Analysis: An Examination of Terror Attacks from an Engineering Perspective – MSI Turkish Defence Review / MSI TDR Reference Magazine of Turkish Defence and Aviation Industry

Analysis: An Examination of Terror Attacks from an Engineering Perspective

21 Ekim 2016
2016_09_07_msi_s

From the Suicide Bombing at the Intersection Near Ankara Railway Station on October 10, 2015 to the Coup Attempt on July 15, 2016

Fikret KURAN * / f.kuran@vgm.gov.tr

*Civil Engineer M.Sc.

Süleyman POLAT * / suleymanpolat2009@gmail.com

*Chemical Engineer, Ammunition and Explosive Expert, Judicial Referee

Summary

In recent times, three suicide bombings have been carried out in Ankara The first of these attacks was carried out by two suicide bombers from DAESH, on October 10, 2015 in front of the Railway Station, where 102 people were killed and more than 100 people were injured. The second attack was carried out by the terror organisations PKK (the outlawed Kurdistan Workers Party) and TAK (Kurdistan Freedom Falcons) on February 17, 2016, with a bomb-laden vehicle that was detonated near an army bus carrying military personnel. In the attack, 29 people were killed and 60 people were injured. Using the same method, members of the same terror organisations carried out the third attack on March 13, 2016. This time, civilians waiting at the bus stop near Güvenpark, Kızılay were targeted. In this attack, 36 people were killed and more than 100 people were injured. Finally, a coup attempt was carried out by agents of the Gülenist Terror Organisation (FETÖ) on the night of July 15, 2016, and during the incidents, a total of 246 people were killed, including civilians, soldiers and police officers, while over 1,500 people were injured. This study analyses the terror attacks carried out in Ankara, and shows that it is possible, within minutes of the terrorist bombings taking place, to identify the exact timing of the attack at a sensitivity level of a split second, as well as the number of explosions, the type of explosive used and its TNT equivalent amount. The timing of the bombings conducted by F-16 fighter jets in various locations during the coup attempt of July 15 were calculated chronologically to the split second, and the sonic booms and bomb attacks were distinguished from one another. It is considered that the methods described in this article, which allow information to be gathered in a very short period of time, can be used in emergency management as well as in judicial and administrative investigations.

1.Introduction

Man-made explosions result in earthquake-like vibrations. These tremor-like jolts are called artificial earthquakes. During the Cold War period, nuclear tests were monitored through seismograph records, allowing the other countries to identify the test locations and the size of nuclear tests. In addition to monitoring nuclear tests, seismic records have also been used in many extraordinary incidents of great social impact, such as terrorist bomb attacks, mine accidents, aircraft crashes, explosions in fuel pipelines, and fuel tank explosions. Due to the increase in large industrial accidents and large-scale terror attacks, particularly in recent years, studies aiming to identify the size of explosions, as well as the amount of explosives detonated, have gained importance. In the last decade, comprehensive studies have been conducted on this issue, including experimental trials by the USA and Israel. Theoretical information on analysing explosions by making use of seismic and acoustic records is available in the references Kuran (2014-a) and Kuran & Polat (2015).

In general, TNT and improvised explosive devices (IED) are used in the bomb attacks carried out in Turkey. The composition of IEDs is generally a mixture of fuel oil and ammonium nitrate (AN) or fertilisers containing ammonium nitrate. Low amounts of explosives with high detonation velocities, such as TNT, RDX or C4 are used in order to explode IEDs with ammonium nitrate content. The detonation velocity is reduced in cases where the ratio of fuel oil in the mixture is lower or higher than a certain value. No empirical studies have been conducted in our country on the identification of the detonation velocities of IEDs. Conducting empirical studies to identify the technical features of these kinds of explosives is considered to be important. The detonation velocities of pure aluminium nitrate and fertilisers during the explosion that occurred at the fertiliser factory of ABS Consulting company in Texas, USA, in 2013 are described as being within the ranges shown in Table 1. As Salzano et al. (2014) states that the detonation velocities of Triacetone Triperoxide Peroxyacetone (TATP) type IEDs is 1,400 m/s, we believe that the detonation velocity of the IEDs (with AN content) used in the bomb attacks in Ankara can be assumed to be 1,350 m/s. This value is certainly likely to vary in each IED, depending on the knowledge and skills of the person designing it.

