Report of the Secretary-General's Panel of Experts on Accountability in Sri Lanka/III Nature and Scope of Alleged Violations

III. Nature and Scope of Alleged Violations
48. The Panel now turns to the allegations of violations of international humanitarian and human rights law committed during the final stages of the war (September 2008 – May 2009) and its immediate aftermath, pursuant to its mandate to provide advice to the Secretary- General “having regard to the nature and scope of alleged violations”. It is the nature and scope of the allegations that determine the nature of the advice that the Panel provides on accountability. In chapter IV, the Panel appraises those allegations from a legal perspective to determine whether the alleged acts, if proven, would amount to violations of international law.

A. Methodology for evaluating allegations
49. The Panel’s assessment is based on a careful examination and weighing of the allegations of fact that have been made regarding the final stages of the war. The Panel’s examination included both written sources of information as well as interviews with various individuals. The written sources included reports, documents and other written accounts by the various agencies, departments, funds, offices and programmes of the United Nations, other inter-governmental organizations, NGOs and individuals, such as journalists and experts on Sri Lanka. It included satellite imagery, photographs and video materials of the final phases of the war. It also included submissions received by the Panel during the course of its work in response to its notifications posted on the United Nations website. While these could not be individually verified, at times they served to corroborate other sources. Some relevant media sources, referring, for example, to statements of the Government of Sri Lanka or other public statements, are cited in this chapter, but serve only to corroborate the information gathered by the Panel. A number of NGO reports exist on events in the Vanni. While the Panel reviewed some of these reports, it did not rely on them to compile these allegations, but rather carried out its own assessment of the nature and scope of allegations.

50. The Panel consulted a number of individuals with expertise or experience related to the armed conflict, including officials of international organizations, NGOs, journalists, diplomats, academics and other individuals, some of whom were in Sri Lanka or in the Vanni during the relevant period.

51. While the Panel’s mandate precludes fact-finding or investigation, the Panel believed it essential to assess whether the allegations that are in the public domain are sufficiently credible to warrant further investigations. Determining the scope and nature of these allegations allows the Panel to properly frame the accountability issues, which arise from them. The Panel has determined an allegation to be credible if there is a reasonable basis to believe that the underlying act or event occurred. This standard used by the Panel – that of a reasonable basis to believe that the underlying act or event occurred – gives rise to a responsibility under domestic and international law for the State or other actors to respond.

52. To determine whether an allegation is credible, the Panel considered the totality of the information in its possession, with careful regard to the relevance, weight and reliability of each of the sources as well as its relationship to the body of information, as a whole. Allegations are only included as credible when based on primary sources that the Panel deemed relevant and trustworthy. These primary sources were corroborated by other kinds of information, both direct and indirect. The allegations laid out below are based on credible and consistent sources of information. In fact, many of the allegations would appear to meet a higher standard of proof.

53. The Panel has chosen to present the allegations it finds credible in a narrative account rather than listing the various allegations under their legal classification, so as to provide a greater sense of context and perspective. This account should not be taken as proven facts, and any effort to determine specific liabilities would require a higher threshold.

B. Background on military strategies and operations
54. In order to have the necessary context for understanding the credible allegations, which are laid out in sections C to F of this chapter, the Panel examined the political and military strategies and capacities of both the Government of Sri Lanka and the LTTE during the final phases of the war. For its analysis of the general strategy, the Panel relied on a range of sources such as publicly available political and military analyses, the Sri Lankan Defence Ministry website and international military experts from the United Nations Office of Military Affairs. Where the Panel examined specific allegations, it did so based on primary sources.

1. Government of Sri Lanka
55. In the aftermath of the failed peace process in 2006, the Government of Sri Lanka devised a comprehensive multi-pronged strategy to defeat the LTTE. This strategy included diplomatic and political components, measures to control information about and access to the combat zones, as well as more strictly military components.

56. In its diplomatic efforts, the Government drew on the favourable global environment for support from a number of States, in the context of the “war on terror” and, in the region, gained increased collaboration from the Indian Government due to the LTTE’s assassination of Rajiv Gandhi. In one significant example, the Indian Navy directly assisted Sri Lankan forces in intercepting the floating warehouses used by the LTTE to maintain its supplies by sea.

57. Internally the Government ensured cohesive political and military leadership. The President appointed himself to be Minister of Defence and his brother, Gotabaya Rajapaksa, as Secretary of Defence. He then appointed Lieutenant General Sarath Fonseka as Army Commander. Both the Secretary of Defence and Lieutenant General Fonseka had extensive combat experience against the LTTE and each survived assassination attempts by the LTTE in 2006. The President then obtained parliamentary approval for major increases in the military budget, which grew to USD1.8 billion in 2008, representing almost 20 per cent of the national budget.

58. With regard to the military, Lieutenant General Fonseka greatly bolstered and revitalized the Sri Lanka Army (SLA). The size of the Armed Forces was almost tripled to 300,000, and regular rotation ensured a steady supply of fresh troops to the battlefront. The army procured new equipment and weapons, strengthening its arsenal of Multi-Barrel Rocket Launchers (MBRLs), mortars and howitzers, MIG-29, Kfirs and helicopter gunships. The Sri Lanka Air Force (SLAF) acquired and used several models of unmanned aerial vehicles (UAVs) for surveillance, target acquisition and subsequent battle damage assessments.

59. The SLA generally relied on its vastly superior firepower and troop numbers as well as its air supremacy to maintain a relentless offensive, on many simultaneous fronts, from which the LTTE could not recover. All attempts to negotiate were declined, and no ceasefires were given, as the Government argued that the LTTE would use any ceasefires to regroup.

60. The SLA also employed flexible, intelligence-driven and guerrilla-like tactics to advance its infantry and inflict maximum damage on the enemy. To this end, the Army Commander reorganized and retrained his forces, including the formation of self-reliant Special Infantry Operations Teams (SIOTs), trained in commando techniques and able to coordinate surveillance, artillery and air strikes. The SLA also used clandestine operations behind enemy lines, carried out by Deep Penetration Units, also known as Long Range Reconnaissance Patrols, composed in part of Tamil militants formerly associated with LTTE commander Mahattaya, to gather human intelligence. In order to reduce military casualties as it advanced, some SLA operations were designed to “soften up the ground” and pierce LTTE defensive lines with heavy artillery. While the Government spoke of its efforts to liberate civilians from the area in a humanitarian operation and to incur zero civilian casualties in the process, in practice, ground forces appear to have been given significant discretion to use a barrage of artillery as they advanced.

61. Prior to shelling, UAVs were often used to identify potential targets. A live stream of footage from UAVs was directly delivered to ground operations, which enabled commanders to make decisions, using that information with virtually no delay. In some areas the physical conditions such as jungle and foliage would have hindered visibility, but generally the UAVs transmit high resolution images. The UAVs used by the SLAF have the capacity to identify single targets, such as individuals and their movements or positions, and to depict terrain features, thereby providing ground troops with validated, near real-time information. Through the use of UAVs, the SLAF had the ability to detect enemy formations both day and night, in various topographic areas. The use of UAVs also enabled the SLAF to identify individuals and civilian installations, such as hospitals.

62. Lieutenant General Fonseka himself commanded the war effort from the Joint Operations Headquarters in Colombo and handpicked seasoned commanders to lead the campaign. The Vanni Security Forces were headed by Major General Jagath Jayasuriya, headquartered in Vavuniya. In addition, six major battalions were active in the final stages of the war, including the 53rd Division (commanded by Major General Karmal Guneratne) and the 55th Division (commanded by Brigadier Prasanna Silva), which progressed south from the Jaffna peninsula; the 56th Division (commanded by Major General H.C.P. Gunalithaka), the 57th Division (commanded by Major General Jagath Dias), the 58th Division (commanded by Brigadier Shavendra Silva) and the 59th Division (commanded by Major General Nandana Udawatta), which all progressed from the south and south-west. Special Forces and commandoes also played a crucial role. Strong leadership, new training and a strong sense of purpose bolstered by numerous victories decreased desertion rates and improved morale among SLA troops.

63. In addition to its regular military operations, the Government employed clandestine operations to uncover LTTE safe houses, dismantle the LTTE networks in the South and eliminate persons believed to be associated with the LTTE. A potent symbol of these operations was the “white van”. White vans were used to abduct and often disappear critics of the Government or those suspected of links with the LTTE, and, more generally, to instil fear in the population. An elite unit within the Special Task Force (STF) of the police is implicated in running these white van operations. Those abducted were removed to secret locations, interrogated and tortured in a variety of ways, including through beatings, forced nudity, suffocation with plastic bags, partial drowning, extraction of finger or toe nails, or administering electric shocks. Many were killed and their bodies disposed of secretly. Human rights workers, journalists, newspaper editors and humanitarian workers accused of being “Tiger sympathizers” were also caught in the net. In the period between 2006 and the end of the war, 66 humanitarian workers were either disappeared or killed.

