Crowdsourced Moon maps get accuracy approval

Crowdsourcing already gets products off the ground and figures out where traffic congestion is, but CosmoQuest is turning the power of group-work to map the moon. Using high-resolution images beamed back from NASA's Lunar Reconnaissance Orbiter (LRO), the public science project allows anybody to register and then sift through, flagging up craters, boulders, and other features of the rocky surface. Now, new research indicates the crowdsourced mapping can be just as accurate as when trained experts do it.

The study, led by a team at the University of Colorado-Boulder, did statistical comparisons between CosmoQuest's MoonMapper project and the handiwork of eight experts.

Researchers found that both the expert and crowdsourcing groups showed widely ranging differences in how many craters were counted, compared to their in-group counterparts. For instance, even when looking at the same image, some experts counted twice as many craters as other experts saw.

However, averaged out across the groups, the populations of craters spotted proved to be statistically similar. Although the experts might be better individually than non-experts, when scaled up to crowdsourced-size projects, the inexperience of the non-experts was smoothed out.

"Our study results mean we can now use the power of crowd-sourcing to gather more data than we ever thought possible before" Research Scientist Stuart Robbins of CU-Boulder's Laboratory for Atmospheric and Space Physics said of the findings. "A very large group of volunteers was able to chart these features on the moon just as well as professional researcher."

Although it may seem like we already know what the moon looks like, there's comparatively little understanding of the actual topography and how it was shaped by the formation of the solar system. In a sense, scientists say, it's a look back in time to a more tumultuous period, the evidence of which has not been erased by plate tectonics as on the Earth.

IMAGE: NASA