Satellite imagery as evidence for international crimes


The acquittal of Mathieu Ngudjolo Chui at the International Criminal Court (ICC) demonstrates the challenges to collect strong and sufficient evidence to bring perpetrators of mass atrocity crimes to justice. As Alison Cole wrote in heropening blog post in the technology for truth series, with its lack of enforcement powers, the ICC is largely dependent on voluntary state cooperation to gather evidence. At the same time, as she rightly states, it is also an ideal venue to “involve evidence processed through technological means.“ Remote sensing, the use of earth orbiting and aerial platforms to collect data, can sometimes provide unique, otherwise unavailable evidence of alleged mass atrocity crimes. Because these events often occur in non-permissive environments, over large geographic areas, and across long and multiple time frames, high resolution satellite imagery, in particular, has value for international tribunals investigating these crimes.

Satellite imagery was first used in international criminal proceedings during theSrebrenica trials at the International Criminal Tribunal for the former Yugoslavia. Yet, the full value that this unique form of evidence may possess is still underused comparative to its potential. Practical matters of collecting and analyzing satellite based evidence, as well as the still nascent legal precedents related to admissibility and probative value, are major barriers to its more frequent and effective use in trials.

Satellites, in the simplest sense, capture light reflecting from the surface of the Earth. Once the light is captured and digitized, it is processed into an image. These types of images are those that can be found on Google Earth.  The images can be adjusted and processed to isolate, amplify, or suppress specific wavelengths of light that can provide the analyst with important information beyond what the human eye can perceive. Through imagery interpretation and analysis, visual information can be used to cross-corroborate complementary non-imagery information – witness testimony, reports and newspaper articles, or information gathered by crowdsourcing platforms. The result of this correlation of data streams is later published as a finding, which can be offered as evidence.

These analyses can often uniquely show the widespread, systematic character and long-term extent of reported violence. Some of the phenomena that can be detectedinclude cratering consistent with bombardment; the destruction of civilian dwellings; targeting of humanitarian and other forms of internationally protected infrastructure; the locations, size, and composition of displaced persons camps; excavations consistent with mass graves; and other probative phenomena.

With this capability, satellite evidence can provide critical crime-specific and circumstantial evidence. It should not be assumed, however, that a specific fact in issue can be proven on the basis of satellite imagery analyses alone. In many instances corroborative evidence is needed, which is often integrated into the analysis.

The main challenges for the use of satellite imagery as evidence before the ICC are as follows:

  1. The limited availability of appropriate imagery data;
  2. Restricted budgets and cognitive bias of external organizations conducting the analysis;
  3. No accepted forensic standards and methodologies, and
  4. A lack of familiarity of judges with this kind of evidence.

Limited Available Data
The availability of appropriate imagery of an area during the critical time period a crime was allegedly committed may often be limited. This challenge can result from meteorological phenomena obstructing what can be seen in the image, government restrictions on collecting imagery of a certain area, or the high financial cost of obtaining imagery. Almost all potentially relevant high-resolution imagery is collected by commercial providers that only collect data when it is of current or potential future value. Whether imagery of a specific situation is pertinent for a criminal investigation can often only be known months, sometimes years later. Hence commercial providers have a low economic incentive to routinely collect areas where mass atrocities are allegedly occurring. Relatedly, third party investigators have low economic capacity to task or purchase such imagery or to create economic incentives for its collection. Once the critical moment to acquire timely imagery is passed, there is an almost insurmountable barrier to data access and much potential evidence is already lost and cannot be recaptured.

Reliance on External Capacity
The ICC has limited in-house capacity and expertise to conduct the satellite imagery analysis that may be required by the many cases on which it works simultaneously. Thus, the majority of this work is often performed ad hoc by third party analysts who are not affiliated with a specific court or prosecutor’s office. Some of the groups that perform this type of work include Human Rights Watch, UNOSAT, Amnesty International, the Signal Program at the Harvard Humanitarian Initiative, andAAAS. Often operating under restricted budgets for conducting these investigations, the economic dynamics inherent in the sector often constrain what potential contributions they can make. These institutions also often have their own focuses and agenda that rarely matches the evidentiary needs of courts.

No Accepted Forensic Standard and Methodologies
Despite the growing number of satellite-analysis of human rights violations, there is little formal pedagogy specific to this field and hence a lack of professionalization and standardization of methods. Analysts currently operate without an accepted forensic standard for remote sensing of alleged mass atrocities.

Little Familiarity of Judges with Geospatial Evidence
Satellite imagery analyses are results of a complex technical and methodological processes based on the division of labor. Only when judges understand the underlying techniques and methods are they able to determine admissibility and weight of this evidence. Currently, there is a lack of familiarity of judges with the analytical process. It is important that the introduction of the evidence into is comprehensible for the judges. The current practice of presenting images and findings in pdf-printouts or on the in-court computer screens does not fully satisfy this need.

In order to overcome these challenges, the following steps should be taken:

  • Because of the digital and technological nature of satellite based evidence, it is particularly important to establish its reliability for it to be admissible (the legal aspects that affect the use of satellite based evidence are discussed in more detail here.). For this purpose, the accuracy, objectivity, and authenticity of the data must be proven along the whole process.
  • Since satellite based analyses are processed by a number of systems and individuals before submitted to court, it is important to preserve the chain of custody of the data along the process. This requires the development of standardized methods for data acquisition, storage, interpretation, integration of cross-reference data, and visualization for the specific purposes of international criminal investigations.
  • At the same time, the practical circumstances of the collection and analyses of satellite images must be addressed if any advancement in its use as evidence is to be achieved.
  • Further, it would be helpful to familiarize lawyers and especially judges with this process so that they can more easily evaluate its reliability and probative value.

If these challenges can be overcome and the evidentiary potential of satellite based evidence is fully exploited, prosecutions at the ICC could rely on a unique form of evidence, which would enhance the quality of international prosecutions and contribute to the fight against impunity of mass perpetrators.

Dr. Patrick Kroker is an international lawyer based in Berlin, Germany and 2014 Fellow at the Harvard Humanitarian Initiative.

This post originally appeared on the International Justice Monitor.