64. The strategy also involved stricter controls on the media and the flow of information, imposing a media blackout and stifling critical views of the war effort. From 2006, independent journalists were not allowed to travel to LTTE-controlled areas and certain journalists were named as “Tiger sympathizers” on the Ministry of Defence website. More detailed guidelines on reporting on the war were established in 2008. Journalists who disobeyed these rules or who were otherwise critical of the Government were subject to arrest and/or severe levels of threat. In June 2008, the Defence Secretary directly threatened two prominent journalists of the Associated Newspapers of Ceylon Ltd (Lake House). One of them, Poddala Jayantha, was subsequently picked up in a white van and severely beaten in June 2009. On 8 January 2009, a prominent newspaper editor, Lasantha Wickrematunge, was killed by unidentified assailants in Colombo. He had collected information alleging corruption in military procurements made by the Defence Secretary, which he was about to publicize as part of a lawsuit. Two days earlier, the independent Maharajah Television / Broadcasting Network was attacked by armed gunmen, who destroyed equipment while holding security guards at gunpoint. None of these events was investigated. At least 10 journalists were killed between 2006 and early 2009. The combined effect of these censorship measures and the fear of being killed or beaten had a deleterious effect on the independent oversight role played by the media. The Defence Secretary was quoted as saying: “I have only two groups - the people who fight terrorism and the terrorists”.

65. The international community was also subject to continuous public criticism by the Government for having been too tolerant of the LTTE and “supporting terrorists”. It was eventually excluded from the conflict zone altogether as part of the strategy, and international diplomatic efforts to put a halt to the fighting or to call for ceasefires were rebuffed or ignored.

2. The LTTE
66. By September 2008, the LTTE’s military capabilities were severely diminished compared to its past strength. Although its exact size at the time is not known, at its peak it was not larger than 20,000; its core fighters consisted of only a fraction of that in the final stages of the war, perhaps up to 5,000. In the south, its networks and sleeper cells in Colombo and elsewhere had been weakened and its ability to carry out suicide actions, although still existent and active, was reduced by the Government’s counterinsurgency operations. The LTTE bid to hold territory as the State of “Eelam” required a departure from guerrilla tactics and a switch to fighting a conventional war with frontlines and fixed-site battles. In the final stages of the war, its efforts were largely defensive, and its surprise counterattacks, which had success in the past, did not materialize. As a result, the LTTE was increasingly cornered into an ever-contracting territory and sought to defend itself behind barriers of earth bunds, fortifications, minefields and ditches, using ambush techniques, booby traps and improvised explosive devices.

67. Its status as a proscribed terrorist organization in some 32 countries limited its international operations and support. The LTTE was a fraction of the size of the SLA, and many of its cadre were inexperienced, but its basic command structure remained intact, with a military wing and, under it, a political wing. Both were headed by a central governing committee led by Velupillai Prabhakaran, overseeing the Sea Tigers (headed by Soosai), the Air Tigers (headed by Charles Anthony, Prabhakaran’s son and Rathnam Master), an elite fighting unit known as the Charles Anthony Regiment, the Black Tigers, an intelligence unit headed by Pottu Amman and a political office headed by Nadesan. Most of these leaders were killed in the final stages of the war.

68. The LTTE mainly relied on forced recruitment in an attempt to maintain its forces. While previously the LTTE took one child per family for its forces, as the war progressed, the policy intensified and was enforced with brutality, often recruiting several children from the same family, including boys and girls as young as 14. Civilians were also enlisted by the LTTE into their war effort in other ways, using them, for example, to dig trenches and build fortifications, often exposing them to additional harm.

69. Although the LTTE’s supply chains had been disrupted, especially after the loss of its floating warehouses, it still had access to some stockpiles of weapons, including some artillery and mortars and a few MBRLs. It used these to offer stiff resistance from behind its fortifications and earth bunds and also launched waves of suicide attacks. Its Black Tigers continued to operate and increasingly engaged in suicide missions against the attacking SLA as well as perpetrating some attacks outside the conflict zone, with numerous civilian casualties. The Sea Tigers still had some of their naval equipment and boats, but the LTTE’s remaining operational Zlin 143 aircraft were shot down during a suicide mission over Colombo in February 2009.

70. Retaining the civilian population in the area that it controlled was crucial to the LTTE strategy. The presence of civilians both lent legitimacy to the LTTE’s claim for a separate homeland and provided a buffer against the SLA offensive. To this end, the LTTE forcibly prevented those living in the Vanni from leaving. Even when civilian casualties rose significantly, the LTTE refused to let people leave, hoping that the worsening situation would provoke an international intervention and a halt to the fighting. It used new and badly trained recruits as well as civilians essentially as “cannon fodder” in an attempt to protect its leadership until the final moments.

3. Civilians in the conflict area
71. In September 2008, the SLA was advancing into the Vanni from multiple directions, trapping increasingly large numbers of civilians who were not able to leave the area. The Government regularly dropped leaflets urging civilians to leave the area (but providing no specific information on how to do so); however, most remained. Apart from the LTTE pass system, numerous other factors contributed to civilians becoming trapped in LTTE-controlled areas. For most of these people, the Vanni was their home. Many had experienced the military occupation of Jaffna and had moved with the LTTE since 1995. From experience, they feared what would happen to them if they crossed into Government-controlled areas, knowing that they would be subject to internment. They also feared the white vans and feared being raped or tortured by the army. In addition, the LTTE forced recruitment practices meant that many families had relatives in the LTTE. In any case, crossing into Government held areas would have, in many instances, required heading into, rather than away from, incoming artillery fire, active combat and minefields.

72. As the SLA shelled its way further into the Vanni, internally displaced persons (IDPs) moved deeper and deeper into LTTE-controlled territory, until they had nowhere left to go. Some IDPs had moved repeatedly, some for as long as two years, all the way from Mannar in the west. Each time they moved, the IDPs loaded tractors, bicycles or carts with all their belongings, taking their domestic animals if they had them. Increasingly they had to abandon belongings or, in extreme cases, even relatives. The living conditions for displaced civilians were poor and deteriorated with repeated displacements; basic necessities, including food, were increasingly scarce. The population became increasingly vulnerable and had to rely on humanitarian assistance for food and shelter. A large part of the displaced population was especially vulnerable, including women travelling alone, widowed or pregnant women, children and the elderly. The simple shelters, in which the IDPs lived, constructed of cloth or old tarpaulin, provided insufficient security. There were no adequate sanitation facilities, which imposed particular hardships for women and increased the risk of sexual violence; many people became ill. Displaced children were particularly affected by the hardships and instability caused by displacement, and were unable to continue their schooling.

1. Prelude: Exclusion of the international presence from the Vanni
73. In September 2008, the Government’s military campaign reached an advanced stage when the 57th and 58th Divisions advanced on Kilinochchi, the main LTTE stronghold and its de facto capital. The United Nations still maintained a humanitarian hub in LTTE territory in Kilinochchi, with its offices and those of other international organizations situated mainly in an area within the town known as the “Kilinochchi box”.

74. By late summer 2008, the Kilinochchi box was subject to several artillery and aerial attacks, in spite of its designation as a safe area, whose parameters were well-known to the Sri Lankan Government. Then, on 8 September 2008, the Government announced that it could no longer ensure the safety of humanitarian workers in the Vanni. It requested that the international staff of the United Nations and international non-governmental organizations (INGO) leave Kilinochchi by the end of that month. The threat to the United Nations at the time, however, was mainly posed by the SLA offensive, thus undermining the credibility of the Government in maintaining that it was not able to guarantee security. Instead, it was unwilling to do so.

75. The United Nations decided to suspend operations in the Vanni and move its offices from Kilinochchi to Vavuniya. Other international organizations withdrew their international staff as well. Nonetheless, the LTTE refused to grant permission to allow United Nations national staff to leave. A large number of the national staff from several INGOs, around 320 in total, and their dependents also remained in the Vanni. As the United Nations international staff prepared to leave Kilinochchi, aerial attacks were staged in close proximity to the United Nations premises. On the day of their departure, on or about 15 September 2009, a large crowd of civilians gathered around them, begging them not to leave, afraid of what their absence would mean.

76. The withdrawal of international staff of the United Nations and INGOs from Kilinochchi represented a pivotal point in the final stages of the war. From that moment on, there were virtually no international observers able to report to the wider world what was happening in the Vanni. The only journalists who continued to report were those embedded with the SLA or those working with the LTTE. There were reports emerging via textmessages, e-mails, phone calls and other sources originated from national staff of international organizations, religious leaders, local government employees, doctors or Tamil Net, a pro-LTTE website. But all of these sources were Tamil and were regularly contested or dismissed by the Government.

77. In January 2009, the Government scored a number of highly significant victories. In November 2008 the SLA had captured the strategically important Pooneryn and the bulk of the west coast, reopening the A32. Then on 2 January 2009, the 57th and 58th Divisions of the SLA captured Kilinochchi. Both the President and the international community urged the LTTE to lay down its arms. On 9 January 2009, the SLA 53rd and 55th Divisions captured the Elephant Pass and freed the A9, bringing the entire highway under Government control for the first time in 23 years. Later that month, on 25 January, the 59th Division captured Mullaittivu, another important LTTE base. These events marked a new stage in the acceleration and intensification of the armed conflict, one in which the ultimate defeat of the LTTE was imminent.

2. Government restrictions on humanitarian access
78. After the United Nations international staff left Kilinochchi, the United Nations Resident / Humanitarian Coordinator and the head of the World Food Programme (WFP) in Sri Lanka secured an agreement with the Government, which allowed the United Nations to continue its humanitarian assistance with weekly convoys into the Vanni to deliver food, shelter and medicine. Nonetheless, the Ministry of Defence imposed extensive restrictions on convoy participants as well as on non-food items, such as tarpaulins, which they argued could be used for military purposes. They also put limitations on food and medical supplies, as discussed in further detail below. The first convoy entered the Vanni on 3 October 2008. In total 11 convoys went into the Vanni over a period of almost 5 months, delivering a total of 7,435 metric tons of food, which was not enough to sustain the civilian population.

