This week the U.S. Department of State publicly accused Syria’s government of killing prison inmates and systematically destroying their bodies. Citing newly declassified photographs, Washington alleges Syrian forces built a crematorium on prison grounds—just 45 minutes from the capital, Damascus—to eliminate evidence of their human rights abuses. Acting Assistant Secretary of State for Near Eastern Affairs Stuart Jones told reporters on Monday satellite photos and other evidence suggest, in recent years, a building was modified at Syria’s Saydnaya military prison complex to dispose of bodies. The photographic evidence purportedly shows modifications to the building and snowmelt on its roof—when all other nearby buildings were snow-covered—underscoring its alleged purpose. “At this point, we are talking about this evidence and bringing it forward to the international community, which we hope will put pressure on the regime to change its behavior,” Jones told reporters. To discuss how satellite photos can be verified and used to help confirm abuses, Scientific American spoke with Eric Stover, a human rights expert who has led forensic investigations into war crimes in more than a dozen countries over the past 32 years. Stover serves as faculty director of the Human Rights Center at the University of California, Berkeley. [An edited transcript of the interview follows.] How can satellite photos like these be used to make a determination about what we’re looking at here? Usually, satellite imagery and remote sensing are valuable when there are situations where you can’t get in to investigate on the ground. What you are looking for with remote sensing are significant changes. So, for example, when I was taking teams into Croatia and Bosnia and looking for mass graves, we would look at something like a satellite images of a farm. You could see the difference between a photo of a fallow field and when the earth was moved three days later—and you knew some activity had taken place. The way to look at this in a proper forensic investigation is you try to triangulate evidence and collect testimonial evidence and collect documentation of that. That could be anything from radio intercepts, e-mails or reports from meetings. And you would want to try to collect physical evidence if bodies are exhumed and identified. So how does that relate to this specific case? In this situation the American intelligence may have picked up on certain movements. We don’t know if they also have intelligence reports of people that worked at that site or survived and saw it or someone who worked at the site and defected and gave intelligence. We don’t know the behind the scenes triangulation here. It seems to me that to release that evidence to the public, they are fairly confident of what they are seeing. In investigations I have worked on, often what would happen is we would get satellite imagery and we would find during war a lot of dirt was moved, for example. Sometimes there were false leads but sometimes that was significant. We’re not talking about dirt here, however, because you are looking at changes to buildings, not lands. But you’d still try to triangulate your evidence and make as sound a case as you can. The building changes the State Department has talked about between 2013 and 2017 are things like snowmelt on the roof and changes including new air-discharge stacks. What suggests that this might be a crematorium and not something else, like a factory? That’s why we would call on the government to get further evidence backing up the claim. It would be good to know how this evidence has been verified and if there have been intercepted documents as well. We will need good scientists to be careful about these claims. How much historical precedent is there for doing this type of investigation in real time—while such atrocities may be occurring? That’s what happened during the war in Bosnia in the early 1990s: The Serb forces moved in and took over large portions of Bosnia. The general of the Bosnia–Serb Army ordered his troops to enter into a silver-mining town not far from the Serbian border, and the women and children fled to a factory, where the Dutch troops were, to take refuge. As the men tried to flee to a Muslim-held area, the Serbian troops stopped them, executed them and buried them in a mass grave. It was within a few days that satellite images became available, and also some journalists were in the region who were reporting by word of mouth what was happening. There was enough there to say something was happening—with potential crimes against humanity. When there are only photos of an event and no testimony to back it up, where’s the line where an investigator can say this is enough evidence? At our Human Rights Center we are running open-source investigations now, validating photographs and video clips—many of which are from social media—which make claims that there was an attack by Syrian forces. We train our students to carefully study these and look for the details in the frame—like if there is a mosque in the background—and look for earlier GPS photos and see that same location for comparison. You can look at shadows to indicate times of day and other factors, too. This work is so important. We verify these images for international courts or law firms or Amnesty International. They send these photos to us and we validate them based on the training we’ve had, to say either, “Yes, I can say this took place on December 2, 2014, and it does appear to be a specific village outside Aleppo” or, “No, we cannot reasonably say this is the case.” We believe today, more than ever, that facts are important and fact checking is even more important. Do today’s photo-altering capabilities make such verification work even more difficult? Potentially, yeah. Though it would take an awful lot of work to fabricate things, because sometimes with photos you can see a bridge where the siding has crumbled, for example. It’s hard to catch every little element there. If this is truly a crematorium where bodies are being disposed of, and if you or other investigators eventually get access to investigate on the ground, what sort of evidence would still be available to study? In court cases that looked at Auschwitz and other camps, examinations were based on things like seeing smokestacks that remained even after everything else collapsed. You can go in now with DNA analysis and take soil samples and see the residue of human remains, too. It’s very difficult to destroy that sort of site in a way that there would not be any way on the ground to say there were human remains. A body leaves a lot of traces behind. What are the next steps that should be taken to verify these claims? I think immediately if there is some way to monitor e-mail in particular—because superiors are always communicating with subordinates—that would be good. And look at all the ways they are communicating on social media, too. If this site is large and is being used as a crematorium, there always tend to be people involved who will be having second thoughts about what they are doing. So getting someone to corroborate that this is happening or getting someone who survived, or perhaps local police, to speak out or share photos would be important. We need to step back and validate this. If it’s going to a court, you need to triangulate as much as you can. No prosecutor will get up in court and present a weak case. For something like this to be taken to an international criminal court, you are going to need to build as strong a case as you can—and that means finding inside witnesses and getting as much documented evidence as you can [by] exploring the site.
