What Satellite Images Can Show Us about ‘Re-education’ Camps in Xinjiang

A Q&A with Shawn Zhang

The government of the Xinjiang Uighur Autonomous Region, located in China’s northwest near Kazakhstan, Kyrgyzstan, and other Central Asian countries, is holding ethnic Muslims en masse in detention facilities the Chinese government euphemistically calls “transformation through education centers.” We know this from a year-long trickle of news reporting from Xinjiang: from citizens bearing witness to Radio Free Asia about what they’re seeing firsthand, from international reporters doggedly traveling throughout the region despite heavy surveillance and intimidation from security personnel, from released detainees describing their experiences in these camps, and from reports from relatives of those detained who are also coming under Chinese government pressure, as well as from more scholarly analyses looking at government procurement bids, public security hiring, and regional security spending that reveal the scope and scale of the Party-state’s mobilization of resources to establish these camps.

Yet, without free and consistent international access to all parts of Xinjiang, let alone to the facilities in question, the Chinese government remains at liberty to deny their existence—or at least their carceral nature. Indeed, Chinese officials recently did just that at a session of the United Nations Committee on the Elimination of Racial Discrimination, arguing that “[t]here are no such things as ‘re-education centers’ or ‘counter-extremism training centers’ in Xinjiang . . . With respect to criminals involved only in minor offenses, the authority provides them with assistance and education by assigning them to vocational education and employment training centers to acquire employment skills and legal knowledge, with a view to assisting in their rehabilitation and reintegration.”

Claims that the camps are merely vocational training centers seem even less credible after one looks at the work of Shawn Zhang. A law student focusing on jurisprudence at the University of British Columbia in Canada, in May Zhang began scouring Google Earth for evidence of detentions in Xinjiang—matching up the addresses he found in documents related to the camps that he found online with satellite images; learning to recognize the distinguishing characteristics of camps, both in written materials and in images so that he could confidently identify them even in cases where details were scant. His collection of satellite images is a key and oft-cited piece of evidence of the camps’ existence, laying bare the hard physical contours of a campaign of ethnic and religious repression. Zhang, a Chinese citizen, continues to do this research despite the fact that the Chinese police have previously pressured his family related to other content he has posted online.

Many of Zhang’s findings cannot be independently verified due to restrictions on international reporting and foreign government access (though recent reporting from The Wall Street Journal strongly suggests that Zhang’s work is indeed on the right track). But it is precisely because of these restrictions that Zhang’s research has been so vital to shedding light on what is happening in Xinjiang. The following is an edited transcript of several phone conversations I had with Zhang about his research.

Google Map compiled by Shawn Zhang of the locations of what he believes are re-education camps.

Jessica Batke: How did you get the idea to start looking for the re-education camps on Google Earth?

Shawn Zhang: Early this year, I saw people talking about these re-education camps on social media. At first, I thought they were maybe exaggerations, because it’s kind of horrible stuff to be happening right now in the 21st century. We’ve only heard of this kind of re-education camp in history textbooks. So I did some searches on Google to find out what was really happening there, and I found some government tender notices for construction [of “transformation through education” centers]. In the notice, it will say the construction address of the camps, like in which village, in what place, so I searched [Google Earth] to find them . . . and published [the images of these locations] on my blog and on Twitter. After that, I searched for more tender notices like [the first few I had found], and I located more and more camps in Xinjiang.

What are the government tender notices for? Are they all for soliciting construction bids?

Most of them are soliciting construction. Some are soliciting equipment or services.

And do the tender notices say jiaoyuzhuanhua [教育转化, “education through transformation”] or something like that? Do they indicate what the building is for?

Yes. Some tender notices say they are building several teaching buildings, [教学楼, jiaoxuelou], and several student dorms [宿舍, sushe]. Often they will note the construction area, like how many square meters for this or that building, so I can see if the size of the buildings from the satellite images matches the specs in the tender notices.

Are you still looking for these types of government tender notices? Do you have a sense that there are more out there? Or have you found most of them?

Most tender notices I’ve found are from 2017 or 2016. I think the Xinjiang government began to realize [that people were tracking these notices] and stopped publishing them. They also deleted a lot of tender notices—sometimes I’ll look back to see if I missed any information on an old tender notice and find that the link I had originally used doesn’t work anymore. So generally, it’s become more and more difficult to find tender notices that have location information for the re-education camps.