79. The circumstances surrounding the convoys’ travel were increasingly hazardous. During the ninth and tenth convoys, shells fell 200 metres from the road, and both the SLA and LTTE were using the cover of the convoys to advance their military positions. On 16 January 2009, the United Nations deployed its eleventh food convoy to Puthukkudiyiruppu (PTK) from Vavuniya via Omanthai. Apart from delivering humanitarian aid, some on the convoy hoped to negotiate the release of United Nations national staff and dependents by the LTTE. Convoy 11 included 7 international staff and was comprised of approximately 50 lorries, carrying essential goods such as rice, sugar, oil and wheat. The convoy off-loaded the supplies, but was not given permission to leave due to heavy fighting along the road to Vavuniya, which continued for four days.

3. SLA shelling of civilians in the first No Fire Zone
80. On 20 January 2009, the Government unilaterally declared a No Fire Zone (NFZ); Commander for the Vanni, Major General Jayasuriya announced by notice that “the Army Headquarters has demarcated this safe zone, as the Security Forces are fully committed to provide maximum safety for civilians trapped or forcibly kept by the LTTE in the un-cleared areas of Mullaittivu.” Maps of the NFZ and its coordinates were disseminated by the Government Agents (GAs). The LTTE did not accept the NFZ as binding. The rationale for the location of the NFZ, which encompassed the LTTE’s western and southern defensive lines, and the boundary of which along the A35 was only 800 metres north of the advancing SLA frontline, was not clear.

81. On or around 19 to 21 January, SLA shells hit Vallipunam hospital, located in the first NFZ, killing patients. Throughout the final stages of the war, virtually every hospital in the Vanni, whether permanent or makeshift, was hit by artillery. Particularly those which contained wounded LTTE were hit repeatedly.

82. On 21 January 2009, Convoy 11 attempted to leave PTK for Vavuniya with national staff and their dependants on board. However, the LTTE refused the convoy permission to proceed to Vavuniya due to the presence of national staff. Most of the international staff then returned to Vavuniya, leaving behind two international United Nations staff who chose to remain with the national staff.

83. On 23 January 2009, the United Nations staff relocated to the first NFZ, as a large SLA offensive on PTK seemed imminent. They set up a hub near Suthanthirapuram Junction along the A35 and relayed their coordinates to the Vanni commander. A large number of civilians also relocated to the NFZ and set up their shelters around the United Nations hub. Most civilians settled in just north of the A35, since other parts of the NFZ were not suitable for erecting shelters. The Additional Government Agent (AGA) established a food distribution centre nearby. During the day, shells fired from Government-controlled areas in the south started landing occasionally in the NFZ. In the evening, shells fell on the food distribution centre, killing and wounding a large number of civilians.

84. In the early morning hours of 24 January, hundreds of shells rained down in the NFZ. Those with access to the United Nations bunker dove into it for protection, but most IDPs did not have bunkers and had nowhere to seek cover. People were screaming and crying out for help. The United Nations security officer, a highly experienced military officer, and others present discerned that the shelling was coming from the south, from SLA positions. He made frantic calls to the head of United Nations Security in Colombo and the Vanni Force Commander at his headquarters in Vavuniya as well as the Joint Operations Headquarters in Colombo, demanding that the shelling stop, which sometimes resulted in a temporary adjustment of the shelling before it started again. Heavy shelling continued over night, and shells continued to hit the United Nations hub and the distribution centre, killing numerous civilians.

85. When United Nations staff emerged from the bunker in the first morning light at the first opportunity, mangled bodies and body parts were strewn all around them, including those of many women and children. Remains of babies had been blasted upwards into the trees. Among the dead were the people who had helped to dig the bunker the previous day.

86. Although LTTE cadre were present in the NFZ, there was no LTTE presence inside the United Nations hub. The LTTE did fire artillery from approximately 500 metres away as well as from further back in the NFZ, but the area where the United Nations was based was very clearly civilian. The Government never gave an explanation for its shelling of the United Nations hub, which was the only international presence in the NFZ.

87. Heavy shelling continued unabated. On 24 January, the Udayaarkaddu Hospital, also located in the NFZ and clearly marked with emblems, was hit by several shells. During the night of 25 January, the first NFZ and area around the United Nations hub continued to be pounded with shells. During the two days of shelling in the first NFZ, hundreds of civilians were killed and many more injured. To escape the intense shelling, civilians started to flee the NFZ, away from the SLA, heading north to Iranapalai or back to PTK. The United Nations contingent, too, decided to leave the NFZ to return to PTK, along with the International Committee of the Red Cross (ICRC) and the AGA. After they left, hundreds of desperate civilians ran to seek shelter in the deserted bunker.

88. The scene inside the NFZ along the road to PTK, the A35, was one of great destruction, and even the vegetation was shredded. Dead or severely injured civilians lay along the roadsides, amidst shattered shelters, strewn belongings and dead animals. Hundreds of damaged vehicles also lay along the road; ambulances parked by Vallipunam Hospital were seriously damaged.

89. This was in stark contrast to the situation outside the NFZ, across the Yellow Bridge (Manjal Palam), further along the A35, where there were few signs of shelling. Paradoxically, while PTK was outside the Government-designated NFZ, that area did not seem to have been shelled, in spite of the presence of a large number of LTTE and far fewer civilians.

4. SLA shelling of PTK Hospital
90. Fighting in the area intensified as part of the expressed efforts by the 55th and 58th Divisions to capture PTK by 4 February, the day commemorating Sri Lanka’s independence. PTK hospital was the only permanent hospital left in the Vanni, and its neutrality was recognized by the Government and the LTTE. The medical staff, including five doctors, was stretched beyond its capacity, and medical supplies were very limited. The shelling in the first NFZ had marked a turning point in the conflict, and civilian casualties were rising. PTK hospital was packed with hundreds of injured civilians from the NFZ. More than 100 new patients were arriving each day, many from the NFZ. Many had severe or life-threatening injuries caused by artillery fire or burns. The casualties, many of them babies, young children and the elderly, were packed in every conceivable space – on beds, under tables, in hallways and outside in the driveway.

91. On 29 January 2009, the two remaining United Nations international staff left for Vavuniya, without the national staff members, who were still not allowed to leave by the LTTE. The ICRC dispatched a separate convoy, which evacuated about 200 wounded patients. Immediately thereafter, in the week between 29 January and 4 February, PTK hospital was hit every day by MBRLs and other artillery, taking at least nine direct hits. A number of patients inside the hospital, most of them already injured, were killed, as were several staff members. Even the operating theatre was hit. Two ICRC international delegates were in the hospital when it was shelled on 4 February 2009. The shelling was coming from SLA positions.

92. The GPS coordinates of PTK hospital were well known to the SLA, and the hospital was clearly marked with emblems easily visible to UAVs. On 1 February 2009, the ICRC issued a public statement emphasizing that “[w]ounded and sick people, medical personnel and medical facilities are all protected by international humanitarian law. Under no circumstance may they be directly attacked.”

93. The Ministry of Human Rights and Disaster Management responded by accusing the ICRC of “either wilful ignorance or naiveté”. Initially, the Government denied shelling the hospital, but on 2 February 2009, the Defence Secretary gave the following statement in an interview on Sky News:


 * If they [reports] are referring to the [PTK] hospital, now there shouldn’t be a hospital or anything because we withdrew that. We got all the patients to Vavuniya, out of there. So nothing should exist beyond the No Fire Zone … No hospital should operate in the area, nothing should operate. That is why we clearly gave these No Fire Zones… For the LTTE …to crush the terrorists, there is nothing called unproportionate.

94. After the fall of Kilinochchi, PTK was a strategic stronghold in the LTTE’s fight against the SLA. As a result, the LTTE had a sizable presence in the PTK area and maintained a separate ward for wounded cadres in PTK hospital, but they were not armed. The frontline was nearby, and as the fighting in the PTK area increased, more LTTE wounded started to come into the hospital. The LTTE also fired mobile artillery from the vicinity of the hospital, but did not use the hospital for military purposes until after it was evacuated. Yet, in its eagerness to capture the area, the SLA repeatedly shelled the hospital and surrounding areas. Due to the incessant shelling, the Regional Directors of Health Services (RDHS), the United Nations, the AGA and the ICRC decided to evacuate some 300 patients in PTK hospital to Putumattalan, around 6 to 8 kilometres away, on the coastal strip next to the Nanthikadal lagoon. Ponnambalam Hospital, a private hospital used in part by the LTTE, was shelled on 6 February 2009, causing part of it to collapse.

95. The SLA suffered significant setbacks and many casualties at PTK, in battles with LTTE forces, who put up a fierce resistance. The 58th and 53rd Division did not capture PTK until 5 April 2009. The important Sea Tiger stronghold of Chalai was captured by the 55th Division on 5 February.

96. At that time, large numbers of civilians, trying to escape fierce fighting in Anandapuram, Iranapalai and Thevipuram, fled towards the coast, since it was the last remaining haven. On 12 February 2009, the Government declared a second No Fire Zone (also referred to as a Civilian Safety Zone), covering a 12 km-long strip along the coast, including the villages of Ampelavanpokkanai, Karayamullivaikkal, Putumattalan, Valayanmadam, and Vellamullivaikkal.