Table 1. Detonation velocities of pure AN and fertilisers containing AN (Source: ABS Consulting Report)

Type of Explosive Winding Material Mixture Material Booster Velocity of Detonation (VoD) m/s
Pure AN Steel N/A 100 gr PA 1,230
Pure AN Steel N/A 50 gr Tetryl 1,310
Pure AN Steel N/A 60 gr Tetryl 1,470
Pure AN Steel N/A 250 gr Tetryl 2,700
AN Powder Steel Carbon C4 and RDX 1,250
Fertiliser Steel Wax-clay 50 gr Comp. A3 1,370
Fertiliser Paper Unknown 225 gr Comp. C 1,100
Fertiliser Steel Unknown 50 gr Comp. A 1,350

 2.Seismic and Acoustic Records of the Terrorist Bomb Attacks Carried Out in Ankara

Since the amount of explosives used in the suicide bombing conducted by the DAESH terror organisation on October 10, 2015 was very limited, there are no seismic or acoustic records of that incident. This attack was hence analysed by using the sound recordings obtained from the video images of the moment of the explosion. On the other hand, seismic and acoustic records are available for the bombings carried out by the PKK and TAK terror organisations with bomb-laden cars on February 17, 2016 and March 13, 2016. The closest seismograph stations to both attack locations are:

  • The Kandilli Observatory and Earthquake Research Institute’s seismograph station, with the code of KO-LOD located in Ankara-Lodumlu,
  • The METU (Middle East Technical University) station of the Disaster and Emergency Management Authority (AFAD), and
  • The United States Geological Survey’s (USGS) seismograph station, codename ANTO that is operated by AFAD on METU’s land in Yalıncak.

The digital recordings of these stations were taken from the websites of European Integrated Data Archive (EIDA) and AFAD, and analysed accordingly (www.orfeus-eu.org/eida; www.deprem.gov.tr).

2.1. Suicide Bomb Attack at the Intersection Near the Railway Station on October 10, 2015

At 10.04 a.m. on October 10, 2015, twin suicide bombings were conducted by the DAESH terror organisation during the peace rally attended by various civil society organisations at the intersection near Ankara Railway Station. The attacks were carried out consecutively by two suicide bombers who were DAESH militants, and 107 people in addition to the bombers were killed in the explosion. Official statements reported that four kilos of TNT strengthened with steel balls was used in the first blast, while five kilos of TNT was used in the second blast.

The first blast is seen in the eight second, 389 millisecond long video image of the attack. The spectrogram graphs of the sound record that were generated from this video image in time-frequency medium allows us to get the sound record of the second blast clearly, just before the end of the record where there are no images. It can be seen that the second blast occurred 1.987 seconds – in other words 13 milliseconds less than two seconds – after the first one. In the first blast, the time elapsed until the sound reached the recorder was longer than the time elapsed in the second blast. Thus, the time difference between the two blasts was calculated as 13 milliseconds less than two seconds. Due to the time difference of 13 milliseconds, it was calculated that the person who recorded the video was four or five metres closer to the second explosion point compared with the first one.

In 2013, a terror attack was carried out in Boston, USA, in the form of two blasts that occurred with an interval of 13 seconds. Gunpowder type explosives were used to fill a pressure cooker and the pressure cooker was put among the audience. Although this terror incident was not a suicide bombing, it also involved two sequential blasts in a manner similar to the Railway Station attack in Ankara. It was also stated in the FBI’s reports that the blasts occurred with an interval of 13 seconds. It can be seen in the video recordings on the Internet that the second blast occurred 12.967 seconds after the first one. The reason for the interval between the blasts, which was 33 milliseconds less than 13 seconds, is believed to be due to the fact that the person who recorded the video was, just as in the Ankara attack, closer to the second point of explosion by about 13 metres While the two second interval between the two blasts in the Railway Station attack appears to be the product of coincidence, the possibility that a timer and a remote control device might have been used, as was the case in the Boston attack, should not be ruled out.

 msi-sekil1                                   Figure 1. Sound recordings of the Ankara Attack dated October 10, 2015:

