When a star reaches its final stages of life, it ends in a bright and powerful explosion called a supernova and scatters material across the universe. A new study led by a University of Hawai‘i at Mānoa researcher discovered that the Moon’s mixed-up soil can be read as a cosmic time capsule for exploding stars.
Emily Costello, research scientist at the Hawai‘i Institute of Geophysics and Planetology in the UH Mānoa School of Ocean and Earth Science and Technology , and co-authors acted as physicist-cryptographers and developed a mathematical model to decode the history recorded in the lunar surface.
“Deep-sea deposits on Earth preserve interstellar debris, but only back about 10 million years,” Costello said. “The lunar regolith [soil], however, acts as a long-term cosmic archive that can preserve history spanning 80 to 100 million years or more. Understanding the physics of regolith mixing ensures that when future astronauts return deeper cores, we can properly read the scrambled layers to reconstruct the history of our Solar System's journey through the galaxy.”
Costello and co-authors developed a “unified stochastic model”, a specialized model that unscrambles ‘impact gardening’, the continuous process where crater-forming impacts flip, mix, and redistribute the Moon's surface soil over time. This is a highly random process driven by meteorites ranging in size from microscopic grains of dust to giant asteroids. Because any single lunar core sample is affected by the unique impact history, separating the broader interstellar signal from local variability requires highly sophisticated statistics.
“To model impact gardening, we have to balance a complex web of physical mechanisms, including impact compaction, excavation, radioactive decay, and space weathering, all operating simultaneously within a single elegant continuum model,” Costello said. “Our mathematical model treats lunar impact gardening as a competition between forces burying the soil and impacts digging it back up. It also accounts for radioactive decay of the star remnants while mapping exactly when and where new stardust was delivered by episodic supernovas.”
Based on radioactive isotopes in deep-sea sediments on Earth and in lunar soil samples returned by Apollo, it is known that supernova explosions hundreds of light-years away scattered radioisotopes across the Earth and Moon, with pulses of activity about 2.3 million and 7.3 million years ago. But once the radioisotopes arrived, they started getting mixed into the lunar surface by impact gardening.
Costello validated her model using real-world data, showing that the model reproduced depth-concentration profiles of the radioisotopes from Apollo core samples whose ages were independently constrained by cosmic ray tracks and radionuclide benchmarks. To further test the model, the team then combined the validated model with known historical timelines of supernova pulses on Earth to forward-model how those exact interstellar events would be preserved at different depths in the lunar soil.
The team discovered that the model can accurately predict the depth-concentration profiles of Iron-60 found in Apollo regolith samples. They then extended the model to predict how other heavy elements, such as Plutonium-244, Iodine-129, Hafnium-182, and Curium-247, are buried over time.
“When I first shared my model results, my colleagues were surprised by how well-matched the model and the measurements were,” Costello said. “This level of fidelity between empirical observations and a physics model is exciting and remarkable.”
A main goal of the study was to provide guidance for future core samples that will be collected on the Moon. In the near future, NASA’s Artemis Program will return humans to the Moon, and they will bring back new precious samples of lunar regolith which preserve the history of stardust.
“These future samples taken from the moon, when considered in conjunction with our gardening model, could reveal new insights into an untold chapter of supernova history!” Costello shared. “I think it’s beautiful that the remains of past stars can be used to navigate the vast history of our Earth-Moon neighborhood, if we have knowledge of how to read the stardust.”
Physical Review Letters
Computational simulation/modeling
Not applicable
Gardening on the Moon: An Advection-Diffusion Model to Guide the Search for Supernova Debris in the Lunar Regolith
14-Aug-2026