Acting Assistant Secretary of State for Near Eastern Affairs Stuart Jones told reporters on Monday satellite photos and other evidence suggest, in recent years, a building was modified at Syria’s Saydnaya military prison complex to dispose of bodies. The photographic evidence purportedly shows modifications to the building and snowmelt on its roof—when all other nearby buildings were snow-covered—underscoring its alleged purpose. “At this point, we are talking about this evidence and bringing it forward to the international community, which we hope will put pressure on the regime to change its behavior,” Jones told reporters.
To discuss how satellite photos can be verified and used to help confirm abuses, Scientific American spoke with Eric Stover, a human rights expert who has led forensic investigations into war crimes in more than a dozen countries over the past 32 years. Stover serves as faculty director of the Human Rights Center at the University of California, Berkeley.
[An edited transcript of the interview follows.]
How can satellite photos like these be used to make a determination about what we’re looking at here? Usually, satellite imagery and remote sensing are valuable when there are situations where you can’t get in to investigate on the ground. What you are looking for with remote sensing are significant changes. So, for example, when I was taking teams into Croatia and Bosnia and looking for mass graves, we would look at something like a satellite images of a farm. You could see the difference between a photo of a fallow field and when the earth was moved three days later—and you knew some activity had taken place. The way to look at this in a proper forensic investigation is you try to triangulate evidence and collect testimonial evidence and collect documentation of that. That could be anything from radio intercepts, e-mails or reports from meetings. And you would want to try to collect physical evidence if bodies are exhumed and identified.
So how does that relate to this specific case? In this situation the American intelligence may have picked up on certain movements. We don’t know if they also have intelligence reports of people that worked at that site or survived and saw it or someone who worked at the site and defected and gave intelligence. We don’t know the behind the scenes triangulation here. It seems to me that to release that evidence to the public, they are fairly confident of what they are seeing. In investigations I have worked on, often what would happen is we would get satellite imagery and we would find during war a lot of dirt was moved, for example. Sometimes there were false leads but sometimes that was significant. We’re not talking about dirt here, however, because you are looking at changes to buildings, not lands. But you’d still try to triangulate your evidence and make as sound a case as you can.
The building changes the State Department has talked about between 2013 and 2017 are things like snowmelt on the roof and changes including new air-discharge stacks. What suggests that this might be a crematorium and not something else, like a factory? That’s why we would call on the government to get further evidence backing up the claim. It would be good to know how this evidence has been verified and if there have been intercepted documents as well. We will need good scientists to be careful about these claims.
How much historical precedent is there for doing this type of investigation in real time—while such atrocities may be occurring? That’s what happened during the war in Bosnia in the early 1990s: The Serb forces moved in and took over large portions of Bosnia. The general of the Bosnia–Serb Army ordered his troops to enter into a silver-mining town not far from the Serbian border, and the women and children fled to a factory, where the Dutch troops were, to take refuge. As the men tried to flee to a Muslim-held area, the Serbian troops stopped them, executed them and buried them in a mass grave.
It was within a few days that satellite images became available, and also some journalists were in the region who were reporting by word of mouth what was happening. There was enough there to say something was happening—with potential crimes against humanity.
When there are only photos of an event and no testimony to back it up, where’s the line where an investigator can say this is enough evidence? At our Human Rights Center we are running open-source investigations now, validating photographs and video clips—many of which are from social media—which make claims that there was an attack by Syrian forces. We train our students to carefully study these and look for the details in the frame—like if there is a mosque in the background—and look for earlier GPS photos and see that same location for comparison. You can look at shadows to indicate times of day and other factors, too. This work is so important. We verify these images for international courts or law firms or Amnesty International. They send these photos to us and we validate them based on the training we’ve had, to say either, “Yes, I can say this took place on December 2, 2014, and it does appear to be a specific village outside Aleppo” or, “No, we cannot reasonably say this is the case.” We believe today, more than ever, that facts are important and fact checking is even more important.
Do today’s photo-altering capabilities make such verification work even more difficult? Potentially, yeah. Though it would take an awful lot of work to fabricate things, because sometimes with photos you can see a bridge where the siding has crumbled, for example. It’s hard to catch every little element there.
If this is truly a crematorium where bodies are being disposed of, and if you or other investigators eventually get access to investigate on the ground, what sort of evidence would still be available to study? In court cases that looked at Auschwitz and other camps, examinations were based on things like seeing smokestacks that remained even after everything else collapsed. You can go in now with DNA analysis and take soil samples and see the residue of human remains, too. It’s very difficult to destroy that sort of site in a way that there would not be any way on the ground to say there were human remains. A body leaves a lot of traces behind.
What are the next steps that should be taken to verify these claims? I think immediately if there is some way to monitor e-mail in particular—because superiors are always communicating with subordinates—that would be good. And look at all the ways they are communicating on social media, too. If this site is large and is being used as a crematorium, there always tend to be people involved who will be having second thoughts about what they are doing. So getting someone to corroborate that this is happening or getting someone who survived, or perhaps local police, to speak out or share photos would be important. We need to step back and validate this.
If it’s going to a court, you need to triangulate as much as you can. No prosecutor will get up in court and present a weak case. For something like this to be taken to an international criminal court, you are going to need to build as strong a case as you can—and that means finding inside witnesses and getting as much documented evidence as you can [by] exploring the site.
