The Impact of Israel’s Use of Cluster Munitions in Lebanon in July and August 2006

During Human Rights Watch’s visits to south Lebanon in August, September, and October 2006, researchers saw dozens of towns hit by cluster munitions and hundreds of submunition duds littering backyards and fields.87 The teams also witnessed UN, nongovernmental, and Lebanese Army deminers struggling to cope with a problem of unprecedented magnitude. Israel had hit only the peripheries of some towns with cluster munitions but had elsewhere blanketed built-up areas. As civilians returned home immediately after the ceasefire, they found their property had become de facto minefields. Villagers and deminers discovered unexploded cluster duds inside houses, in the streets, in gardens, on roofs, on patios, and hanging from trees and fences. In Tebnine, a hospital had been struck, and the hundreds of duds strewn across the entryway trapped doctors and patients inside (see Case Study below). MACC SL reported, on January 15, 2008, 192 civilian casualties, including 20 killed and 172 wounded.88 Exploding duds were still injuring civilians in the south in December 2007.

A senior Human Rights Watch military analyst who arrived in south Lebanon immediately after the ceasefire had surveyed cluster munitions on the ground in both Kosovo and Iraq. The sheer number and density of dud fields in urban areas dwarfed anything he had ever seen before.

The IDF’s cluster munition strikes were spread over an area of approximately 1,400 square kilometers north and south of the Litani river, an area comparable in size to the US state of Rhode Island (1,214 sq km). Of the 1,400 square kilometers affected by the cluster munitions, an aggregate area of 38.7 square kilometers, including 4.3 square kilometers of urban areas, 20 square kilometers of agricultural land, and 4 square kilometers of woodland, has been confirmed by deminers as directly contaminated by submunitions.89 However, the lives of civilians in the entire 1,400 square kilometer area have been severely affected, as they cannot live in safety until demining crews clear and inspect their homes and fields.

Shocking Scope: Number of Submunitions and Strikes

In the first two days after the ceasefire, UN deminers beginning emergency survey and clearance work in south Lebanon identified 10 locations where Israel used cluster munitions. A UN official said he feared it could be only the “tip of the iceberg.”90 By January 2008, the number of strike sites identified was 962,and continued to grow as clearance professionals pushed into new corners of south Lebanon.91

MACC SL has estimated that Israel used cluster munitions (artillery shells, ground rockets, and air-dropped bombs) containing between 2.6 and four million submunitions in Lebanon.92 It arrived at that estimate in the following fashion. First, it calculated that Israel fired some 16,000 to 32,000 artillery cluster shells containing a total of 1.4 to 2.8 million submunitions.93 To those figures, it added 1,800 MLRS rockets carrying 1,159,200 M77 submunitions, which Israeli soldiers reported to Ha’aretz newspaper.94 It also noted that Israel dropped an unknown number of aerially delivered CBU-58B cluster bombs, each containing 650 BLU-63 bomblets.95 Given the high failure rates of these different types of submunitions, the UN has estimated that the cluster barrages left behind hundreds of thousands, possibly up to one million, hazardous duds.96

Outside of the UN estimate, Israeli soldiers told Human Rights Watch that the 1,800 MLRS rockets accounted for only those fired by a reserve MLRS battalion, and that an active duty battalion fired 1,000 more, which would contain 644,000 submunitions, bringing the number of rocket submunitions to more than 1.8 million.97 This additional information could raise the estimated total of Israeli submunitions fired into Lebanon to some 3.2 to 4.6 million submunitions.

Israel’s use of cluster munitions was the most extensive use of the weapon anywhere in the world since the 1991 Gulf War and was concentrated in a relatively small geographical area. The number and density of cluster munitions used in Lebanon vastly exceeded their use in prior wars in Kosovo, Afghanistan, and Iraq where Human Rights Watch also conducted investigations. NATO air forces used 1,765 cluster bombs with about 295,000 bomblets in Kosovo in 1999, the US Air Force used 1,228 cluster bombs with about 248,000 bomblets in Afghanistan in 2001 to 2002, and Coalition forces used about 13,000 cluster munitions with about 1.9 million submunitions in Iraq in 2003.98

“I’ve seen every single cluster use since 1991 and this is more than I have ever seen,” Chris Clark, program manager of MACC SL, told Human Rights Watch. “A similar amount of ordnance was thrown in Iraq, but south Lebanon is much smaller.”99 Israel’s use of cluster munitions in Lebanon compares most closely to Coalition use in Iraq in 2003 because in both cases most of the attacks were ground-launched and included counter-battery fire. Still, the use of so many cluster munitions in such a small area is shocking and unprecedented.