What is your process for confirming that what you’ve found is a re-education camp versus, say, a detention center, a prison, or a school—all facilities than can be used to house large numbers of people? [Detention centers (看守所, kanshousuo) are facilities run by police that hold people under investigation or prosecution for committing a criminal offense. Prisons (监狱, jianyu) are run under the auspices of the Ministry of Justice and hold people who have already been convicted at trial. As noted previously, “re-education camps” are extra-legal facilities holding ethnic and religious minority citizens in Xinjiang; the Chinese government has denied their existence.]

First, it’s relatively easy to tell the difference between detention centers and other buildings. Detention centers all over China look very similar. I can search the Internet for many detention centers by their addresses, like some detention centers in Urumqi, or in Beijing, or in Shanghai, or other major cities. They all look similar, so you can tell [what they are]. You can tell the difference between detention centers and other buildings.

You can also see the difference between prisons and other buildings because prisons have a much higher level of security. In Xinjiang, a lot of new prisons really look like schools. Many re-education camps also look like schools. Camps have school-like teaching buildings and dorms. The major difference between these three [types of facilities] is schools have very limited security facilities/equipment, and they usually only have metal fences at perimeters of the compound. Re-education camps have higher solid walls, and there are razor wire fences around the buildings inside the compound. Prisons have a much higher security level. The perimeter walls are stronger, with several additional layers of razor wire fences both inside and outside these walls.

Most re-education camps have less security than detention centers and prisons, but we can see razor wire fences inside the camps [for example, in Kona Sheher (Shufu) and Peyzawat (Jiashi) Counties]. And some camps also have watchtowers. Many public buildings also have razor wire fences on the outermost walls of their compounds, but they don’t tend to have razor wire fences around the buildings inside the compounds. But in the re-education camps, many of the buildings inside the compounds are surrounded by razor wire fences. This is a feature you can use to distinguish re-education camps from other buildings. For example, many vocational schools in Xinjiang have been turned into re-education camps. In satellite images from two or three years ago, there are no razor wire fences inside the campuses, but now we can see lots of razor wire fences surrounding the buildings of the vocational schools. The razor wire fences suggest that these vocational schools have been turned into re-education camps.

Satellite image of a re-education camp in Kona Sheher (Shufu) county, Xinjiang Uighur Autonomous Region. (Image courtesy of Google Earth)

Did you have previous experience looking at satellite images like this, or was this something that you taught yourself in the course of looking for re-education camps?

This was the first time I did so many searches for satellite images.

What is your confidence level in the images that you have identified as re-education camps? How confident are you that any given image is a re-education camp, versus a prison or something else? 90 percent sure? 60 percent sure?

For most of the camps I write about on my blog I have some corroborating evidence, like tender notices, or news reports suggesting there is a camp there, or some news pictures showing the camp. So most of the camps I post on my blog I’m 90 percent sure about. I’ve also discovered some camps for which I don’t have any corroborating evidence, but I’m pretty sure they’re camps.

Satellite images of before and after construction of what Shawn Zhang believes is a re-education camp in Kona Sheher (Shufu) County, Xinjiang. (Images courtesy of Google Earth. Interactive by Muyi Xiao)

Satellite images of before and after construction of what Shawn Zhang believes is a prison in Changji, Xinjiang. (Images courtesy of Google Earth. Interactive by Muyi Xiao)

What do you think other people can be doing to also help further this type of work and bring attention to, or to further research, re-education camps?

Some people are sending me the locations of the re-education camps. Many people in Xinjiang are sending the locations that they know. That’s really helped me to find the new camps. I think most of the camps don’t have tender notices [associated with them that are easily searchable on the Internet], so it really depends on local people’s knowledge to find them.

Each city usually has several camps. In one city in southern Xinjiang [Makit (Maigeti) County], government documents say they built one new re-education camp and turned five old buildings into re-education camps. But I can only find the newly-built one and not any of the old transformed buildings. They’re really difficult to find.

Do you have any thoughts more generally on where you think this whole situation is moving in Xinjiang? I mean, this is so much more radical and further than I thought the Xinjiang government would ever go. Something I think about a lot is that once you have facilities, and once you have equipment, they beg to be used. So say even if a year from now the Xinjiang government were to say, “Everyone’s been appropriately ‘re-educated’ and everything’s fine,” it’s hard for me to imagine that all these buildings would just empty out and they would never be used again. I worry that this is kind of the starting edge of something much worse.

