Since the International Space Station (ISS) has been in orbit, astronauts have taken images of our Earth in all of its unending beauty. With some technological improvements, the images have greatly improved from the first snaps. There are over 1.3 million images taken from the ISS that are available to the public through The Gateway to Astronaut Photography of Earth. Of all of these amazing images, 30% of them were taken at night. The resolution of images from the ISS is quite good, but the downside is that very little location information is associated with them. To deal with this issue, Complutense University of Madrid (UCM) started a crowd souring project called, “Cities at Night” that is composed of three sub projects.
The most basic of the three is called Dark Skies, where the public can logon, sort through images and catalog them as of cities, stars or other objects. The human eye is more capable of discerning the difference between a star field and a city at night then any algorithm. No expertise is required, so anyone can participate in this particular project to help weed out photos that are not of Earth at night.
The next level up is the Night Cities project where people can review night images of their own cities and identify the location of certain points and where they correspond on Earth. The idea behind this is that a person is more likely to identify areas and objects in the places where they are most familiar, so people can focus their attention on areas where they are likely to have the most success. These images will be used to create light maps of cities.
The most tedious and difficult project is Lost at Night. When an image is taken from ISS, the only information collected is where the station is in orbit, the camera however could be pointing in any direction. This leaves an area that is 310 square miles from which the image could have been taken in. Some large cities are easy to identify, while other areas are more rural with just a spattering of light, and little indication of its location. The user’s job then is to identify where the center of the image is on a map so it can be given a more accurate location.
As you can imagine, it will take a plethora of individuals to correctly classify each image to increase the accuracy. Volunteers can log on to these systems and start helping Cities at Night. The intention is to create a library of images for researchers, the public and media to access at any time. Images at night help researchers identify different kinds of light pollution, calculate energy consumption and can be used in conjunction with outside data to assess city growth and health.