical weight. Future crewed missions under NASA's Artemis program will land on the lunar surface knowing that these tremors will occur with absolute regularity—predictable enough that astronauts could, in theory, use them as a natural timer. The quakes are far too faint to be felt by a human standing on the Moon, but understanding their origin and frequency will inform how engineers design new landers and how mission planners schedule surface operations.
The Apollo 17 seismic data, collected between October 1976 and May 1977, had remained largely dormant in archives until Caltech's team decided to revisit it. The Moon is the only planetary body besides Earth to have hosted multiple seismometers simultaneously, making it a unique laboratory for understanding how other worlds respond to thermal and mechanical stress. As Husker noted, that opportunity is precious: "It's important to know as much as we can from the existing data to design experiments and missions to answer the right questions." The tremors that puzzled scientists for decades have become a resource—a window into how spacecraft behave in an alien environment, and a natural rhythm that future explorers will learn to read.
Notable Quotes
Every lunar morning when the sun hits the lander, it starts popping off. Every five to six minutes, another one, over five to seven Earth hours. They were incredibly regular and repeating.— Allen Husker, Caltech researcher
It's important to know as much as we can from the existing data to design experiments and missions to answer the right questions. The Moon is the only planetary body other than the Earth to have had more than one seismometer on it at a time.— Allen Husker
The Hearth Conversation Another angle on the story
So these moonquakes—they're not actually quakes at all?
Not the ones from the lander, no. Those are just metal expanding and contracting. But the ones from the regolith itself are real seismic events, driven by temperature. The Moon swings wildly between day and night, and the ground responds.
Why did it take fifty years to figure this out?
The data was there, but the tools weren't. Machine learning let them sort through the noise and see patterns that human analysis might have missed or taken much longer to spot. Sometimes the answer is waiting for the right question.
Is this dangerous for future astronauts?
Not at all. The tremors are too faint to feel. What matters is that they're predictable. Astronauts can plan around them, and engineers can design landers knowing exactly how thermal stress will behave.
What does this tell us about the Moon itself?
That it's mechanically responsive in ways we're still learning. The regolith isn't inert—it's constantly adjusting to temperature changes. Understanding that helps us understand the Moon's interior structure and composition.
Could this apply to other planets?
Absolutely. Any body with a surface and a day-night cycle will experience thermal stress. This is a template for understanding how other worlds respond to their environments.