SAR analysis, data mining & GIS mapping of the Central Mediterranean to document the violations of migrants' rights & deaths at sea

Summary

Forensic Oceanography (FO) is a project led by Charles Heller, Lorenzo Pezzani and Situ Studio as part of the European Research Council project “Forensic Architecture” directed by Centre for Research Architecture, Goldsmiths, University of London (www.forensic-architecture.org). It aims to use geospatial technology to document the violations of migrants’ rights and deaths at sea. GISCorps volunteers contributed this summer to the analysis of Synthetic Aperture Radar (SAR) imagery towards a legal case aiming to determine responsibility for the death of 63 migrants of the Libyan coast in 2011. GISCorps volunteers also contributing to data mining and GIS mapping towards the pilot phase of WatchTheMed (watchthemed.crowdmap.com), a participatory mapping platform aiming to monitor the Mediterranean so as to document violations and eventually prevent them from occurring.

Deaths at the Borders of Europe. Olivier Clochard and Philippe Rekacewicz, Le Monde diplomatique, June 2010

The death of migrants at sea: a longstanding phenomenon

The death of migrants at the EU’s maritime borders is a sadly longstanding phenomenon. It is at the end the 1980s, when increasing restrictions on the entry of non-European migrants into EU territory entered into force, that the first bodies of migrants washed ashore on Mediterranean beaches. Faced with a legal wall and increasingly militarised means of control, migrants resorted to clandestine and dangerous means of entry – amongst others embarking on unseaworthy vessels. Further more, the refusal of coastal states to disembark migrants and the criminalisation of seafarers who have assisted migrants have become a strong disincentive for seafarers to comply with their obligation to rescue migrants in distress. As a result, between 1988 and May 2012, Fortress Europe has counted 13.448 deaths of migrants at the maritime borders of the EU, and 6.255 in the Sicily Channel only.

Until 2011, a few organisations were involved in the fundamental work of “counting the dead”, producing analysis of their causes and denouncing them. The map above was instrumental in making the wider public aware of the scale of these deaths and the strong degree of responsibility of EU migration policies. However the number of deaths continued to mount and reached its climax in 2011 following the Arab Spring, when important movements of population occurred in particularly precarious conditions.

2011: the “deadliest year in the Mediterranean"

The UNHCR defined 2011 as the “deadliest year in the Mediterranean” since the organisation began recording these statistics in 2006, estimating that over 1,500 migrants died within the year while fleeing Libya during the initial stages of the violent conflict.^{^[1]}Forteress Europe further counted 334 deaths amongst migrants from Tunisia. However this record number of deaths occurred while western states’ military ships and patrol aircrafts were deployed of the Libyan coast towards the international military intervention under the United Nations Security Council Resolution 1973. The largest number of death at sea thus occurred within the most surveilled waters on earth. As such, the deep and ongoing political responsibility of the EU was doubled by another level of direct responsibility on the part of the military forces of the worlds’ most powerful states who could not not witness the distress of migrants and yet were failing to assist them

This paroxysmal context led organisations defending migrants’ rights to resort to new strategies to which the Forensic Oceanography project based at Goldsmiths College, University of London contributed to and which GISCorps volunteers supported in fundamental ways. The first outcome is a contentious legal case attempting to draw direct responsibility for the death of 63 migrants in the “Left-to-die boat” case, the second is an international campaign called Boats4People aiming to draw attention to the death of migrants at sea and attempt to prevent them.

Chain of events of the Left-to-die boat case

A) Departure point at Port of Tripoli between 00:00 and 02:00 UTC on 27 March. Boat first spotted by a French aircraft at 14:55 GMT on 27 March at position LAT 33°40’ N, - LON 13°05' E

(B) GPS location of vessel at 16:52 GMT on 27 March 2011 at position LAT 33 58.2 N – LON 12 55.8 E as determined by the MRCC based on locations established by the satellite phone provider Thuraya.

(C) The GPS position of the boat was determined a second time (at 19:08 GMT on 27 March at position LAT 34 07.11 N – LON 12 53.24 E, again based on information provided by Thuraya.

(D) The vessel began to drift within a 4.3 nm radius of position 34 24.792 N – 12 48.576 E at approximately 07:00 GMT on 28 March.

E) Between 3 and 5 April the migrants encounter a military ship. On 10 April the boat lands

1^st initiative: The “Left-to-die boat” case

Amongst all the incidents that occurred in 2011, one particular event, reported by the international press, provoked widespread public outrage. In the case of what is now referred to as the “Left-to-die boat”, 72 migrants fleeing Tripoli by boat on the early morning of March 27 2011 ran out of fuel and were left to drift for 14 days until they landed back on the Libyan coast. With no water or food on-board, only nine of the migrants survived. In several interviews, these survivors recounted the various points of contacts they had with the external world during this ordeal. This included describing the aircraft that flew over them, the distress call they sent out via satellite telephone and their visual sightings of a military helicopter which provided a few packets of biscuits and bottles of water and a military ship which failed to provide any assistance whatsoever. The events, as recounted by these survivors, appeared to constitute a severe violation of the legal obligation to provide assistance to any person in distress at sea, an obligation sanctioned by several international conventions.