Timing and Targets: When and How Cluster Munitions Were Used

The Early Phases of the War

Israel launched sporadic cluster munition attacks on south Lebanon in the first two weeks of the war. Human Rights Watch first confirmed Israeli use of cluster munitions when it reported on a July 19 attack on Blida that left one civilian dead and at least 12 wounded.100 Meanwhile, Human Rights Watch researchers observed large numbers of artillery-fired cluster munitions in the arsenals of the IDF artillery teams deployed in Israel’s border with Lebanon.

Attacks increased in the days after the 48-hour partial cessation of air strikes of July 31 to August 1. Israeli soldiers serving with an MLRS unit told Human Rights Watch that it was in August when they fired many of their cluster rockets.101 Through field visits and other sources, Human Rights Watch identified strikes that had taken place before the last three days of the war in about 10 towns other than Blida: `Ainata, `Aitaroun, Deir Qanoun, Hasbayya, Hebbariyeh, Kfar Dounine, Kfar Hamam, Rashaya al Foukhar, Sawane, and Tair Debbe.102

Human Rights Watch’s investigations and interviews indicate that Israel aimed some of its cluster strikes prior to the last days at Hezbollah rocket launch sites, largely in olive groves and tobacco fields. Some villagers told Human Rights Watch researchers that Hezbollah fighters used such fields to fire rockets into Israel.103 Others who suffered cluster attacks, such as those in Blida, said there was no Hezbollah military activity nearby.

Israeli soldiers told Human Rights Watch that their radar would locate Hezbollah launch sites while the rocket was airborne, and the IDF would then fire cluster munitions in the vicinity of the launch area, using the area-effect weapons in an attempt to kill the launch crew and destroy its launchers as they tried to escape. Shooting back—typically with a “six-pack” of US M26 rockets—at Hezbollah rocket launch sites generally occurred within one to 1.5 minutes of receipt of the launch detection coordinates.104

Civilian casualties from cluster munitions at the time of these strikes seem to have been fairly limited, reflecting the fact that so much of the population had vacated south Lebanon or hid in their basements, and that much of the Hezbollah rocket fire and Israeli counter-battery fire occurred in fields and valleys where civilians were not present at the time. However, the exact number of injuries and deaths from these cluster strikes may never be known, as hospital staff were too overwhelmed at the time to ask questions about the specific causes of injury or death.

The Final Barrage

Over the final days of the conflict, the Israeli use pattern changed dramatically. According to the UN, Israel fired 90 percent of its cluster munitions during the last 72 hours, after the UN Security Council had passed Resolution 1701 calling for a ceasefire on August 11, but before the ceasefire took effect at 8 a.m. on August 14.105 During this period, there was also an intensification of bombardment by other weapons, including artillery strikes as well as the aerial strikes on civilian homes with 500-pound bombs. The increase coincided with an increase in Hezbollah rocket strikes on Israel. According to Israeli government statistics and news accounts, Hezbollah increased its rocket attacks in the final days, and on the last day of the war launched 252 rockets, the highest daily toll of the conflict.106 However, even given an increase in Hezbollah attacks, the use of more than four million submunitions to strike at hundreds of rocket launches posed a high likelihood of indiscriminate and disproportionate attacks, particularly when so many of the submunitions hit built-up areas, predictably leaving behind thousands of duds.  

Witness testimonies from south Lebanon’s villages also indicate that there was a massive increase in cluster munition attacks in the last few days before the ceasefire. The head of Tair Debbe municipality, `Ali Moughnieh, said that in the last several days of the war, “it started raining cluster bombs.”107 Hassan `Abass Hattab, the mukhtar (a local official with administrative responsibilities) of Habboush, similarly said that Israel launched cluster munitions on his village during “the last four days of the war.”108 Several others, including the mukhtars of both Tebnine and Kfar Rommane, echoed these statements.109

Soldier testimony further attests to the IDF’s heavy use of cluster munitions in the final hours of the war. “In the last 72 hours we fired all the munitions we had, all at the same spot, we didn’t even alter the direction of the gun,” an IDF soldier said. “Friends of mine in the battalion told me they also fired everything in the last three days—ordinary shells, clusters, whatever they had.”110 UN Interim Force in Lebanon (UNIFIL) fire mission data supports these assertions.111

