Google Earth screenshot of wrongly stitched photos of bridge over the river

Seeing the world through Google's eyes

An article that looks at Google's various platforms and technologies and how they capture everything indiscriminately, sometimes resulting in bizarre situations being recorded.

Before 2005 the only people to have seen the entire world from above were astronauts and those who viewed whatever grainy imagery they sent back down.

Since satellite imagery became more readily available in 2005, images taken from Google Earth, Google Street View, along with other platforms such as Bing Maps, have been used to see from above to highlight prison complexes, factory farming, environmental degradation and secret military bases.  

In this article we first look at how these various platforms and technologies are capturing everything indiscriminately, sometimes resulting in bizarre situations being recorded or the uncovering of hidden infrastructure, violence or degradation. We then focus on a few examples of Google Earth and Bing Maps being used in investigations and look at how these investigations were carried out, the tools used and any legal or copyright issues to consider.

Recording everything from glitches to hidden infrastructure

Imagery from above is mostly being taken by unmanned machines who are charged with capturing everything, whether they are satellites for Google Earth or in the case of Google Street View by 15-lens cameras attached to the top of cars.

These cameras record everything, with no particular significance given to anything. This can result in images of bizarre situations or the uncovering of evidence of hidden infrastructure, violence and degradation. For example, the series of images collected by Jon Rafman who curates a series of strange situations captured by Google Street View, one of which is shown below.

Tiger in 1500 Pearl St, Boulder, CO, United States by Jon Rafman. On later investigation the tiger featured in the image was made out of fibreglass and was a permanent fixture outside this shop until recently

Another example of these glitches in the system comes from Nicholas Mirzoeff in his 2015 book How to See the World who explains that "Google Earth and Street View use a process called 'stitching' to link enormous numbers of individual images into what appears to be a continuous depiction. At certain points, the illusion in these softwares breaks down because of a glitch in the system."

Since 2010 Clement Valla has been finding these 'stitching' glitches which he put together for a series entitled Postcards from Google Earth

Clement Valla, Postcards from Google Earth, 2010

Uncovering evidence

There are various artists and activists who are mining Google Earth, Google Map and other platforms such as Bing Maps to find satellite imagery that shows something about the world that was otherwise hidden.

Mishka Henner used satellite imagery to illustrate the scale of the beef industry in the United States in his 2013 series Feedlots. Henner's images demonstrate the environmentally destructive nature of industrial beef farming.

Initially Henner was researching satellite images of oil fields in the US, which is when he came across “strange-looking structures, like a big lagoon, or all these dots that look like microbes”, as he told Fast Company. At first it's easy to miss the tiny specks that show the cows that live here.

Tascosa Feedyard, Mishka Henner (2013)

Josh Begley was also interested in an industrialised US system, this time the prison system. Through using a database created by the Prison Policy Initiative and Prisoners of the Census, he located the latitude and longitude of every prison facility in the United States. This includes state and federal prisons but also local jails, detention centres and privately-run facilities.

Josh, however, wanted to have a greater sense of these institutions and to find out what the geography of incarceration in the prison capital of the world actually looked like in reality. Through using Google Maps he used these coordinates and collected satellite images of 4,916 facilities and displayed 700 (14%) of these images on online.

Josh Begley Prison Montage

Grid view of 35 prisons featured in Josh Begley's Prison Map

In 2013 Wired reported on a possible secret US drone base in Saudi Arabia based on satellite imagery found on Bing Maps. A former American intelligence officer told Wired “I believe it’s the facility that the U.S. uses to fly drones into Yemen. It’s out in eastern Saudi Arabia, near Yemen and where the bad guys are supposed to hang out. It has those clamshell hangars, which we’ve seen before associated with U.S. drones.”

Through checking the dates of the satellite images from the source (this function is explained in greater detail here, Wired wrote, “the commercial imaging company Digital Globe flew one of its satellites over the region on Nov. 17, 2010, there was no base present. By the time the satellite made a pass on March 22, 2012, the airfield was there. This construction roughly matches the timeline for the Saudi base mentioned in the Post_and in the _Times.”

These satellite images are only available on Bing Maps. In Google's repository of satellite images a message returns to the user to say 'Sorry, we have no imagery here.”

Satellite images taken in 2013 from Bing Maps in Saudi Arabia as published by Wired

The How

No data source is an island 

The question here is how are these images found? Is it through manually trawling through Google Maps and Street View for hundreds of hours? Or is it through getting 'lucky' and finding these images?