5. LTTE forced recruitment and forced labour in the second No Fire Zone
97. Increasingly, LTTE forces, mounting their last defence, moved onto the coastal strip in the second NFZ, particularly in the Mullivaikkal area, where the LTTE leadership had a complex network of bunkers and fortifications and where it ultimately made its final stand. The LTTE was no longer mobile and established a series of defensive earth bunds throughout the zone. Its positioning of mortars and other artillery among IDPs often led to retaliatory fire by the Government, often resulting in civilian casualties. LTTE cadre were not always in uniform at this stage. Since the loss of the Jaffna and Kilinochchi fronts, the LTTE’s supply lines and logistical systems began to fail, as it was almost impossible for them to shift supplies and maintenance installations to the Mullaittivu district. As a result, the LTTE lost access to fuel, ammunition and some of its food supplies. While the LTTE had some weapons caches left, they had only limited heavy artillery, including a small number of MBRLs. The SLAF had complete air superiority, and the LTTE had to camouflage its installations to make them more difficult to identify from the air.

98. In spite of the futility of their military situation, the LTTE not only refused to surrender, but also continued to prevent civilians from leaving the area, ensuring their continued presence as a human buffer. It forced civilians to help build military installations and fortifications or undertake other forced labour. It also intensified its practice of forced recruitment, including of children, to swell their dwindling ranks. As LTTE recruitment increased, parents actively resisted, and families took increasingly desperate measures to protect their children from recruitment. They hid their children in secret locations or forced them into early arranged marriages. LTTE cadre would beat relatives or parents, sometimes severely, if they tried to resist the recruitment. All these approaches, many of them aimed at defending the LTTE and its leadership, portrayed callousness to the desperate plight of civilians and a willingness to sacrifice their lives.

99. Nonetheless, as the situation in the second NFZ worsened, growing numbers of civilians sought to escape LTTE-controlled areas. Civilians waded long distances through the lagoons or across mine-ridden territory, often in the dead of night. Inevitably people stepped on landmines and lost their limbs or were fatally injured. Beginning in February, the LTTE commenced a policy of shooting civilians who attempted to escape, and, to this end, cadre took up positions where they could spot civilians who might try to break out.

6. SLA shelling in the second No Fire Zone
100. From as early as 6 February 2009, the SLA continuously shelled within the area that became the second NFZ, from all directions, including land, air and sea. It is estimated that there were between 300,000 and 330,000 civilians in that small area. The SLA assault employed aerial bombardment, long-range artillery, howitzers and MBRLs as well as small mortars, RPGs and small arms fire, some of it fired from a close range. MBRLs are unguided missile systems designed to shell large areas, but if used in densely populated areas, are indiscriminate in their effect and likely to cause large numbers of casualties.

101. At the time, the Defence Secretary stated: “We are taking casualties to prevent civilians getting hurt. This is a factor we are very concerned about. Otherwise we could have used so much artillery and just moved on.” The Government announced on 25 February, and again on 27 April, that the SLA was no longer using heavy weapons in the second and third No Fire Zones. But what was happening on the ground indicated the opposite. Intensive artillery fire had been a core tactic in the SLA’s military campaign from the outset. As victory neared, this tactic was not abandoned, but rather its use was intensified, even though the LTTE was now immobilized and surrounded in an area of high civilian density. The intensive shelling also caused many civilians to attempt to flee the area, meeting another of the Government’s objectives, to put pressure on civilians to get out of the way. Despite Government pronouncements, satellite images in Annex 3 show that SLA artillery batteries were constantly adjusted to increasingly target the NFZs. The LTTE had fewer heavy weapons left and less space to fire them from.

102. The coastal strip became increasingly crowded, and liveable spaces were in short supply. Much of the land where IDPs set up shelters was beach territory, with sandy, waterlogged land unsuitable for human habitation, and it was difficult for IDPs to construct makeshift bunkers to protect themselves. Daily life for the IDPs at that time took place mostly inside the bunker, although some IDPs hoisted white flags over their shelters in an attempt to protect themselves. Fresh water was scarce and food was in such short supply that a few people died of starvation. When the seasonal rains came, many bunkers were flooded, adding to the general misery of the people.

7. Shelling of Putumattalan Hospital
103. When the PTK hospital relocated to Putumattalan, the Government stated that “there are now no hospitals functioning in uncleared areas in the Vanni”. Nonetheless, the second NFZ had three makeshift hospitals, including Putumattalan, a small clinic at Valayanmadam and a hospital in Mullivaikkal. All of their coordinates were known to the Government, and they were clearly marked with emblems. Government doctors continued providing their services there. Putumattalan hospital was severely overcrowded with hundreds of newly injured civilians. As the Government did not allow basic medical supplies into the Vanni, conditions in Putumattalan hospital were so poor that a large number of amputations were performed without anaesthetic, using butcher knives rather than scalpels. Sanitary pads and cotton cloths were used as bandages, and intravenous drips were hung from the trees, with the severely-injured patients lying on the ground under them. In spite of the significant efforts of the few available doctors, many patients died due to lack of access to proper medical care, and scores of bodies were deposited in front of the hospital each day.

104. On 9 February 2009, shells fell on Putumattalan hospital, killing at least16 patients. The shells came from SLA bases in Chalai, but subsequently shells were also fired from SLA positions across the lagoon (even though the hospital was clearly visible to the SLA based there). While some wounded LTTE cadre were treated at Putumattalan hospital, they were few in number and were kept in a separate ward. Putumattalan hospital was shelled on several occasions after that, in February and March. RPGs were fired at the hospital around 27 March killing several civilians. In addition to civilian casualties, the operating theatre, makeshift ward and roof all sustained damage.

105. While individual incidents of shelling and shooting took place on a daily basis, destroying the lives of many individuals or families, the SLA also shelled large gatherings of civilians capable of being identified by UAVs. On 25 March, an MBRL attack on Ambalavanpokkanai killed around 140 people, including many children. On 8 April 2009, a large group of women and children, who were queued up at a milk powder distribution line organized by the RDHS, were shelled at Ambalavanpokkanai. Some of the dead mothers still clutched cards which entitled them to milk powder for their children.

8. Hindrance of humanitarian assistance via the ICRC ships
106. The ICRC continued to play a leading role in alleviating the plight of the civilian population in the Vanni, by evacuating wounded civilians from the coastal strip by ship, starting on 10 February 2009. In total, 16 ICRC ships came to the conflict zone in the final months. The international ICRC staff that had remained in Putumattalan left on the first ship, but they returned and stayed onshore for a few hours each time the ships came back. The Government did not allow United Nations staff on the ships.

107. The LTTE issued passes for injured civilians and some of their dependents to leave the area on ICRC ships, but the wounded had to be ferried on small boats, as the ship was not allowed to come closer than a kilometre offshore. The wounded were lined up on the beach, but several times came under fire. Shells fired by the SLA sometimes fell in the sea near the ICRC ships. Around 22 April, shelling near a ship forced the captain to return to deeper waters.

108. The ICRC’s ships were also the only means for delivering food, but the supplies they were allowed to bring by the Government were inadequate. As conditions in the NFZ became more desperate, on 17 March, a large crowd of IDPs surrounded an international ICRC staff member who came ashore, begging him to save their lives by taking them out of the Vanni. The LTTE forcibly dispersed the crowd. The final ICRC ship came to the Vanni on 9 May 2009. On 15 May 2009, a ship approached, but had to turn back due to the intensity of the fighting. In all, ICRC evacuated 14,000 wounded persons and their relatives from the second and third NFZs and delivered around 2,350 metric tons of food to Mullivaikkal. Those evacuated were all civilians, as the LTTE did not permit its cadre to leave the conflict area for treatment.

9. SLA shelling including Mullivaikkal Hospital
109. By early April 2009, after their defeat at Anandapuram, the LTTE’s remaining forces had almost entirely retreated onto the coastal strip. The Government shelling intensified inside the NFZ. Although it relied mainly on mortars, throughout the conflict, until the final moments, the Government continued to use heavy weapons such as MBRLs and aerial bombardment, although it said that it was conducting a “humanitarian rescue” of the hostage civilian population at that stage. On 19 April 2009, the area between Putumattalan and Amparanpokkanai was shelled intensively, and the SLA 58th Division came onto the coastal strip for the first time, breaking through LTTE defences, dividing the NFZ into two, but inflicting heavy civilian casualties at the same time. The division of the NFZ into two parts enabled a group of around 100,000 civilians to escape to Government-controlled territory, in addition to the 70,000 or so who had already come out. At least another 130,000 civilians remained trapped further south.

110. After the SLA captured the north of the NFZ, Mullivaikkal Hospital was the only remaining hospital in the conflict zone. There were no LTTE cadre in uniform in the hospital, nor did anyone bring weapons inside. Conditions were extremely poor. The hospital had four doctors and ran two improvised operating theatres. Some of the patients, including those with serious head injuries and other obvious fatal injuries, were merely made comfortable, but no attempt could be made to save them. With few beds available, wounded patients often remained in front of the hospital, some on mats and others lying on dust and gravel, under sheets set up for shelter, cradled by their loved ones or alone. With a severe shortage of gauze or other sterile bandages, old clothes or saris were used as bandages. No gloves were available, and the conditions were grossly unhygienic, giving rise to a high risk of infections. In this hospital, amputations were also performed with butcher knives, due to the lack of surgical equipment, and amputated limbs were collected in piles. On many occasions amputations were performed to save the life of the patient, as there was simply no other way to treat wounds. Due to the severe shortage of anaesthetics, the little that remained was mixed with distilled water, but many amputations were performed without anaesthesia. In spite of widespread malnutrition, some people continued to donate blood, but a general shortage of blood meant that a patient’s own blood was often used, caught in a plastic bag, to be filtered through a cloth and re-transfused back into the same patient.

