The moon has many more craters than we thought, new research has shown.
More than 109,000 new craters have been discovered in low- and medium-latitude lunar regions using artificial intelligence (AI) that received data collected from Chinese lunar orbiters.
The number of craters recorded on the lunar surface is now more than a dozen times larger than it was before. The findings were published Dec. 22 in the journal Nature Communications.
“It is the largest lunar crater database with automatic extraction for medium- and low-latitude moons,” said study lead author Chen Yang, associate professor of earth sciences at Jilin University in China, in an email to Live Science.
Related: China on the moon! History of Chinese lunar missions in pictures
Impact craters, formed during a meteor impact, cover most of the Moon’s surface.
Impact craters can be considered the lunar equivalent of a “fossil,” which “records the history of the solar system,” Yang said.
However, these “fossils” can vary dramatically in size and shape, and over time they can overlap and erode. This makes recognizing and getting to know them extremely difficult and time consuming. The process is also subjective, leading to inconsistencies between existing databases.
Yang and her team approached these problems using machine learning. They trained a deep neural network (where a computer uses layers of mathematical calculations that fit together) with data from thousands of previously identified craters and learned an algorithm for finding new ones. The network was then applied to data collected by lunar orbiters Chang’e-1 and Chang’e-2, revealing 109,956 additional craters on the lunar surface.
A significant number of craters identified in this study are classified as “small” to “medium” in size, although from an Earthling perspective they are still quite large, ranging from 1 to 100 kilometers. diameter. The relatively small size of the crater is probably the reason why they were not discovered earlier.
But the AI program also spotted much larger irregularly shaped craters eroding – some of them up to 550 km in diameter.
The algorithm was also estimated when nearly 19,000 craters were formed based on their characteristics, such as size and depth, and by assigning each geological time period. These craters span all five lunar lunar geological periods, and some go back about 4 billion years.
The team hopes to improve its crater detection algorithm by sending it data from the recently launched Chang’e 5 spacecraft, which recently returned lunar samples to Earth.
Researchers also want to adapt and apply their machine learning approach to other bodies in the solar system, including planets like Mars.
“This prediction will typically take several minutes, followed by several hours of post-processing on standard computer hardware,” the researchers wrote in the study.
Originally posted on Live Science.