For the most part, these images in the examples above were found through first identifying locations through using other sources such as online databases and eye-witness testimonials, cross-corroborating these sources through various methods and then specifically searching for these locations through using various satellite imagery platforms.

For example, Josh Begley, when working on making visible the locations of United States military installations, first got the data from various journalistic reports, which he then, with the longitudes and latitudes already identified, searched on Google Maps for these specific locations. 

Mishka Henner identified locations for his Feedlots images by consulting State and National Feedyard directories and associations available online. He told us that, as far as he could remember, “I looked at every feedlot throughout the US, large and small. Only in the latter stages did I focus on seven large feedlots. These weren't the largest in the US but they suited what I wanted to convey. The good thing about feedlot association directories is they often include data about a feedlot's capacity and the type of livestock they house.”

He further explores Google Earth through his 2011 project Dutch Landscapes, which exhibits Google Earth images of government and military locations that have been camouflaged, rather stylistically compared with other more subtler governments, by the Dutch government. 

Mishka Henner, Dutch Landscapes, 2011.

For his Dutch Landscapes project he uses crowd-sourced data that had been uploaded to the GE Community and then he went online to search for local landmarks, barracks, and so on to corroborate the data and spent time checking the locations were accurate. Henner's process is echoed by James Bridle, who as part of his interview series with Exposing the Invisible talks about the process of uncovering hidden landscapes:

"There's something fascinating to me about that, that it was uncovered almost inadvertently by civilian technology. We've put this incredible array of technological power into space, satellites are constantly photographing the Earth, and it's a huge amount of information that any of us can access. But it still requires very human research, it still requires us to think about where to look and spend genuine time exploring it.  There is an incredible unconscious collaboration possible there between alternative information gathering systems and directed human research."

Finding the imagery 

"I'm equipped with the lenses of orbiting satellites and roving Street View cars, suddenly able to see things that I would never be able to see walking around on the ground with the camera." Mishka Henner

In our how-to, Starting Satellite Investigations, we introduce how to start thinking about finding satellite imagery online. For these examples featured above, mostly tools such as Google Maps or Bing Maps are used rather than downloading satellite imagery from landsat registries. This is often due to a different set of skills being needed with landsat registries in terms of knowing how to access the imagery and how to make sense of it. However, if you find something worth taking a closer look at and have the capacity to take some time to learn, or work with a geographer, Landsat and Sentinal data are highly recommended as they are free and open to the public.

Many of these platforms such as Google Maps and Bing Maps also have API's for people to utilise. For example Josh Begley for his Prison Maps project used the Google Maps API as it allows users to enter any latitude and longitude into its Static Maps service and it will produce an image of that location. Josh goes onto explain here that by writing a simple script in Processing he automated the process of collecting all the available images of 4, 916 facilities rather than doing it manually. 

Different platforms do contain different imagery. For example the earlier Saudi Arabia image of a possible drone base was only visible through using Bing Maps, on Google Maps a message returns to the user to say 'Sorry, we have no imagery here.” Other images are taken at different time periods which can show a different landscape resulting from an environmental disaster or a change in landscape resulting from war.

Copyright and legal issues 

Often images of industrial farming are difficult, if not impossible, to take or disseminate. There have been various reports of photographers being arrested for trespassing while attempting to document industrial farming. “Ag-gag” laws in the U.S. attempt to make taking photos or recording video illegal in animal farms. This makes Henner's images even more important as the aerial shots are made up of satellite imagery that is free to use for non-commercial purposes, so can be distributed. These images taken from above could be protected by the legal recognition that the altitude that satellites take images is considered public space. 

Exposing the Invisible asked Mishka Henner whether his work has been used to illustrate industrial farming by legal and advocacy groups. He replied “I've received requests almost weekly since the series was launched in 2013 and have struggled to keep up with demand. The work has appeared in medical journals, healthy eating campaigns, on PETA's social media campaigns and even in Meat Packing Journal - a magazine dedicated to serving the US meat industry. I generally send images out for free or tell people to just copy them from my site.”

As a general rule for non-commercial use of their imagery (i.e. for revenue-generating purposes) Google are happy for people to use their maps and imagery, as long as their terms of service are being followed and there is the proper attribution. The guidelines for attribution can be read here.

Many of the images you'll see not only feature the visible Google copyright icons but also still contain, in the case of Google Street View, the white directional compass arrows and zoom icons that continue to remind viewers where these images originally come from, never letting the viewers forget the eyes that Google has on the world. 

From the Jon Rafman Google Street View collection