In response to this incident, a coalition of NGOs was formed (constituted primarily by GISTI (www.gisti.org), FIDH (www.fidh.org) and Migreurop (www.migreurop.org), Human Rights Watch (www.hrw.org) to seek accountability for the non-assistance of migrants at sea during and in the aftermath of Arab Spring in general and in the case of the “left-to-die boat” in particular. The Forensic Oceanography project offered its expertise in spatial analysis to this initiative and published a report on 11 April 2011, which was the basis for a first case filed against the French military and will be used in forthcoming procedures in the courts of other countries.

More information on this initiative can be found here: www.fidh.org/Death-of-63-migrants-in-the www.forensic-architecture.org/publications/report-on-the- left-to-die-boat/

The report FO produced involved an innovative methodology, particularly considering the field of migrants’ rights. By combining survivors’ testimonies, GPS coordinates published by the Italian Coast and a drift model of the boat’s trajectory, we were able to reconstitute with precision the position of the migrants’ vessel over the duration of this event. The distress signals sent out by the Italian Coast guards on the 27 and 28 March 2011 also allowed us to determine in which area ships were informed of the migrants’ distress. The question that remained was thus: “which ships were present at the time?”. It is here that Synthetic Aperture Radar (SAR) imagery and analysis became central. SAR imagery was crucial in providing a snapshot of the maritime context at the time and in the area of events.

For the analysis of the acquired SAR imagery, FO reached out to GISCorps. GISCoprs recruited Lawrence Fox III, Humboldt State University Emeritus Professor of Remote Sensing and Consultant to provide SAR imagery analysis. The report on this first contribution of GISCorps can be accessed here: http://giscorps.org/index.php?option=com_content&task=view&id=117&Itemid=63

The support of the GISCorps volunteer had been fundamental to the first phase of our project and as such it was only natural to reach out to GISCorps volunteers for its next steps.

SAR analysis by Rossana Padeletti, indicating vessel returns in vicinity of migrant boat

New SAR data for Left-to-die boat case

In summer 2012 we were able to acquire a new SAR imagery relevant to the Left-to-die boat case. While we had previously relied on Envisat data of 75m resolution, we were able to acquire Radarsat-1 data of 50m resolution and concerning a key moment in the chain of events: the 4^th of April 2011. This was the day when, according to our reconstruction based on several sources of evidence, the migrants’ boat at drift since already 8 days encountered a still unidentified military vessel. At this point in the chain of events, almost half of the 72 people had already died. Despite witnessing the distress of the migrants – the vessel came up to 10m and members of the crew onboard took photographs of the migrants - the military vessel provided no assistance whatsoever, effectively letting the migrants die in all knowledge of their tragic fate.

The new SAR imagery we have acquired and its refined analysis provided by Rossana Padeletti, GIS and Remote Sensing Specialist and GISCorps volunteer, still does not allow us to identify the military vessel in question. The image is first of all taken between 5:08:42 and 5:10:08 UTC time, whereas the encounter with the vessel occurred towards the end of the day according to the survivors. Further more, the relatively low resolution allows us only to detect returns for vessels in the area at the time and does not enable us to define the identity of these ships. However it does allow us to say with certainty that only a few hours before this encounter there were at least 78 vessels of over 50m in the waters of the coast of Libya. A number of these were certainly military vessels, since according to a document released on 24 March by the U.S. Department of Defense were 38 military ships deployed of the coast of Libya and all of them were above 50 meters in length. Four of these were less than 50nm away and could have easily reached the migrants’ boat. The SAR analysis will be released shortly as new evidence towards the Left-to-die boat case.

2^nd initiative: Boats4People and WatchTheMed

A second important initiative that emerged amongst migrant’s rights networks in response to the record number of deaths at sea was the Boats4Peope campaign (www.boats4people.org) that took place in July 2012, linking both shores of the Mediterranean through the travel of a solidarity boat accompanied by many activists. At every stage of the boat’s travel, intense meetings occurred between migrants, activists, journalists and politicians. B4P gathered in particular with Tunisian families of dead and disappeared migrants who demand information and justice, as well as with the refugees and rejected of the Choucha camp in Tunisia who demand protection, resettlement and better living conditions. After this first campaign, Boats4People is currently consolidating into a permanent structure allowing to monitor and eventually prevent the deaths of migrants at sea.

watchthemed.crowdmap.com, with layers produced by Steve Etherington and Keith McCrary (red: Search and Rescue zones; light green: territorial waters; blue ovals: Frontex operations; orange: Italian coastal radars

During the B4P campaign, the Forensic Oceanography project also launched the pilot phase of an almost real time and participatory mapping platform called WatchTheMed (watchthemed.crowdmap.com). The aim here was to make the mapping tools used towards the report described above available to the wider movement, so that they could be used towards more cases and increase accountability at sea. The main elements we already had – new tool for documentation of violations occurring in a very large space and their spatialisation so as to determine responsibility – were coupled with the tools of “crowdsourcing” used in crisis mapping so as to come closer to a documentation in real time and in a participatory mode. With this our hope was not only to document violations occurring in the past but to tend towards the real time so as to actually prevent violations and death from occurring.