a) Sound and Spectrogram Graphs for Both Blasts

b) Spectrogram Graph of the First Blast with a 10 Times Extended Time Scale

c) Spectrogram Graph of the Second Blast with a 10 Times Extended Time Scale

The sound recordings of both blasts in the Railway Station attack were slowed down 10 times, and spectrogram graphs of the blasts were generated in the time-frequency medium (Figure 1). The time lag of the secondary shock wave of the second blast can be clearly seen on this graph, and is calculated to be 11 milliseconds. On the other hand, although it is not very distinct, it was calculated from the graph of the first blast that the time lag of the secondary shock wave was also 11 milliseconds. If the attack was conducted using an ANFO (Ammonium nitrate-fuel oil) type explosive, then the first explosion would have required 0.75 kg TNT equivalent ANFO (0.90 kg ANFO), while the second explosion would have required 0.90 kg TNT equivalent ANFO (1 kg ANFO). These values are inconsistent with the size of the blast and the number of casualties it caused. On the other hand, if the type of explosive used was TNT, the amount of TNT used in the first and the second explosion are calculated as 4.5 kg and 5.5 kg, respectively. When the calculated amounts of explosives and the impact of the explosion are taken into consideration, it is more likely that TNT was used in the attack. The types and amounts of explosives that were identified based on the sound recordings also match the information provided by the official authorities.

2.2 Seismic and Acoustic Recordings Relating to the Terrorist Attack Conducted at Merasim Street on February 17, 2016

A terrorist bomb attack was conducted against service buses carrying Turkish Armed Forces (TAF) personnel on Wednesday February 17, during the evening rush hour. In addition to the suicide bomber, 29 people were killed and 60 people were injured in this terror attack. According to official statements, the attack occurred at 18:31 local time, and was carried out by a suicide bomber sitting in a passenger car.  The distances between the location of the terror attack and the closest seismograph stations, namely the METU, ANKO and KO_LOD stations, were 6,600, 6,610 and 7,150 metres, respectively (Figure 2).

msi-sekil2

Figure 2. The geographic locations of the terror attacks in Ankara that were carried out using vehicles and the geographical locations of the seismic stations

The examination of the recordings revealed that both the ANTO and METU stations recorded the seismic and acoustic waves caused by the explosion in the terror attack, while no vibration recordings were observed at the KO-LOD seismic station. In the analysis of the seismic and acoustic recordings, a Butterworth filter was applied within the range of 0.5-8 Hz. Based on the analysis of the seismic waves, the time of the attack was calculated to be 18:31:15.34±0.2 local time. In other words, the terror attack of February 17 took place 31 minutes 15 seconds and 34 milliseconds past 18:00 (Table 2).

As seen in the Ankara bombing incident of February 17, seismograph stations can record acoustic (sound) waves as well as seismic waves if the blast occurs on the surface (Koper et al., 1999; 2002; Gitterman and Hofstetter, 2012; Gitterman, 2013; Kristekova et al., 2008; Evers et al., 2007). Similar acoustic waves were observed in the Afyonkarahisar Ammunition Depot Explosion that occurred on September 5, 2012, and the Gunpowder Depot Explosion in Hasandede, Kırıkkale, that occurred on August 20, 2013 (Kuran 2014-a). While the duration of acoustic waves is much shorter in surface explosions compared with seismic waves, the magnitude of these acoustic waves are higher. This phenomenon, which was observed during the terror attack of February 17, indicates that the blast occurred either on the surface, or at a certain height from the surface. As most of the liberated energy was released into the atmosphere, the seismic energy was lower (Figure 3).

Table 2. Parameters obtained from the seismic and acoustic recordings of the attack of February 17, 2016

 

 

Calculated Time of Attack

18:31:15.341±0.2

 

Seismic Stations METU ANTO
Distance to the Location of Attack 6,600 m 6,610 m
Arrival Time of the P Wave 18:31:16.765 18:31:16.846
Arrival Time of the Acoustic Wave 18:31:34.445 18:31:34.464
Travel Time of Sound Wave 19.10 s 19.12 s
Speed of Sound Wave at 150C 340.35 m/s
Average Speed of Shock Sound Wave 345.55 m/s 345.71 m/s
Time Lag Before Secondary Shock Wave (∆t=tMS-tSS) 0.183 s

 

msi-sekil3a

 

 

 

 

 

 

 

 

 

(a)

msi-sekil3bc

 

 

 

 

(b)

Figure 3. During the attack of February 17; a) The seismic and acoustic waves recorded by METU and ANTO stations b) Spectrogram graphs of the seismic and acoustic recordings of METU and ANTO stations in time-frequency medium