111. Due to the heavy shelling that hit the hospital on numerous occasions, the RDHS moved to a second location at Vellamullivaikkal. On 11 or 12 May, the second hospital was also hit by SLA shells, killing many people, although it, too, was prominently marked. The conditions in the second hospital were as poor as the first, and some of the hospital staff members were killed by SLA shelling.

10. LTTE killing of civilians and forced recruitment
112. As the situation in the second NFZ worsened, large numbers of civilians tried to escape LTTE-controlled areas, but the LTTE sought to prevent this with increasing brutality. Some LTTE cadre would let fleeing civilians through, but others opened fire on them with AK47s, killing men, women and children, alike. The IDPs, who attempted escape, desperately tried to run away and to reach SLA lines, carrying their children or luggage or dropping them in their panic. Some were killed on the spot; others flailed in the shallow water or incurred terrible injuries from stepping on landmines. Small children and others drowned in the lagoon. While it is not known precisely how many people died this way, the number was significant and rose as the armed conflict progressed.

113. Desperate for new troops, the LTTE again stepped up its policy of forced recruitment, dragging away more and more youngsters, including the under-aged, to be used in the first lines of defence. On one occasion in mid-April, LTTE cadre, led by the former Trincomalee Political Wing leader known as Ezhilan, forcibly recruited hundreds of young people from Valayanmadam Church and put them on buses to Mullivaikkal. Parents begged and cried for them not to be taken away to fight and to an almost certain death, but to no avail.

114. On 26 April the LTTE declared a unilateral ceasefire, but it was not accepted by the Government, which referred to it as a “joke”. On 27 April 2009, the Government announced for the second time that combat operations against the LTTE had concluded and that it was ceasing the use of heavy weapons. On or about 8 May 2009 the Government declared a third and final NFZ, which was very small section in the south of the second NFZ.

11. SLA shelling during the final days (13-18 May)
115. On May 13, the 58th Division was pushing its way forward from the east towards the coastline with the aim of advancing south from there, with the 53rd Division marching east along the A35 road towards the lagoon. Troops from the 55th Division pushed further south from Putumattalan. At that point, the United Nations estimated that 100,000 civilians remained trapped within three square kilometres, whereas the Government claimed there were only 10,000.

116. The final days of the armed conflict saw a steep rise in the number of civilian casualties. At the hospital at Vellamullivaikkal, hundreds of patients were lying on the ground, bleeding from terrible wounds. The dead lay intermingled with the gravely injured. The relatives caring for the wounded were themselves malnourished and weak.

117. The shelling within the third NFZ was such that it was impossible for the ICRC to conduct any more maritime rescues. As the SLA neared the hiding places of the senior LTTE leadership, its offensive assumed a new level of intensity, in spite of the thousands of civilians who remained trapped in the area. The LTTE leadership, in turn, sent many cadre to die in their defence, including through suicide missions.

118. Due to the lack of space in the third NFZ, civilians had nowhere to hide from the shelling, which was coming in from all sides. Shells rained down everywhere and bullets whizzed through the air. Many died and were buried under their bunkers or shelters, without their deaths being recorded. Black smoke and the stench of dead bodies filled the air. Some people begged for food for their starving children or for help for the wounded or dying. The scene was described as reminiscent of hell.

119. In spite of many desperate telephone calls by the AGA and doctors to stop the shelling to allow them to attend to the wounded and dead, no reprieve was forthcoming from the SLA. After 14 May 2009, the doctors could no longer go to the hospital due to the intensity of the shelling, and it had to be closed. Dozens of patients who could not be moved were left behind. All survivors huddled together in rudimentary shelters. Cooking was impossible and leaving the shelter even for sanitary purposes meant risking one’s life. Some civilians tried to stage a mass breakout, but were shot at and shelled by the LTTE. Those who managed to escape were helped across by individual SLA soldiers.

120. On 15 May, the LTTE began destroying their communications equipment. On 16 May, a large explosion rocked the LTTE-area, and a fire destroyed hundreds of IDP shelters. That same day, the 58th and 59th Divisions of the SLA linked on the coastline, and Army Commander Lieutenant General Fonseka declared victory against the LTTE. The 53rd Division continued to make its way south, along the Nanthikadal lagoon. The remaining LTTE, including many of the top leaders and around 250 hard-core fighters, were locked into a small area of around 3 square kilometres at Vellamullivaikkal. The end was near; the circumstances surrounding the deaths of many of those leaders are the subject of controversy.

121. On 18 May 2009, Defence Ministry sources said that Prabhakaran, Soosai and Pottu Amman were killed while trying to break out of the NFZ. Charles Anthony, Nadesan and Pulidevan were also among the dead. On 19 May, the Government of Sri Lanka officially announced that Prabhakaran and his key aides had been killed and showed their bodies on television. Many photographs of the corpses emerged later. The same day the President gave a speech in Sri Lanka’s Parliament declaring victory over the LTTE.

122. Between 16 and 19 May, the remaining civilians trapped in the zone made their way south, out of the coastal strip, crossing the Vadduvahal Bridge into the Government controlled area. The shelling continued and large fires were burning (including destroyed arms caches or weaponry from the LTTE). The dead were strewn everywhere; the wounded lay along the roadsides, begging for help from those still able to walk, but often not receiving it. Some had to be torn away from the bodies of their loved ones left behind. The smell of the dead and dying was overwhelming.

123. 18 May 2009 marked the end of the armed conflict in the Vanni. In the words of the ICRC, the final days had culminated in “unimaginable humanitarian catastrophe”.

D. Disputing IDP figures as a basis to deny humanitarian assistance
124. Throughout the final stages of the armed conflict, particularly from January to May 2009, the Government downplayed the number of civilians present in the LTTE-controlled area, using the low estimates to restrict the amount of humanitarian assistance that could be provided, especially food and medicine.

125. At the outset of the final phase, on 13 January 2009, the Government website reported that, according to independent verifications, the number of civilians in the Vanni was between 150,000 and 250,000. The United Nations estimate at the time was 250,000 (although its subsequent estimates were higher). Later in January 2009, the Ministry of Defence said that the number of civilians present in the Vanni was between 75,000 and 100,000, “on a high estimate”. However, the Government had more than sufficient information at its disposal during the final stages of the armed conflict to accurately estimate the actual number of civilians in the Vanni. Each month the GAs continued to collate data on IDPs in order to make requests for dry rations from WFP. Prior to September 2008, numbers compiled by the GAs of Mullaittivu and Kilinochchi indicated that there were around 420,000 people in the LTTE-controlled area at that time. While these numbers may have been inflated, the United Nations Children’s Fund (UNICEF) estimates of school children registered in the Vanni were 70,000, which was approximately the same as to the Government’s estimate for the total IDP population.

126. The subsequent numbers given by the Ministry of Defence varied, but in general they were deliberately kept low, and some Government employees working in the zone were reprimanded, when they provided other figures or different calculations of need. For instance, on 2 February 2009, the AGA based in the second NFZ sent a situation report to the Ministry of Public Administration and Home Affairs stating that there were about 81,000 families present in Mullaittivu District at that time, totalling some 330,000 persons. However, on 18 March, the AGA received a response from the Secretary of the Ministry of National Building and Estate Infrastructure Development, stating that the figure of 330,000 was “arbitrary and baseless” and that the Government would be “reluctantly compelled” to take disciplinary action against him for providing “wrong information to any source especially in regard to IDP figures”.

127. At the end of February 2009, the United Nations Country Team informed the Government that, in its view, there were 267,618 civilians present in the LTTE-controlled area, basing the estimate, in part, on UNOSAT Quickbird and Worldview satellite images, used to count the number of IDP shelters. At the end of April, United Nations estimates were that 127,177 civilians still remained trapped, whereas the Government said there were only 10,000 persons left at the time. The number of IDPs who eventually emerged from the area and were housed at Menik Farm and in other camps was approximately 290,000. The discrepancy in these figures has not been adequately explained by the Government.

128. As a result of the Government’s low estimates, the food delivered by WFP to the Vanni was a fraction of what was actually needed, resulting in widespread malnutrition, including cases of starvation. Similarly, the medical supplies allowed into the Vanni were grossly inadequate to treat the number of injuries incurred by the shelling. Given the types of injuries sustained in the second NFZ, the doctors requested medical supplies such as anaesthetics, blood bags for transfusion, antibiotics, surgical items, gloves and disinfectant. Only a small quantity of these items was allowed into the Vanni. Instead, they received items such as Panadol, allergy tablets and vitamins. As the casualty figures rose in March 2010, the absence of the needed medical supplies imposed enormous suffering and unnecessarily cost many lives. The RDHS doctors repeatedly spoke out about the inadequacy of medical supplies, in letters and televised interviews. They also compiled and communicated photographs and lists of the names of the injured and dead. They were warned by the Ministry of Health to stop speaking to the media and stop complaining, or be punished. Drs. Sathyamoothy and Varatharajah forwarded a report, “Undue Deaths due to Non-Availability of Essential Drugs at Mullaittivu”, to the Government on 16 March, stating:


 * Most of the hospital deaths could have been prevented if basic infrastructure facilities and essential medicines were made available … We have been supplied with no antibiotics, no anaesthetics and not even a single bottle of IV fluid, leaving us in a desperate situation of not being able to provide even lifesaving emergency surgery.

129. On 19 March 2009, the Secretary of the Ministry of Healthcare and Nutrition replied that only strong painkillers and intravenous fluids could be dispatched, since Mullivaikkal Hospital did not have trained anaesthesiologists. The letter also warned the doctors not to violate protocols, by addressing copies of their letters to the Indian High Commission or the Chief Minister of Tamil Nadu, or else disciplinary action would be taken “for violating procedure and embarrassing the Government”.