The use of cluster munitions in the last 72 hours elicited outrage from UN officials. The UN’s humanitarian coordinator in Lebanon, David Shearer, said, “The outrageous fact is that nearly all of these [cluster] munitions were fired in the last three to four days of the war…. Outrageous because by that stage the conflict had been largely resolved in the form of [UN Security Council] Resolution 1701.” He said it “defied belief” that Israel had used so many cluster munitions in the last hours of the war.112 The UN’s then emergency relief coordinator and under-secretary-general for humanitarian affairs,Jan Egeland, called Israel’s use of cluster munitions “completely immoral.”113 

IDF lawyers told Human Rights Watch that the ceasefire negotiations did not change operational decisions over the last three days of the war because the IDF considered itself still in combat. Maj. Dorit Tuval, head of the strategic section in the IDF’s International Law Department, said, “As a lawyer, it was not important. It was a legitimate decision to be taken by commanders. As far as we know, the use was legal.”114

Attacks on Population Centers

Many cluster munitions struck population centers. According to a land use study commissioned by the UN Development Program (UNDP), cluster munitions contaminated about 4.3 million square kilometers of urban areas during the conflict.115 A senior UN demining official said he had “no doubt” that Israel had deliberately hit built-up areas with cluster munitions, stating, “These cluster bombs were dropped in the middle of villages.”116 The program manager of MACC SL told Human Rights Watch that “the vast majority of clusters were used in towns.”117

Human Rights Watch field research corroborated the widespread use of cluster munitions in population centers. In the first week after the ceasefire, Human Rights Watch visited about 30 villages and towns that the IDF attacked with cluster munitions and visited more than a dozen more in October. Cluster munitions landed in large villages such as Tebnine and Nabatiyah. Towns that were especially hard hit include: Ain B’al, Bar’achit, Bint Jbeil, Majdel Selm, Kfar Tebnit, Sawane, Srifa, Tebnine, Yohmor, Zawtar al-Gharbiyeh, and Zawtar al-Sharkiyeh.

The IDF has since acknowledged that it targeted built-up areas with cluster munitions. A statement released by the IDF Spokesman’s Office in November 2006 said that “the use of cluster munitions against built-up areas was done only against military targets where rocket launches against Israel were identified and after taking steps to warn the civilian population.”118 In July 2007, IDF lawyers reiterated this position in a meeting with Human Rights Watch. “In cases where there was a need to direct cluster munitions toward the vicinity of a built-up area, they were always directed toward places where rockets were shot from toward Israel. It was always after messages to leave the area and then we made sure distinction and proportionality were applied,” Major Tuval said.119 She added, “Even if they were used in the vicinity of built-up areas, it was much less than necessary. Operational considerations were hurt because of our efforts.”120

The IDF statement in December 2007 reporting the results of the second internal inquiry echoed these statements. It said that investigating officer Maj. Gen. Gershon HaCohen found that “cluster munitions were fired by the IDF on built-up areas only in direct response to Hizbullah’s firing of rockets from within those areas…. Furthermore, the munitions were fired on villages only when the forces understood them to have been almost completely evacuated, hence the anticipated harm to civilians was small.”121

However, soldiers have offered eyewitness accounts with a very different description of the targets. A commander of an IDF MLRS unit told a Ha’aretz reporter, “What we did was insane and monstrous, we covered entire towns in cluster bombs.” He said that to compensate for the cluster rockets’ imprecision, his unit was ordered to “flood” the area with cluster munitions. In one case, his unit was ordered to fire cluster rockets toward “a village’s outskirts” in the early morning because “people are coming out of the mosques and the rockets would deter them.”122

Failure Rates

As described earlier, the presence of duds is an inevitable result of the use of cluster munitions. “It’s a bad weapons system. So many things need to happen to deploy and arm properly,” the program manager for MACC SL said.123 A BACTEC deminer added, “A lot of things can go wrong.”124 Given the vast number of submunitions used in Lebanon, the only result could be a huge number of duds.