The main aims of the pilot phase were to:

set up the platform
create the main layers for the map and research data that was still missing
try out the live monitoring of the maritime space and crowdsourcing

The work of three GISCorps volunteers was fundamental in all stages of this project.

Platform

The contribution of GISCorps in fact began before the official start of the volunteers’ missions, since Shoreh Elhami, co-founder of GISCorps, already gave us precious advice on the choice of the platform. We opted for Crowdmap as an easy to set up and use tool ideal for this pilot phase.

Layers

Through the static layers of the map, the aim was to provide the most complete picture of maritime life in relation to migration: different actors and technologies aiming to control migrants, but also jurisdictions as well as other actors who populate the sea and encounter migrants.

While SAR zones and territorial waters layers had been produced in advance, Steve Etherington and Keith McCrary, two GIS specialists and volunteers from GISCorps, contributed to search for the data and draw the following map layers:

oil and gas offshore development
patterns of route traffic both cargo and passenger ferry services
GSM/3GSM mobile coverage at sea
Italian Coast Guard radars coverage
Frontex (the EU border police) operation locations

Another layer that we attempted to indicate was live AIS tracking data for large vessels. While the process was not straightforward and the data could not be included in this phase, Steve Etherington and Keith McCrary managed to define a procedure that could be used in the future.

Monitoring incidents

In terms of gathering live information on incidents at sea, we were able to collect Hydrolant distress signals accessible online, and some members of B4P organisations also received distress calls from migrants at sea. It is important to note that we intentionally limited crowdsourcing since we did not want to set up connections we could not sustain in the long run – that is before B4P-WTM became further institutionalised. In the future, we will rely further on migrants’ networks as well as reach out to sailors.

We were able to document incidents shortly after they occurred thanks to two B4P volunteers who scanned the press systematically. B4P volunteers were dispatched to collect testimonies of migrants who had survived the crossing. While Rossana Padeletti, Remote Sensing Specialist, searched for satellite imagery following reports of incidents, pertinent imagery was not found. This clearly demonstrated some of the limits of monitoring based on satellite imagery since very few images are actually taken and the area is extremely vast.

Outcome

During B4P’s solidarity boat’s 3 weeks at sea, WTM acted as a kind of a “civilian watchtower” over the Mediterranean. With this system we were able to document 400 arrivals and the death of over 100 people at sea. The information gathered through WTM was well publicised by the Italian and French press during the entire operation, while one particularly dramatic incident was publicised internationally. This was an incident in which Abbas, the sole survivor of 56 migrants, was rescued near the coast of Tunisia after having left from Tripoli and drifting at sea for 14 days. WTM reporters were able to meet him shortly after he was rescued and reported on the incident: https://watchthemed.crowdmap.com/reports/view/26, It was further publicised by newspapers such as the Guardian, Malta Today and Suddeutsche Zeitung. The international coverage helped Abbas achieve particular attention from the UNHCR and states to find a resettlement solution as quickly as possible. The B4P network and the WTM platform were however not able to prevent these tragic deaths from occurring.

Example of 06.09.2012 incident documented with WTM

One important incident occurred after the project’s pilot phase had ended, but the tools developed towards the platform were used in Google Earth to help document it effectively. In the afternoon of the 6 of September 2012, a boat carrying over 100 Tunisian migrants drowned 10nm off the coast of Lampedusa, despite having sent out several distress calls by mobile phone to the Italian Coast Guards. Only approximately 50 migrants were eventually rescued. This is an incident B4P is continuing to inquire into and many questions remain open, but the tools of WTM proved very useful to better understand the unfolding of events. We were able to find the coordinates of the incident provided in the distress signal sent out by the Italian Coast Guards. These coordinates, once inserted into the different layers we had designed, showed that the boat was located with the Italian territorial waters (blue), within the Italian Search and Rescue zone (red line), within the perimeter patrolled by the Italian border police (black line). The fact that the incident occurred within these zones, which we expect to be carefully monitored, may indicated that a failure occurred. Finally, the incident occurred within the zone of GSM coverage, which corroborates the fact that the migrants called for help with a mobile phone. The very simple map bellow was useful for B4P to demand further information and accountability to Italian and Tunisian authorities.

In summary, this was a good pilot phase in that a participatory mapping system was created and tested, allowing both to see the usefulness of the platform as well as its limits. The structure seems to be there – even though it should be further refined – but the real time monitoring and capacity to intervene in the unfolding of events must be further developed. The contribution of GISCorps volunteers to this phase was fundamental, as they produced important components of the project and proved able to work creatively with the element of unpredictability that accompanies any pilot project. We hope that GISCorps volunteers will be able to continue to support the WTM initiative in the future and help it develop into an effective warning and documentation mechanism for violations and situations of distress at sea.

Authors: Charles Heller and Lorenzo Pezzani, Centre for Research Architecture, Goldsmiths, University of London, Situ Studio, New York & GISCorps Volunteers: Rossana Padeletti, Steve Etherington, & Keith McCrary

Reprinted with permission, GISCorps