130. When the doctors exited the conflict zone on 16 May, they were detained and interrogated for several months. In early July 2009, the doctors gave a press conference, in which they said that there were, in fact, very few civilian deaths and injuries during the war and that they had been forced to lie about it by the LTTE. This retraction contradicts what they had said in interviews, e-mails and public statements while they were still in the Vanni. The Panel believes they were put under pressure by the Government, and that these retractions do not affect the veracity of their earlier statements.

131. Despite its access to first-hand information regarding the size of the civilian population and its needs, the Government of Sri Lanka deliberately used greatly reduced estimates, as part of a strategy to limit the supplies going into the Vanni, thereby putting evergreater pressure on the civilian population. A senior Government official subsequently admitted that the estimates were reduced to this end. The low numbers also indicate that the Government conflated civilians with LTTE in the final stages of the war.

E. The number of civilian deaths
132. There is no authoritative figure for civilian deaths or injuries in the Vanni in the final phases of the war. Several factors make it very difficult to calculate a reliable casualty figure:
 * (a) the number of persons in the conflict area remains uncertain, although it was likely to have been as many as 330,000;
 * (b) the lack of an accurate count of the number of persons who emerged from the Vanni, due to the lack of transparency in the screening process;
 * (c) lack of certainty on the numbers of LTTE combatants, complicated further by the increase in forced recruitment in the final phase; and
 * (d) the fact that many civilians were buried where they fell, without their deaths being registered, in some cases, unobserved.

133. Some have developed estimates based on the statistics of the injured and dead collected by the doctors, which were collated by the hospitals and the District Disaster Management Unit. One estimate is that there were approximately 40,000 surgical procedures and 5,000 amputations performed during the final phase. Depending on the ratio of injuries to deaths, estimated at various times to be 1:2 or 1:3, this could point to a much higher casualty figure. Others have put the estimate at 75,000, a figure obtained by subtracting the number of people who emerged from the conflict zone (approximately 290,000) from the estimate of the number thought to have been in the conflict zone (approximately 330,000 in the NFZ from January, plus approximately the 35,000, who emerged from the LTTE-held areas before that time).

134. The United Nations Country Team is one source of information; in a document that was never released publicly, it estimated a total figure of 7,721 killed and 18,479 injured from August 2008 up to 13 May 2009, after which it became too difficult to count. In early February 2009, the United Nations started a process of compiling casualty figures, although efforts were hindered by lack of access. An internal “Crisis Operation Group” was formed to collect reliable information regarding civilian casualties and other humanitarian concerns. In order to calculate a total casualty figure, the Group took figures from RDHS as the baseline, using reports from national staff of the United Nations and NGOs, inside the Vanni, the ICRC, religious authorities and other sources to cross-check and verify the baseline. The methodology was quite conservative: if an incident could not be verified by three sources or could have been double-counted, it was dismissed. Figures emanating from sources that could be perceived as biased, such as Tamil Net, were dismissed, as were Government sources outside the Vanni.

135. The number calculated by the United Nations Country Team provides a starting point, but is likely to be too low, for several reasons. First, it only accounts for the casualties that were actually observed by the networks of observers who were operational in LTTE controlled areas. Many casualties may not have been observed at all. Second, after the United Nations stopped counting on 13 May, the number of civilian casualties likely grew rapidly. Due to the intensity of the shelling, many civilians were left where they died and were never registered, brought to a hospital or even buried. This means that, in reality, the total number could easily be several times that of the United Nations figure.

136. It is worth noting that the United Nations raised casualty figures in private entreaties with the Government, but never publicized its specific estimates. Government officials strongly refuted the figures provided by the United Nations, stating that the numbers were fabricated and that this was not the business of the United Nations. Publicly the United Nations referred to the “heavy toll” of the fighting on civilians, or that the casualty figures were “unacceptably high”, but that the actual figures were not verifiable. The decision not to provide specific figures made the issue of civilian casualties less newsworthy. However, this position was maintained by senior United Nations officials until 13 March 2009, when the High Commissioner for Human Rights publicly stated that 2,800 civilians may have been killed and more than 7,000 injured since 20 January, many of them inside the NFZs. Pressure from the Government of Sri Lanka and fears of losing access may have resulted in a general under-reporting of violations by United Nations agencies. Some have criticized the failure of the United Nations to present figures publicly as events were unfolding, citing it as excessively cautious in comparison with other conflict situations.

137. In the limited surveys that have been carried out in the aftermath of the conflict, the percentage of people reporting dead relatives is high. A number of credible sources have estimated that there could have been as many as 40,000 civilian deaths. Two years after the end of the war, there is still no reliable figure for civilian deaths, but multiple sources of information indicate that a range of up to 40,000 civilian deaths cannot be ruled out at this stage. Only a proper investigation can lead to the identification of all of the victims and to the formulation of an accurate figure for the total number of civilian deaths.

F. Credible allegations relating to events outside the conflict zone and in the aftermath
138. The plight of civilians who had survived the conflict in the Vanni did not end when they entered Government-controlled areas. In spite of Government pronouncements that it was ready to receive a mass exodus of civilians from the Vanni as early as January 2009, the Government failed to prepare adequately for the time when large numbers did emerge and then had trouble coping. In general, the Government gave priority to security considerations over the humanitarian needs and well-being of the IDPs.

139. When they emerged from the conflict zone, many civilians were fearful of the reception they would receive. They were severely traumatized and exhausted as a consequence of their recent experience. Many of them were newly widowed, orphaned or disabled. Tens of thousands of IDPs had conflict-related injuries, with at least 2,000 amputees among them. The situation, as large numbers exited, was chaotic, and many family members were separated from each other. In the process, many families were divided and placed in separate camps; provision for family tracing and reunification was inadequate, and the ICRC was not authorized to play a role in this regard.

140. Family separation left many women on their own and vulnerable to sexual violence. Pregnant or lactating women had suffered from lack of adequate nutrition, medical care, and enormous psychological strain while in the conflict zone. Forced recruitment of children also took a heavy toll on mothers.

141. The conflict took a particular toll on the young. Children as young as 14 had been the target of forced recruitment by the LTTE. Measures to avoid recruitment, including early marriages, had a detrimental impact on the health of young girls. In addition, thousands of children suffered violations such as killing and maiming, due to the shelling. Some were killed because they had ventured out of the bunker to play. Children were particularly vulnerable to horrific injuries as shrapnel ripped at their small limbs. A Rapid Nutrition Assessment showed that around 25 per cent of children suffered from acute malnutrition.

142. Many children suffered from the adverse psychological impact of multiple displacements. Many had lost their parents, emerging unaccompanied and were not registered. Most children were malnourished, and many babies suffered from dehydration or diarrhoea.

143. Likewise, the elderly were particularly affected by the conflict. In the multiple displacements, the elderly and others who could no longer walk, were often left behind. Some were abandoned when their relatives fled. Others had nobody to care for them in the IDP camps and died of neglect, exhaustion and preventable diseases.

1. Violations during the screening process
144. On leaving the Vanni and arriving in the Government-controlled areas at Vadduvahal Bridge and other locations, survivors of the armed conflict surrendered to the SLA. Incoming civilians were separated into different groups. First, the SLA generally strip-searched and checked them for weapons and explosives. Laptops and cameras (for the few that had them) were confiscated by security forces, leading to the loss of valuable information. People were then transferred, often by foot, to initial screening sites set up in places such as Kilinochchi, Pulmoddai and Padaviya. At these sites, the SLA called those who had been associated with the LTTE, even for a day, to identify themselves and surrender, and promised vocational training and employment abroad for those who did. Instead, those identified as LTTE were taken to separate camps. A significant number of suspected LTTE were women and children.

145. In addition, the Government used former LTTE cadre from the Karuna faction or People’s Liberation Organization of Tamil Eelam (PLOTE) to identify suspected LTTE cadre, who were separated and taken to other locations. The Government purposefully prevented international humanitarian agencies from accessing the initial screening sites.

146. After this initial screening, surviving civilians were transported to a further screening site at Omanthai. Although men and women were screened separately, as part of the screening process, people were generally forced to strip naked, causing humiliation and increased vulnerability, particularly among women and girls. Médecins Sans Frontières (MSF), Office of the United Nations High Commissioner for Refugees (UNHCR) and ICRC had some access to Omanthai, but were not allowed to interview people in private. After July 2009, the ICRC was excluded altogether.

147. Civilians in need of medical attention were transferred to hospitals in Vavuniya or the clinic staffed by Indian doctors at Pulmoddai. Vavuniya Hospital was overflowing with patients, leading to early discharges, and all patients were closely guarded by the SLA and subject to interrogation by police investigators (Criminal Investigation Department, CID, or Terrorist Investigation Department, TID). Some patients disappeared from the hospitals.

148. In particular, the screening process resulted in cases of executions, disappearances, and rape and sexual violence.

(a) Executions
149. Authenticated footage and numerous photographs indicate that certain LTTE cadre were executed after being taken into custody by the SLA. Photographs available to the Panel show many dead bodies of cadre (or possibly civilians), some with their hands tied behind their back. On 25 August 2009, the UK-based Channel 4 News released video footage, which showed the summary execution by Sri Lankan soldiers of several prisoners with their hands tied behind their backs. The prisoners in the footage are naked and blindfolded. They are kicked and forced to cower in the mud before being shot in the head at close range. The film shows several other prisoners who appear to have been killed earlier. A second film of the same scene, also released by Channel 4, on 2 December 2010, pans out over the landscape, showing the bodies of a number of other naked and executed prisoners, male and female. Among them are a young boy and a woman; the woman has been identified as a well-known LTTE media anchor known as “Isaipriya”. Notably, Isaipriya is listed on the Defence Ministry website as killed on 18 May 2009 in a “hostile operation” by the 53rd Division. The extended video shows the faces of some of the soldiers and shows persons filming the scene with cell phones.