Mine clearance personnel in Lebanon report the failure rates for Israeli submunitions to be exceptionally high, with a large number of duds compared to impact sites. The program manager of MACC SL has projected an average failure rate of 25 percent, with up to 70 percent in some locations.125 In some strikes, especially with BLU-63 submunitions, deminers have found dud rates of 90 to 100 percent.126 The dud rates in the field of Israel’s submunitions have been substantially higher than published test data and also substantially higher than those found in previous conflicts, such as Iraq and Kosovo.

Israel has not provided any reasons for the exceptionally high dud rates in Lebanon. It questioned MACC SL’s estimates, claiming instead that the dud rate was under 10 percent, and dismissed the problem as being quickly dealt with by clearance.127 The rates have been documented by deminers, however, and may be the result of the extensive use of older cluster munitions, especially dated US weapons. Explosive materials deteriorate over time, making the weapons increasingly unstable and more likely to fail.

Another possible factor is low trajectory or short-range firing. A report in Ha’aretz said that in some cases the IDF fired M26 MLRS rockets “at a range of less than 15 kilometers, even though the manufacturer’s guidelines state that firing at this range considerably increases the number of duds.”128 An IDF reservist told Human Rights Watch that he thought the reserve MLRS unit shot 20 to 25 percent of the M26 rockets at minimum ranges of 13 to 15 kilometers.129 According to the US Army, the M26 rocket’s M77 “submunition dud rate increases significantly at ranges less than 10 km.”130 The high dud rate may also be partially attributable to landscape characteristics, such as the soft ground of agricultural fields, and the density of trees and vegetation, which may catch cluster submunitions as they fall.131

The large number of Israeli-produced M85 submunition duds is particularly striking since one model of that submunition, used extensively during the conflict, has a self-destruct feature that reportedly reduces the failure rate to some 1.3 to 2.3 percent under testing conditions. As mentioned earlier, many experts have pointed to it as one of the most reliable submunitions in the world.132 However, a report in December 2007 by three organizations that carefully studied the performance of the M85 with self-destruct devices in Lebanon, including the primary institution responsible for defense-related research in Norway, estimated the failure rate to be about 10 percent.133 The study said that the “inescapable conclusion from Israel’s use of M85 bomblets…is that they failed far more often than would have been predicted based on the claims of stockpiling states and manufacturers.”134

Earlier, the program manager of MACC SL had reached a similar conclusion. In April 2007, he stated, “Whilst several military users maintain that the M85 with self-destruct mechanism has a failure rate of less than 1%, the evidence on the ground in South Lebanon clearly shows that this weapon has a reality failure rate of between 5 and 10%. It is common to find at least 3 unexploded submunition grenades from individual carrier shells (M396/49 per shell) equating to a 6% failure rate.”135

The IDF also used a version of the M85 without the self-destruct device, though the ratio of self-destruct to non-self-destruct is unknown at this time. As of January 18, 2008, MACC SL reported that it had cleared 6,892 M85s of all types.136

A UNIFIL deminer holds an Israeli-made M85 submunition that he has rendered safe in Beit Yahoun on October 24, 2006. Some such submunitions had self-destruct devices, but deminers and weapons experts have documented dud rates in the field of 10 percent even for these models. © 2006 Bonnie Docherty/Human Rights Watch

Many of those cleared were the self-destructing types, but the precise number is not known. According to MACC SL the M85 without the self-destruct mechanism is “commonly found with a 15% failure rate on the ground.”137 

In recent years, the United States and several other countries have identified a one percent failure rate as the desirable standard for submunition procurement, but the performance of the M85 in Lebanon calls into question the feasibility and effectiveness of this potential future standard, since even a very low failure rate in test conditions gives way to a much higher failure rate in the conditions of actual combat. 

Israeli soldiers were well aware of the large numbers of duds their cluster strikes were producing. A soldier said that his MLRS commander gave a “pep talk” after a period of heavy fire, saying, “Just wait until Hezbollah finds the little presents we left them.”138Soldiers also told Human Rights Watch that IDF soldiers were taught throughout their training to ignore the manufacturer’s claim of a 5 percent submunition failure (dud) rate for the M77 submunitions contained in the M26 rocket, and to presume a 15 percent submunition failure rate instead.139 A reserve officer told a reporter that there is an IDF regulation prohibiting the firing of cluster munitions on areas the IDF is planning to enter to avoid exposing IDF soldiers to risks of death or injury by duds.140

87 Human Rights Watch has separately reported on violations of international humanitarian law by both Israel and Hezbollah during the 2006 conflict. See Human Rights Watch, Why They Died, and Human Rights Watch, Civilians under Assault.