The time lag before the secondary shock wave was calculated as Δt=0.183 s (183 milliseconds) for the bomb attack (Figure 4). It was calculated that, in the case of factory-made ANFO type explosive being used in the attack, the required amount of TNT equivalent explosive would have been 800 kg (940 kg ANFO). On the other hand, when the relationship determined for TNT type explosives was used, the amount of TNT type explosive required was calculated to be 3,500 kg. The amounts of explosives calculated based on the assumption that either ANFO or TNT was used are inconsistent with the size of the blast. Besides, it is physically impossible to fit such amounts of explosive into a passenger vehicle. This indicates that an IED with lower velocity of detonation was used in the attack. By taking the velocity of detonation of the ammonium nitrate-containing IED as 1,350 m/s, the TNT equivalent weight was calculated as 90 kg. When the conversion coefficients used for the TNT equivalent weight are taken as 0.42 and 0.33, the weight of the IED used in the attack was calculated as 215-270 kg.

msi-sekil4

Figure 4. The time lag of the secondary shock wave calculated based on the acoustic waves at ANTO Station

In the media, it was stated that the terror attack resulted in a two metre-wide crater. Since the attack was carried out with a car, the centre of blast is assumed to be 0.80 m higher than the surface. Using an equation formulated based on previous blasts that have occurred at a certain level from the surface, the TNT equivalent weight of the explosive was calculated as 85 kg. Although we cannot confirm the size of the crater described by the media outlets, the TNT equivalent amount calculated according to this parameter is close to the TNT equivalent amount (90 kg) calculated based on the assumption that the velocity of detonation was 1,350 m/s.

The document called FEMA-426 (2003) ‘Reference Manual to Mitigate Potential Terrorist Attacks Against Buildings,’ published by US Federal Emergency Management Agency, describes the ranges of TNT equivalent amounts of explosives that can be used according to the type of vehicle, as well as the structural and environmental damages that are likely to occur as a result of the shock wave. The data from this document are shown in Figure 5. The possible TNT equivalent amount that can be carried in a passenger vehicle is given as 100-500 lb (45-225 kg). It can be seen that the calculated weight of TNT equivalent of 90 kg is also consistent with the values given in FEMA-426. According to FEMA 426, the area of effect for an IED that is equivalent to 90 kg of TNT can be calculated as follows: The distance limit in which the explosion may result in partial collapse of reinforced concrete in buildings/structures was determined as 6.7 metres. The distance limit from the centre of explosion for fatal injuries was calculated as 33 metres, while the distance limit for non-fatal injuries was calculated as 58 metres, the distance limit for injuries resulting from broken glass as 112 metres, and the distance limit for breaking glass as 160 metres.

msi-sekil5

Figure 5. The TNT equivalent amounts and influence areas of explosives as per FEMA-426

2.3. Seismic and Acoustic Recordings Relating to the Terror Attack in Kızılay on March 13, 2016

On the evening of March 13, a terrorist bomb attack was carried out at a bus stop near Güvenpark, Kızılay. The attack was conducted with a passenger car by two suicide bombers. Other than the suicide bombers, 36 people were killed and 125 people were injured in the terror attack. Official statements describe that the attack was conducted at 18:45 local time. When the relevant recordings are examined, it can be seen that all three seismic stations recorded both the seismic and acoustic waves caused by the blast that occurred during the terror attack. As a result of the analysis about seismic waves, the time of attack was calculated as 18:46:06.62±0.2  local time. In other words, the terror attack on March 13 took place 46 minutes 6 seconds and 62 milliseconds past 18:00. (Table 3).

Table 3. Parameters obtained from the seismic and acoustic recordings of the attack of March 13, 2016

 

 

Calculated Time of Attack

18:46:06.620±0.2

 

Seismic Stations METU ANTO KO-LOD
Distance to the Location of Attack 7,665 m 7,675 m 8,370 m
Arrival Time of the P Wave 18:46:08.273 18:46:08.275 18:46:08.425
Arrival Time of the Acoustic Wave 18:46:28.700 18:46:28.698 18:46:30.841
Travel Time of Sound Wave 22.08 s 22.08 s 24.22 s
Speed of sound at 130C 339.16 m/s
Average Speed of Shock Sound Wave 347.15 m/s 347.60 m/s 345.58 m/s
Time Lag Before Secondary Shock Wave (∆t=tMS-tSS) 0.220 0.224

 

Similarly to the attack of February 17, the magnitude of acoustic waves of the March 13 attack is larger than seismic waves. This indicates that the blast occurred either on the surface or at a certain height from the surface. As most of the liberated energy was released into the atmosphere, the seismic energy was lower (Figure 6). At the moment of the attack, the weather in Ankara was 130C, and the speed of wind was 3.8 m/s at north-northeast direction. The speed of sound waves in the atmosphere is 339.16 m/s at a temperature of 130C and when there is no airblast effect. It can be seen that the speed of the acoustic waves were increased by the airblast effect that occurred after the explosion, with the waves reaching the seismic stations 0.52, 0.55 and 0.45 seconds earlier, respectively. As expected, the average speed of sound wave, calculated based on existing recordings, decreases with distance and becomes closer to reaching the flexible limit.