150. Photographs that appear to be taken before the executions show what appears to be the boy, sitting in a group of prisoners, who were alive, with their hands tied behind their back. The persons in the photograph are clearly terrified. When first detained by the SLA, some suspected LTTE cadre were also tortured. Photographs show bodies with signs of torture; a video shows a young man who has been tied to a tree and is covered in blood. He later appears dead, lying in a grave covered by a Tiger flag.

(b) Disappearances
151. The Government has not provided a public registration of persons at screening sites or Omanthai, neither did it allow international organizations to monitor the process. This makes it difficult to trace persons. During hearings by the Lessons Learnt and Reconciliation Commission (LLRC), a number of women gave accounts of how their husbands or relatives were taken from them when they first entered the Government-controlled area and that they have not been seen since and to date, the Government has not confirmed their whereabouts. At least 32 submissions made to the Panel alleged disappearances in May 2009, some of them dealing with groups of persons rather than individuals. Many of these were persons who had surrendered to the SLA.

(c) Rape and sexual violence
152. Rape and sexual violence against Tamil women during the final stages of the armed conflict and, in its aftermath, are greatly under-reported. Cultural sensitivities and associated stigma often prevented victims from reporting such crimes, even to their relatives. Nonetheless, there are many indirect accounts reported by women of sexual violence and rape by members of Government forces and their Tamil-surrogate forces, during and in the aftermath of the final phases of the armed conflict.

153. Many photos and video footage, in particular the footage provided by Channel 4, depict dead female cadre. In these, women are repeatedly shown naked or with underwear withdrawn to expose breasts and genitalia. The Channel 4 images, with accompanying commentary in Sinhala by SLA soldiers, raise a strong inference that rape or sexual violence may have occurred, either prior to or after execution. One video shows SLA soldiers loading the naked bodies of dead (or nearly dead) women onto a truck in a highly disrespectful manner, in one case, stomping on the leg of a woman who appears to be moving. Rapes of suspected LTTE cadre are also reported to have occurred, when they were in the custody of the Sri Lankan police (CID and TID) or the SLA. International agencies also recorded instances of rape in the IDP camps, but the military warned IDPs not to report cases of rape to the police or to humanitarian actors.

(a) Arbitrary detention of IDPs in closed camps
154. Civilians emerging from the conflict zone were initially housed in a network of 21 IDP sites spread across Jaffna, Mannar, Trincomalee and Vavuniya districts. Most were eventually sent to Menik Farm near Vavuniya, which, at its peak, housed around 250,000 IDPs, making it one of the largest IDP sites in the world and one of the largest population centres in Sri Lanka.

155. Menik Farm and other IDP sites were closed camps, guarded by the military and surrounded by barbed wire. Essentially, the entire Vanni IDP population was detained and not allowed to leave. The Government held that the detention of the entire IDP population was necessary until the screening could be completed and the Vanni sufficiently cleared of landmines. Screening continued inside Menik Farm. Paramilitaries from former Tamil militant groups, often wearing balaclavas, roamed around, often at night, outside the scrutiny of humanitarian organizations, to select and remove people they claimed had links to the LTTE.

156. At Menik Farm, severe restrictions prevented international organizations from doing protection work or speaking to the IDPs in private. ICRC initially had access to Menik Farm for a short period, but was soon excluded. The restrictions suggest an attempt by the Government to prevent those who came out of the conflict zone from relaying their experiences to international agencies and NGOs. The absence of external and independent monitoring also increased the vulnerability of IDPs to violations in the camp, including exposure of women without male relatives and unaccompanied children to sexual and other forms of violence.

157. Prior to the establishment of Menik Farm, international agencies, including the International Organization for Migration (IOM), UNHCR, UNICEF and others, debated amongst themselves about conditioning their provision of humanitarian assistance on the Government’s meeting international standards with regard to the camps. Several communications on the applicable standards were sent to the Sri Lankan Government by agencies, such as UNHCR, and by NGOs. However, when IDPs came out in larger numbers, the international agencies failed to take a common position on the pre-conditions. Many international agencies continued to provide assistance, in spite of the dramatically substandard conditions that prevailed at Menik Farm.

158. The detention of the IDP population lasted for months or in some cases, years. By December 2009, around 149,000 IDPs had been released, with another 135,000 remaining in the camps. By September 2010, the Government said it had released 242,741 IDPs, with 25,795 still waiting to be released.

(b) Inhumane camp conditions
159. While the Government referred to Menik Farm as a “welfare village” for IDPs, it was located in the middle of the jungle, without its own water source. After the large influx of IDPs in April and May 2009, conditions in Menik Farm were far below international standards. These conditions imposed additional unnecessary suffering and humiliation on civilians. New arrivals often had not eaten for days. While many persons suffered from depression, psychological support was not allowed by the Ministry of Social Services, and some IDPs committed suicide. Some died while awaiting passes to get basic medical treatment or died from preventable diseases.

160. Extreme overcrowding in the camps forced some people into unsafe living conditions. Provision for food, water, shelter and sanitation at Menik Farm was highly inadequate to cope with the large numbers of people who arrived in April and May. The shelters consisted of tarpaulins, which became very hot under the blazing sun. People had to wait many hours or sometimes an entire day for food and water. Food was of very poor quality and sometimes was served into bare hands, without plates.

161. Families were often grouped into tents with other families, to whom they were not related. In cases of families headed by women whose husbands were missing or dead, such practice made them vulnerable to abuse by unrelated men living in the same tent. The poor conditions provoked violence by IDPs against other IDPs, including sexual violence and exploitation, particularly considering the high number of women without male relatives and unaccompanied children. Women were not given sufficient privacy, and soldiers infringed on their privacy and dignity by watching them while they used the toilet or bathed. Some women were forced to perform sexual acts in exchange for food, shelter or assistance in camps.

162. While basic conditions at Menik Farm were inhumane, a Western Union (money transfer facility) soon opened, and thousands of people, many of them LTTE with connections among the diaspora, were able to buy their way out of the camps by bribing the military. Conditions in Menik Farm did improve over time after much protest from the international community and threats from donors to cut off funding.

(c) Torture in detention
163. The CID and TID maintained units inside the camps in Menik Farm and conducted regular interrogations. Other individuals were also detained and interrogated for potential links to the LTTE, including the doctors, the AGA and two United Nations staff members. Some of them were tortured as well. The sounds of beating and screams could be heard from the interrogation tents. The UNHCR recorded at least nine cases of torture in detention. Some detainees were taken away and not returned.

3. Arbitrary detention of suspected LTTE
164. During the screening process, the SLA removed those suspected of being LTTE members to separate detention facilities at Boossa and Omanthai, generally under the Prevention of Terrorism Act or the Emergency Regulations. In many cases the SLA did not provide family members with notification for the detention of their relatives; neither did it identify the criteria by which it was identifying suspected LTTE. According to Government figures provided to the Panel, as of September 2010, a total of 11,696 persons who “initially surrendered … are accounted for and are being processed”, although this number cannot be independently verified, as the Government has refused to allow independent oversight by the United Nations, ICRC or the Sri Lankan Human Rights Commission. The tally includes people who did not take part in fighting or who were only recruited in the final weeks or days. Among them, according to the Government’s figures, were 594 children. Initially children were housed with the adults, but were registered by UNICEF; later they were moved to separate child rehabilitation centres. However, many of these were in the south of Sri Lanka, which made family visits difficult.

165. Detainees would be questioned in detail about their links with the LTTE. Some would then be transferred to “Protective Accommodation and Rehabilitation Centres” (PARCs), under the authority of the Commissioner-General for Rehabilitation. This office, established in 2006 under Emergency Regulations, exercises power to detain a “surrendee” upon order of the Defence Secretary for up to two years, for the purpose of “rehabilitation”. The Commissioner-General decides the nature of the rehabilitation in individual cases, and the programme does not comply with international frameworks for Disarmament, Demobilization and Reintegration. While it is known to include vocational training (as selected by the Government), official rehabilitation also includes a psychological component where “surrendees” are “reformed”. According to the September 2010 figures provided to the Panel by the Government, “approximately 6,500” alleged ex-combatants were undergoing “short term” rehabilitation, “around 3,500” were undergoing “longer term rehabilitation”, and “less than 1,500” were identified as “hard core” LTTE and designated for prosecution.

166. The Government submitted documents to the Panel which stated that 5,809 “rehabilitees” had been “reintegrated”, that is, released, as of 8 February 2011, with a further 4,581 undergoing “rehabilitation” under the authority of the Commissioner-General for Rehabilitation in nine different PARC detention facilities. This suggests 1,306 alleged LTTE suspects are still retained in closed detention facilities for criminal investigation and prosecution.

167. There is virtually no information about the conditions at these separate LTTE “surrendee” sites, due to a deliberate lack of transparency by the Government. The fact that interrogations and investigations as well as “rehabilitation” activities have been ongoing, without any external scrutiny for almost two years, rendered alleged LTTE cadre highly vulnerable to violations such as rape, torture or disappearances, which could be committed with impunity.

