Home North pole ice Look into the shadows of the moon with AI

Look into the shadows of the moon with AI



Permanently shaded lunar craters contain water ice, but are difficult to image. A machine learning algorithm now provides sharper images.

The Moon is a cold, dry desert. Unlike the Earth, it is not surrounded by a protective atmosphere, and the water that existed during the formation of the Moon has long since evaporated under the influence of solar radiation and escaped into the space. Nevertheless, the craters and depressions of the polar regions give hope for limited water resources. Scientists at MPS, the University of Oxford and the NASA Ames Research Center have now taken a closer look at some of these regions.

“Near the lunar north and south poles, incident sunlight enters craters and depressions at a very low angle and never reaches some of their floors,” explains MPS scientist Dr. Valentin Bickel, first author of the new article. On this “Eternal Night”, the temperatures in some places are so cold that the frozen water is expected to last for millions of years. Impacts of comets or asteroids could have delivered it, or it could have been degassed by volcanic eruptions, or formed by the interaction of the surface with the solar wind. Neutron flux and infrared radiation measurements obtained by space probes in recent years indicate the presence of water in these regions. Finally, NASA’s Lunar Crater Observation and Detection Satellite (LCROSS) provided direct evidence: twelve years ago, the probe fired a projectile into the shaded crater at the South Pole Cabeus. As a later analysis showed, the dust cloud emitted into space contained a considerable amount of water.

However, permanently shaded areas are not only of scientific interest. If humans are ever to spend long periods on the Moon, natural water will be a valuable resource – and shady craters and depressions will be an important destination. NASA’s unmanned VIPER rover, for example, will explore the South Pole region in 2023 and enter such craters. To get a precise idea of ​​their topography and geology in advance – for mission planning, for example – images from space probes are essential. NASA’s Lunar Reconnaissance Orbiter (LRO) has been providing such images since 2009.

However, capturing images in the deep darkness of permanently shaded areas is exceptionally difficult; after all, the only sources of light are scattered light, such as that reflected off the Earth and surrounding topography, and faint starlight. “Because the spacecraft is in motion, the LRO images are completely blurry at long exposure times,” says Ben Moseley of the University of Oxford, co-author of the study. With short exposure times, the spatial resolution is much better. However, due to the low amount of light available, these images are dominated by noise, making it difficult to distinguish true geological features.

The 17 newly studied craters and depressions are located near the South Pole. While the smallest of these regions (region 11) has an area of ​​only 0.18 square kilometers, the largest (region 9) measures 54 square kilometers. Region 9 is not located in the section of the South Polar Region shown here, but a little further north, in the Schrödinger Basin. The representations of the lunar surface presented here are based on elevation data from the Lunar Reconnaissance Orbiter.]MPS / University of Oxford / NASA Ames Research Center / FDL / SETI Institute

To solve this problem, the researchers developed a machine learning algorithm called HORUS (Hyper-effective nOise Removal U-net Software) which “cleans up” these noisy images. It uses over 70,000 LRO calibration images taken from the far side of the Moon along with information about the camera’s temperature and the spacecraft’s trajectory to distinguish which structures in the image are artifacts and which are real. . This way, researchers can achieve a resolution of around 1 to 2 meters per pixel, which is five to ten times higher than the resolution of all previously available images.

Using this method, the researchers have now reassessed images of 17 shaded regions in the lunar south pole region that measure between 0.18 and 54 square kilometers. In the resulting images, small geological structures only a few meters in diameter can be discerned much more clearly than before. These structures include rocks or very small craters, which can be found anywhere on the lunar surface. Since the Moon has no atmosphere, very small meteorites repeatedly fall on its surface and create such mini-craters.

“With the help of the new HORUS images, it is now possible to understand the geology of the lunar shadow regions much better than before,” says Moseley. For example, the number and shape of small craters provide information on the age and composition of the surface. It also makes it easier to identify obstacles and potential dangers for rovers or astronauts. In one of the craters studied, located on the Leibnitz plateau, the researchers discovered a surprisingly bright mini-crater. “Its relatively bright color may indicate that this crater is relatively young,” says Bickel. Because such a fresh scar offers a fairly loose glimpse into deeper layers, this site could be an interesting target for future missions, the researchers suggest.

The new images do not provide any evidence of frozen water on the surface, such as bright spots. “Some of the regions we have targeted might be slightly too hot,” Bickel speculates. It is likely that lunar water does not exist at all as a clearly visible deposit on the surface – instead, it could be mixed with regolith and dust, or it could be hidden underground.

To answer this and other questions, the researchers’ next step is to use HORUS to study as many shaded regions as possible. “In the current post, we wanted to show what our algorithm can do. Now we want to apply it as completely as possible, ”says Bickel.

This work was made possible by the Frontier Development Lab (FDL.ai). FDL is a cooperation agreement between NASA, the SETI Institute (seti.org) and Trillium Technologies Inc, in partnership with the Luxembourg Space Agency and Google Cloud.

More information

Please follow SpaceRef on Twitter and Like us on Facebook.