88 MACC SL Casualty List. The Landmines Resource Center reported, on January 2, 2008, 220 civilian injuries and 19 deaths from cluster munition duds. LMRC Casualty List.

89 Email communication from Dalya Farran, media and post clearance officer, MACC SL, to Human Rights Watch, January 15, 2008; UNDP, “CBU Contamination by Land Use,” current as of November 29, 2006.

90 See “Lebanon: Israeli Cluster Munitions Threaten Civilians,” Human Rights Watch news release, August 17, 2006,

91 Email communication from Dalya Farran, media and post clearance officer, MACC SL, to Human Rights Watch, January 15, 2008.

92 MACC SL, “South Lebanon Cluster Bomb Info Sheet as at November 4, 2006”; Chris Clark, program manager, MACC SL, presentation at UN Mine Action Service briefing on Lebanon, Seventh Meeting of States Parties to the Mine Ban Treaty, Geneva, September 19, 2006 (notes by Human Rights Watch).

93 MACC SL, “South Lebanon Cluster Bomb Info Sheet as at November 4, 2006.” MACC SL took a press report of 160,000 artillery shells and assumed that 10 to 20 percent of them were cluster munitions containing 88 submunitions.

94 Ibid.; Meron Rapoport, “IDF Commander: We Fired More than a Million Cluster Bombs in Lebanon,” Ha’aretz, September 12, 2006.

95 One source cited an estimate of 500,000 BLU-63 bomblets from CBU-58 cluster bombs. James Brooks, “How Israel Cluster Bombed Future of South Lebanon, with US help,”, (accessed October 18, 2006). As of January 18, 2008, MACC SL reported that 28,136 BLU-63 duds had been cleared. Email communication from Dalya Farran, media and post clearance officer, MACC SL, to Human Rights Watch, January 18, 2008.

96 Email communication from Dalya Farran, media and post clearance officer, MACC SL, to Human Rights Watch, January 15, 2008.

97 Human Rights Watch interviews with IDF reservists (names withheld), Tel Aviv and Jerusalem, Israel, October 2006. The reservists were experienced non-commissioned officers leading platoons with an MLRS unit. They also commanded resupply missions from the active unit to the reserve unit.

98 See Human Rights Watch, Cluster Bombs in Afghanistan, October 2001,; Human Rights Watch, Fatally Flawed, p. 15; Human Rights Watch, Off Target, p. 6.

99 Interview with Chris Clark, program manager, MACC SL, Tyre, September 14, 2006.

100 “Israeli Cluster Munitions Hit Civilians in Lebanon,” Human Rights Watch news release, July 24, 2006,

101 Human Rights Watch interviews with IDF reservists (names withheld), Tel Aviv and Jerusalem, Israel, October 2006. A launcher commander in the reserve MLRS unit said that this unit “did nothing” during the first week of conflict, and “only shot sporadically” during the second week. He said the largest volume of fire for the reserve MLRS unit was during the third week, immediately after the 48-hour ceasefire when civilians were told to leave. He said “we fired tons” during this time and noted that one launcher under his command shot 60 pods (360 rockets, 231,840 M77 submunitions) in a 24-hour period during this time. The volume of fire from his unit was reduced for the remainder of the conflict due to ammunition shortages.

102 Lebanese security forces, UN sources, and medical personnel also identified these sites. Dr. Nasser al-Din Kassir, a surgeon at Hiram Hospital, told Human Rights Watch that during the war the hospital received at least four patients from Deir Qanoun al-Nahr with cluster injuries. Human Rights Watch interview with Dr. Nasser al-Din Kassir, Hiram Hospital, Tyre, August 30, 2006. See also Rym Ghazal and Leila Hatoum, “Investigators Probe Possible Use of Banned Weapons,” Daily Star, July 26, 2006. 

103 Human Rights Watch interview with Shawki Yousif, head of Hebbariyeh municipality, Hebbariyeh, October 22, 2006; Human Rights Watch interview with Khalil Muhammad Hussein, farmer, Kfar Rommane, August 16, 2006; Human Rights Watch interview with Salih Ramez Karashet, farmer, Hammoud Hospital, Saida, September 22, 2006; Human Rights Watch interview with Hassan Muhammad Nasser, construction worker, `Ein Ba`al, September 22, 2006; Human Rights Watch interview with Hussein `Ali Kiki, construction worker, `Ein Ba`al, September 22, 2006.

