New Moon Mapping Algorithm Leads to Discovery of 6,000 Craters
Researchers recently developed a moon mapping algorithm to make their job a little less insufferable. To their surprise, this led to the discovery of 6,000 previously unknown moon craters.
This new algorithm could potentially result in more remarkable discoveries through mapping other planets like Mars, Mercury, asteroids, or moons of entities like Jupiter. It could also help researchers study the solar system’s vast history.
Counting Craters One-by-One
Believe it or not, up until just a few weeks ago, astronomers needed to physically count every single moon crater. If you thought your job was annoying, imagine comparing black and white photos of the moon’s surface.
Postdoctoral fellow in the Centre for Planetary Sciences (CPS) at the University of Toronto Scarborough Mohamad Ali-Dib — rightfully — called the method “archaic.”
“Basically we need to manually look at an image, locate and count the craters and then calculate how large they are based off the size of the image. Here we’ve developed a technique from artificial intelligence that can automate this entire process that saves significant time and effort,” Ali-Dib said in a U of T press release.
How the New Moon Mapping Algorithm Works
Researchers at the University of Toronto Scarborough developed the moon mapping algorithm by using the same technology for self-driving cars. It’s the first algorithm of its kind to actually work correctly.
Similar technology could count and identify lunar craters based on provided data sets. When it came down to the point of analyzing new information, however, the algorithms always fell short.
You wouldn’t think developing an algorithm for monitoring images would be so difficult. After all, researchers have already developed AI to identify works of art by genre and even create their own art. Other projects in the works say they have the technology to translate animal speech to human. Images of the moon are pretty simple: just black and white photos of circles. Nonetheless, developing an algorithm for moon mapping wasn’t so easy.
The U of T research team combined a convolutional neural network (that’s the technical term for AI used with self-driving cars and computer robot vision) with data from satellites and elevation maps. None of the team had ever counted craters themselves before. I guess that’s a job typically left up to the most-hated interns.
Researchers trained the neural network with large chunks of information using data from 2/3rds of the moon’s surface. After that, they tested the technology on the remaining 1/3rd of the moon. They hoped it would work and they probably didn’t expect it to work quite this well.
In fact, the technology identified 6,000 previously unknown craters: that’s twice as many craters as manual counting has identified up to this point.
A New World of Possibilities
The moon contains tens of thousands of unidentified craters. Could you imagine counting each and every one of them by hand? It’s not even possible — half of them you can’t even see. But this technology isn’t limited to moon mapping. Next, the team plans to use this algorithm for studying the surfaces of Mercury, Mars, and other celestial bodies like the icy moons of Jupiter or Saturn.
Unlike the earth, moons and other planets don’t have plate tectonics, water, or even an atmosphere. This means that craters from over 4 billion years ago are still visible — and in pretty good shape — to this day.
Since the moon doesn’t experience erosion, you can actually determine a crater’s age by counting the number of small craters inside of it. This technique only works on planets without air. If you’re interested, an organization called CosmoQuest has open volunteer positions since studies show that amateurs can count craters just as well as 50-year veterans.
Mapping typographic features like craters provides much more information than you’d think. Researchers look at many factors when they study craters: their age, shape, size, distance apart, and so on. This actually provides useful information for learning about how the solar system formed and what happened during its early years.