G. Other allegations
168. In addition to the credible allegations discussed above, the Panel has been presented with a number of other allegations, about which it was unable to reach a conclusion regarding their credibility. Due to their potentially serious nature, these allegations should also be investigated.

1. Allegations of the use of cluster munitions or white phosphorus
169. There are allegations that the SLA used cluster bomb munitions or white phosphorus or other chemical substances against civilians, particularly around PTK and in the second NFZ. Accounts refer to large explosions, followed by numerous smaller explosions consistent with the sound of a cluster bomb. Some wounds in the various hospitals are alleged to have been caused by cluster munitions or white phosphorus. The Government of Sri Lanka denies the use of these weapons and, instead, accuses the LTTE of using white phosphorus.

2. The “White Flag” incident
170. Various reports have alleged that the political leadership of the LTTE and their dependants were executed when they surrendered to the SLA. In the very final days of the war, the head of the LTTE political wing, Nadesan, and the head of the Tiger Peace Secretariat, Pulidevan, were in regular communication with various interlocutors to negotiate a surrender. They were reportedly with a group of around 300 civilians. The LTTE political leadership was initially reluctant to agree to an unconditional surrender, but as the SLA closed in on the group in their final hideout, Nadesan and Pulidevan, and possibly Colonel Ramesh, were prepared to surrender unconditionally. This intention was communicated to officials of the United Nations and of the Governments of Norway, the United Kingdom and the United States, as well as to representatives of the ICRC and others. It was also conveyed through intermediaries to Mahinda, Gotabaya and Basil Rajapaksa, former Foreign Secretary Palitha Kohona and senior officers in the SLA.

171. Both President Rajapaksa and Defence Secretary Basil Rajapaksa provided assurances that their surrender would be accepted. These were conveyed by intermediaries to the LTTE leaders, who were advised to raise a white flag and walk slowly towards the army, following a particular route indicated by Basil Rajapaksa. Requests by the LTTE for a third party to be present at the point of surrender were not granted. Around 6.30 a.m. on 18 May 2009, Nadesan and Pulidevan left their hide-out to walk towards the area held by the 58th Division, accompanied by a large group, including their families. Colonel Ramesh followed behind them, with another group. Shortly afterwards, the BBC and other television stations reported that Nadesan and Pulidevan had been shot dead. Subsequently, the Government gave several different accounts of the incident. While there is little information on the circumstances of their death, the Panel believes that the LTTE leadership intended to surrender.

H. The Government’s version of events
172. The credible allegations above reveal a version of the final stages of the war very different from that of the Government of Sri Lanka. The Government says it pursued a “humanitarian rescue operation” with a policy of “zero civilian casualties”. In a speech on 7 May 2010, the Defence Secretary said that it was the first time in the world that a zero casualty principle was included in military operational orders in a battle. On 18 June 2010, President Mahinda Rajapaksa said in a speech that “we left no room for even one bullet to be fired against ordinary citizens”. On 18 May 2010, the first anniversary of the end of the war, the Defence Secretary recalled:


 * We declared ‘no fire zones.’ We also adopted a self-imposed ban on air bombing, artillery and mortar fire whenever we were confronted with battle zones which were home to civilians. Our field commanders were very mindful of this and restrained themselves often. … Also, at every stage of the battle we made certain that food and medical supplies reached trapped civilians through the World Food Programme, the Red Cross and the United Nations.

173. On 2 March 2009, the Minister for Disaster Management and Human Rights, Mahinda Samarasinghe told the BBC’s Hardtalk: “There is absolutely no justification to use heavy weapons, and, in fact, about ten days ago, the Armed Forces took a conscious decision not to use any heavy weapons. As you know, the LTTE is restricted in fact to a very small area of about 48 sq. kilometres, and we cannot use heavy weapons.” However, other Government officials said that up to 81 mm mortars had been used until the end of the conflict but argued that these are not heavy weapons.

174. On 6 April 2009, in Sri Lanka’s Observer newspaper, Lieutenant General Sarath Fonseka said that the SLA was involved in “the world’s largest hostage rescue” operation. On 18 May 2009, Minister for Human Rights Samarasinghe said that warnings over an “imminent bloodbath” had proved wrong. The Government also maintained that IDPs coming out of the conflict zone were received in “welfare centres”, whereas the former LTTE were subject to a “restorative” process, focusing on their “rehabilitation”.

175. After a rigorous review and assessment of all of the available information, the Panel is unable to accept the version of events held by the Government of Sri Lanka.

I. Conclusions
176. The Panel’s account of the allegations associated with the final stages of the war thus reveal five core categories of potential serious violations committed by the Government of Sri Lanka:

(a) Killing of civilians through widespread shelling.
The Sri Lanka Army (SLA) advanced its military campaign in the Vanni, using large-scale and widespread shelling, at times with heavy weapons, such as Multi-Barrel Rocket Launchers (MBRLs) and other large artillery, causing large numbers of civilian casualties. It shelled in three consecutive No Fire Zones, where it had encouraged the civilian population to concentrate, and after it had indicated that it would stop using heavy weapons. It shelled in spite of its knowledge of the impact, provided through SLA intelligence systems, including UAVs, and through notification by various external actors, including the United Nations and the ICRC. The majority of civilian casualties in the final phases of the war were caused by Government shelling. The Government sought to limit external pressure and observation by excluding international organizations from the conflict zone.

(b) Shelling of hospitals and other humanitarian objects.
The Government systematically shelled hospitals on the frontlines, some of them repeatedly. Some civilians who had been injured in shelling and who had come to the hospital were re-injured or killed due to this shelling. All hospitals in the Vanni were hit by shells and had to be evacuated. This was despite the fact that their locations were well-known to the Government.

(c) Denial of humanitarian assistance.
The Government systematically deprived persons in the conflict zone of humanitarian assistance, in the form of food and basic medical supplies, particularly supplies needed to treat injuries. To this end, it purposefully underestimated the number of civilians that remained in the conflict zone. Particularly the denial of surgical supplies greatly increased the suffering of the civilians and added to the large death toll.

(d) Human rights violations suffered by victims and survivors of the conflict.
Despite referring to its actions as a “humanitarian rescue operation”, the Government subjected victims and survivors of the conflict to further deprivation and suffering after they left the conflict zone. Massive overcrowding led to terrible conditions, breaching the basic social and economic rights of detainees, and lives were lost unnecessarily. All IDPs were detained in closed camps and were not allowed to speak privately with humanitarian organizations. Women were subject to further harassment and exploitation in the camps and in detention. Screening for suspected LTTE took place without any transparency or external scrutiny. Some suspected LTTE cadres were executed and others disappeared. Photos and footage of naked female cadre indicate that they may have been raped or sexually assaulted. Torture during interrogation continued. Suspected LTTE were removed to separate camps where they were held for years, outside the scrutiny of the ICRC, the Sri Lankan Human Rights Commission or other agencies.

(e) Human rights violations outside the conflict zone.
The Government sought to intimidate and silence the media and other critics through a variety of threats, including the use of white vans to abduct and make people disappear.

177. The Panel’s account of the allegations associated with the final stages of the war also reveals six core categories of potential serious violations committed by the LTTE:

(a) Using civilians as a human buffer.
Despite the grave dangers and terrible conditions in the conflict zone, the LTTE refused civilians permission to leave, using them as hostages and at times using their presence as a strategic human buffer between themselves and the advancing SLA. Civilians were increasingly sacrificed as dispensable “cannon fodder” while the LTTE fought to protect its senior leadership. The LTTE’s refusal to allow civilians to leave the area added significantly to the total death toll in the conflict.

(b) Killing civilians attempting to flee LTTE control.
From February 2009 onwards, the LTTE instituted a policy of shooting civilians who attempted to escape the conflict zone, significantly adding to the death toll in the final stages of the war. It positioned cadre along points where civilians were trying to escape and shot at groups of men, women and children whom in their desperation were prepared to wade through the lagoon or cross minefields to try to reach Government-controlled areas. Some drowned in the panic as they tried to escape the shooting.

(c) Using military equipment in the proximity of civilians.
The LTTE fired artillery from the NFZs, in proximity to IDP populations, and fired from or stored military equipment near IDPs or civilian installations such as hospitals. They did this even though they knew that it would provoke a response from the SLA and that any retaliating artillery would cause harm to civilians. Sometimes they fired from among civilians before quickly moving away, leaving the civilians on the receiving end of the return fire.

(d) Forced recruitment of children.
The LTTE operated a policy of forced recruitment throughout the war, but in the final stages greatly intensified its recruitment of people of all ages, including children as young as fourteen. It recruited more than one child per family and beat relatives who tried to resist, in a desperate attempt to prevent their children from being carried away from them to an almost certain death. This policy was enforced with great cruelty and regardless of the hopeless military situation of the LTTE.

(e) Forced labour.
The LTTE forced civilians to bolster their defence lines through digging trenches and other emplacements used for its own defences, thereby contributing to blurring the distinction between combatants and civilians. It thereby exposed civilians to additional harm from shelling.

(f) Killing of civilians through suicide attacks.
During the final stages of the war, the LTTE continued its policy of suicide attacks outside the conflict zone. Even though its ability to perpetrate such attacks was diminished compared to previous phases of the conflict, it perpetrated a number of attacks outside the conflict zone, including a suicide bombing at a screening centre in Mullaittivu on 9 February 2009, in which around 30 people died, and a suicide attack killing Minister Mahinda Wijesekera at Akuressa on 10 March 2009, killing around 15 people.