104 Human Rights Watch interviews with IDF reservists (names withheld), Tel Aviv and Jerusalem, Israel, October 2006. Human Rights Watch was told that for MLRS units Lebanon was divided into two sectors. The first was the tactical battle against Hezbollah forces south of the Litani River under the direction of the IDF’s Northern Command. The second sector was the strategic effort targeting locations north of the Litani River controlled by the operations division of the General Staff Headquarters. Both active-duty and reserve MLRS battalions fired at targets in both sectors. Most MLRS attacks occurred prior to the final 72 hours of fighting.

105 The UN has explained that “UNMACC’s calculations are based both on-the-ground identification of cluster bomb strike locations and extensive conversations with South Lebanon residents.” UNOCHA, “A Lasting Legacy: The Deadly Impact of Cluster Bombs in South Lebanon,” undated, but information as of September 16, 2006, p. 1, fn 3. Chris Clark, program manager of MACC SL, told Human Rights Watch he reached this conclusion based largely on his own firsthand observations of Israeli attacks throughout the war. He also noted the small number of reports of cluster munition attacks and casualties prior to the final days of the war. Human Rights Watch interview with Chris Clark, program manager, Tyre, October 25, 2006.

106 On the last three days of the conflict, Hezbollah launched 115, 70 and 252 rockets respectively. Israeli Police North District, Central Command Center, “War in the North,” powerpoint presentation obtained by Human Rights Watch in October 2006. Israel has not presented information indicating that cluster munitions caused any significant damage to Hezbollah personnel or weaponry.

107 Human Rights Watch interview with `Ali Moughnieh, head of Tair Debbe municipality, Tair Debbe, October 21, 2006.

108 Human Rights Watch interview with Hassan `Abbas Hattab, mukhtar, Habboush, October 25, 2006.

109 Human Rights Watch interview with Yousif Fawwaz, mukhtar, Tebnine, October 24, 2006.

110 Meron Rapoport, “What Lies Beneath,” Ha’aretz, September 8, 2006.

111 Landmine Action, “Foreseeable Harm: The Use and Impact of Cluster Munitions in Lebanon, 2006,” October 2006, (accessed September 3, 2007), p. 11. This report has a chart titled “Fire Missions Observed by UN Observers in UNIFIL Areas of Operation, 16 July-13 August 2006,” based on data provided by UNIFIL. It notes, “Whilst an average of 2,000 fire missions were recorded each day during the conflict, this increased to approximately 6,000 per day in the last three days before the ceasefire.”

112 “UN Calls Israel’s Use of Cluster Bombs in Lebanon ‘Outrageous,’” Ha’aretz, September 19, 2006.

113 “UN Slams Israel as Unexploded Cluster Bombs Discovered,” Irish Examiner, August 31, 2006.

114 Human Rights Watch interview with Maj. Dorit Tuval, head of the strategic section, International Law Department, IDF, Tel Aviv, Israel, July 2, 2007.

115 UNDP, “CBU Contamination by Land Use,” current as of November 29, 2006.

116 Alistair Lyon, “Israel Cluster-Bombed 170 Sites in Lebanon—UN,” Reuters, August 22, 2006 (quoting Tekimiti Gilbert, operations chief for MACC SL).

117 Human Rights Watch interview with Chris Clark, program manager, MACC SL, Tyre, September 14, 2006.The NGO Landmine Action analyzed maps provided by MACC SL and concluded that 60 percent of cluster strikes hit built-up areas, and that there were cluster munition strikes in or near 90 towns and villages. This was based on data as of September 5, 2006, and indicated where the center of the strike hit less than 500 meters from a built up area. Landmine Action, “Foreseeable Harm: The Use and Impact of Cluster Munitions in Lebanon: 2006,” pp. 13-15.

118 Nir Hasson and Meron Rapoport, “IDF Admits Targeting Civilian Areas with Cluster Bombs,” Ha’aretz,November 21, 2006.

119 Human Rights Watch interview with Maj. Dorit Tuval, head of the strategic section, International Law Department, IDF, Tel Aviv, Israel, July 2, 2007.

120 Ibid.

121 Israel Ministry of Foreign Affairs, “Opinion of the Military Advocate General Regarding Use of Cluster Munitions in Second Lebanon War.”