As far as I understand, the Xinjiang government wants to have this kind of re-education campaign for at least several years. Even this year, they are still building more camps. Like the satellite image used in the Congressional hearing—early this year, they actually expanded that camp and tripled its size. So I think the Xinjiang government is still building. The construction has not slowed down. They are still building more camps. I think they want to keep this re-education campaign as a regular routine.

I think it’s really difficult to imagine what will happen in the next few years. I think more and more people will be getting sent to the re-education camps, but some reports have suggested that 10 percent of the Uighur population has already been sent into the re-education camps. That’s a lot. You have to exclude very old people and children. That basically means it’s a great loss for the labor force in Xinjiang. The Xinjiang government, I think, has to invest a lot of money into this re-education campaign because they have to pay [construction] expenses and pay the people living in the camps, like the security staff.

Maybe there will be some corporations with factories that use the Uighur people detained in these camps to do some labor, to work in their corporate factories. Although there is no direct evidence suggesting re-education detainees are required to do similar jobs like in a laogai camp [劳改, “re-education through labor”], the laogai model, or forced labor, is possible. So it is only a guess, there is not so much solid evidence. But one camp in Aksu has factories, and it’s possible detainees have to work in the factories. [As described on Zhang’s blog, “[t]his camp is noteworthy because it has adjacent factories. The tender notice suggests a walnut processing factory . . . and textile factory . . . will be built. It suggests [the] government could potentially use camp inmates to work in these factories.”] So I think they will use these Uighur people for labor in the factories like they did with the prisoners in the laogai.

Satellite images of before and after construction of what Shawn Zhang believes is a re-education camp that has factories in Aksu, Xinjiang. (Images courtesy of Google Earth. Interactive by Muyi Xiao)

Where do you see your project leading?

I’m not sure, but I think it’s becoming more and more difficult to find the kind of tender notices I’ve relied on in the past. Now, I rely more and more on witness accounts of where the camps are. Also, Google is updating its satellite images in Xinjiang, so maybe after that happens I will look and see if I can find more camps.

What do you hope to achieve, ultimately?

I hope my work shows the scale of the detention camps in Xinjiang. It’s a lot of people. One report estimated a million Uighurs have been detained in Xinjiang. Many Chinese people don’t believe it, because it’s a million, and they think “that’s really a lot, it’s impossible to detain one million people there.” I hope my project can provide some evidence that it is possible to detain that many people there, because China is building a lot of detention centers, like re-education camps, in Xinjiang. I hope my project can show the massive scale of the detentions.

Below is a screenshot of a tender notice that Zhang used to find the location on Google images of what he believes to be a camp. The screenshot is followed by our translation.

This particular notice makes very clear that the construction being solicited is for a “legal transformation through education school” (法制教育转化学校). Other notices use phrases such as “vocational education training center” (职业技能教育培训中心), but Zhang notes that when he matches the addresses listed in these notices with buildings he finds on Google Earth, the buildings often include features such as watchtowers and razor wire fences that suggest uses other than “vocational education” and that resemble the buildings that the government does label “transformation through education schools.”

Solicitation of Bids for Construction Design
April 26, 2017
Xinjiang Economic News

Shufu County Legal Education Through Transformation School Construction Project

Scope: New construction, 26,855 square meters, legal education through transformation school; construction to enlarge existing structure, 35,000 square meters, meant to configure relevant facilities and supplement existing facilities to form a complete set.

Address: Yishenlaimuqi Road (behind the new detention center), industrial zone public housing district, security company, Bulakesu Village old police substation, old county detention center, New District Community, Party School Training Center.
Investment approximately 140 million renminbi.

To design a section of this construction project, one must possess a construction project design certification grade B or higher; the project manager must have level two or higher registered architect professional certification.
All those wishing to participate in the bidding process should refer to the registration information provided on http://www.xjsfzb.com.

Work unit responsible for construction: Shufu County Bureau of Justice

Party responsible for receiving bids: Xinjiang Xinjianlian Project Management Information Ltd., Kashgar No. 1 Branch
Contact person: Zhao Feibao
Contact telephone: 18799552304
Supervising unit: Shufu County Construction Project Bid Solicitation and Receipt Management Office
Supervisor telephone: 0998-3252959