A Million ISS Images
We’ve always been a fan of ISS Notify… but when the open.NASA team saw the latest project from Nate Bergey, we wanted to get him to share it directly with you. This project exemplifies why we share our open data with the public as well as the kind of inspiration we’re anticipating at the upcoming Space Apps Challenge. Hopefully this will be the first in a long series of posts by Nate and other innovators who are using space data to create great things.
I’ve always been fascinated with space. I’ve done plenty of looking up at the skies throughout my life, wondering what I would see if I were up there looking back down below.
I use NASA data every day, even though I don’t work for NASA. They have many detailed and rich datasets available for reading/viewing by the general public. The problem though is that they are often published in archaic formats buried under hard to use interfaces. But with a little ingenuity, anyone can get their hands on amazing data about the universe we live in.
For instance, I love the photographs that are being downloaded from the International Space Station’s cameras. Each one inspires me to dream of floating high above our blue planet. Sometimes, I wonder, what do the astronauts mostly take pictures of? In order to find out, I had to get some data.
The Image Science & Analysis Laboratory, at Johnson Space Center in Houston, Texas, maintains a [giant database of all the photographs astronauts have ever taken in space]. Luckily, they also have a database search page that allows you to ask for information about photos from particular missions, and features all kinds of metadata.
It’s possible to screen scrape the search result pages, using popular programing languages like python and tools like Beautiful Soup, creating your personal mini database of the data you care about. In my case, it was the approximate latitude and longitude of the photos taken from the ISS, or at the very least, the position of the station when the photo was taken.
So, I wrote some python scripts to do this for me a few pages at a time, and I ended up with the location of 1,129,177 photos!
The next step was to put the data on a map. To do this, I decided to draw each photo as a single pixel on a blank image. This is actually really easy in python using cairo.
The result is a giant map of Earth, based only on images from the ISS!
Most of the photos are taken of land masses. Coastlines, islands and cities seem to be popular targets. So much, that it’s possible to identify outline of continents. This makes sense, photos of clouds over an otherwise blank ocean get old after a while. I’m sure every astronaut has taken at least one photograph of the town they grew up in.
Now, let’s divide up the dots by mission. Is there any pattern? Here I draw each mission in a different color:
The map is dominated by purple, light blue, and green with some yellow. Also notice that the purple dots make almost uninterrupted orbit lines while most of the other dots seem to fill in randomly. This is because during Expedition 30/31 Don Pettit took dozens of amazing time lapse sequences consisting of hundreds of images taken continuously as the ISS orbited. In fact he’s single-handedly responsible for almost half the images taken on orbit!
One more thing we can do is add a map underneath to see exactly how the photos line up:
You can see that the ISS stays between about 50° and -50° latitude as it orbits the Earth The inclination of the orbit of the station is in fact 51.6°.
Since it’s hard to see the overlapped colors in the above image here is a collection maps with the photos from each mission mapped separately:
This is just some of the data that NASA has to offer. If you’re interested in getting involved in things like this I suggest taking a look at data.nasa.gov and International Space Apps Challenge.
You can also download the data I scraped from my github page.
[giant database of all the photographs astronauts have ever taken in space]: http://eol.jsc.nasa.gov/