122 Rapoport, “When Rockets and Phosphorous Cluster,” Ha’aretz.

123 Human Rights Watch interview with Chris Clark, program manager, MACC SL, Tyre, October 21, 2006.

124 Human Rights Watch interview with Johan den Haan, BACTEC, Tyre, October 25, 2006.

125 Chris Clark, program manager MACC SL, presentation to CCW Delegates, Geneva, August 30, 2006 (notes by Human Rights Watch).

126 Human Rights Watch interview with Andy Gleeson, program manager and technical operations manager, Mines Advisory Group, Kfar Joz, October 25, 2006.

127 Major Tuval said, “There is a certain rate of duds, but one could deal with it by clearing…. Even if we can’t release [the dud rate], we’re not talking about 25 percent. It’s less than 10 percent as far as I know.” Human Rights Watch interview with Maj. Dorit Tuval, Tel Aviv, Israel, July 2, 2007.

128 Rapoport, “When Rockets and Phosphorus Cluster,” Ha’aretz.

129 Human Rights Watch interview with IDF reservist (name withheld), Tel Aviv, Israel, October 2006. He also estimated half of the targets fired at by the reserve unit were close to the maximum range of 38 to 40 kilometers and likely north of the Litani River. He said one time they were shooting at such a flat trajectory, almost horizontal, that they accidentally drilled their rockets into a mountain in Israel.

130 US Department of the Army, “Tactics, Techniques and Procedures for Multiple Launch Rocket System (MLRS) Operations,” FM 6-60, April 23, 1996, (accessed September 3, 2007), chap. 1.

131 The US Marine Corps has stated that DPICMs “should not be fired into wooded areas. Submunitions may become suspended in tree branches and later pose a threat to friendly forces. Firing DPICMs into mountainous areas where the slope is greater than 60 percent increases the dud rate.” US Marine Corps, “Fire Support Coordination in the Ground Control Element,” MCWP 3-16, November 2000, pp. 5-38 and 5-39.

132 The submunition (or variations of it) is found in the arsenals of many countries, including Argentina, Austria, Denmark, Finland, Germany, Greece, India, Italy, Norway, Romania, Switzerland, the United Kingdom, and the United States.

133 C. King Associates, Ltd., Norwegian Defence Research Establishment, and Norwegian People’s Aid, M85: An Analysis of Reliability, pp. 14-22.

134 Ibid., p. 6. The study said that in the three strike sites with the most conclusive information, the failure rates were 9.8, 11.5, and 12.2 percent. The study also concluded: “The specific example of the M85 demonstrates that while SD [self-destruct] mechanisms in general may help to lower failure rates, they are not capable of ensuring against post-conflict contamination at an unacceptable level. The specific example of M85 also illustrates the substantial differences between results obtained during testing and reality seen during operations. This suggests that current testing practices may have little or no utility as a predictor of the risk that will be created to the post-conflict civilian population.” The report also “strongly rejects the distinction between ‘hazardous’ and ‘non-hazardous’ duds as conceptually flawed, misleading and dangerous.” Ibid., p. 5.

135 Chris Clark, program manager, MACC SL, “Unexploded Cluster Bombs and Submunitions in South Lebanon.”

136 Email communication from Dalya Farran, media and post clearance officer, MACC SL, to Human Rights Watch, January 18, 2008.

137 Chris Clark, program manager, MACC SL, “Unexploded Cluster Bombs and Submunitions in South Lebanon.”

138 Human Rights Watch interview with IDF reservist (name withheld), Tel Aviv, Israel, October 2006.

139 Human Rights Watch interviews with IDF reservists (names withheld), Tel Aviv and Jerusalem, Israel, October 2006. This dud rate is consistent with US testing data, which reports a 16 percent submunition failure rate. See Office of the Under Secretary of Defense for Acquisition, Technology, and Logistics, “Unexploded Ordnance Report,” table 2-3, p. 5. The US General Accounting Office reported some lots of M26 in US stockpiles to have dud rates as high as 23 percent, based on testing done to accept newly produced batches. See US General Accounting Office, “GAO/NSIAD-92-212: Operation Desert Storm: Casualties Caused by Improper Handling of Unexploded US Submunitions,” August 1993, pp. 5-6.

140 Rapoport, “A Barrage of Accusations,” Ha’aretz.