(a)

msi-sekil6a

(b)

msi-sekil6b

 

 

Figure 6. Recordings of the METU, ANTO and KO-LOD seismic stations concerning the attack of March 13 (a) and the seismic and acoustic recordings of the ANTO Station (b)

When the attack of March 17 is examined in the frequency medium, scalloping is observed in the spectrogram (Figure 7), indicating that there may be more than one blast at the millisecond level. The analyses carried out in the time and frequency medium also indicate that a second blast with a relatively lower impact took place 1.33 seconds after the main blast.

msi-sekil7

Figure 7. Spectrogram graphs in the time-frequency medium of the acoustic recordings of ANTO Station (on the left), KO-LOD Station (in the middle), and METU Station (on the right) concerning the March 13 attack

By using acoustic waves, the time lag before the secondary shock wave was calculated as Δt=0.224 s (224 milliseconds) by ANTO station, which is located a distance of 7,675 m, and as Δt=0.220 s (220 milliseconds) by METU station, which is located at a distance of 7,665 m from the attack site (Figure 8). No significant secondary shock waves were observed in the KO-LOD station.

msi-sekil8

Figure 8. Three components of the attack of March 13, as recorded by ANTO Station (on the left) and the vertical component from the recordings of METU Station (on the right), the time lags of the secondary shock waves

As was the case with the February 17 attack, the amounts of explosives calculated based on the assumption that either ANFO or TNT was used in the attack are inconsistent with the observed size of the blast. Furthermore, it is physically impossible to fit such amounts of explosive into a passenger car. This shows that an IED with lower velocity of detonation was used in the attack. The TNT equivalent weight was calculated as 165 kg by taking the velocity of detonation of the IED as 1,350 m/s, and by using the time lags before the secondary shock waves recorded by ANTO and METU stations (Table 4). On the other hand, when the coefficient of TNT equivalent weight of IED is taken as 0.33 and 0.42, the weight of the IED used in the attack is calculated as being in the region of 390-500 kg.

Table 4. Amounts of explosives calculated based on the time lag before the secondary shock wave

Seismic Station IED (TNT Equivalent Weight-kg) IED (kg)
ANTO 170 kg 400-515 kg
METU 160 kg 380-485 kg
AVERAGE 165 kg 390-500 kg

According to FEMA 426, the area of effect of an IED that is equivalent to 165 kg of TNT is calculated as follows: The distance limit that may result in partial collapse of reinforced concrete in buildings/structures was determined as 8.5 metres. The distance limit from the centre of explosion for fatal injuries was calculated as 46 metres, while the distance limit for non-fatal injuries was calculated as 64 metres, the distance limit for injuries resulting from broken glass as 145 metres, and the distance limit for breaking glass as 200 metres. Despite being relatively limited. the effect of the blast on the surrounding area was also examined. While there was more broken glass in the buildings at the corner of Kumrular Street than are situated 185 metres from the location of blast, it was observed that a large window on the TMO building at a distance of 198 metres was also broken. No broken glasses were observed in the neighbouring buildings that are located 10-15 metres from this building. Although the amount of empirical data that could be obtained was limited, the impact of the blast on the surrounding area is consistent with the amount of explosive calculated based on FEMA 456.

While three seismic stations recorded the March 13 attack, the February 17 attack was recorded by only two of the stations. Although an evaluation of the acoustic recordings indicates that the temperature was lower and the distance between the location of attack and the station was higher in the March 13 attack, the average acoustic speed of the wave was calculated to be higher compared to the February 17 attack. These findings show that the amount of explosive used in the March 13 attack was higher than the amount of explosive used in the February 17 attack. When the TNT equivalent weights calculated based on the secondary shock wave parameter are taken as a reference, it can be seen that the amount of explosive used in the March 13 attack was 1.8 times greater than the amount used in the February 17 attack.

 

  1. Coup Attempt of July 15, 2016, and the Chronology of F-16 Air Strikes

Members of FETÖ carried out a coup attempt on the night of July 15. Among those who resisted the coup-plotters, 246 people, most of whom were civilians, were killed. During the coup attempt, members of the FETÖ terror organisation conducted bombing attacks with F-16 fighter jets against several locations. In addition, many sonic booms occurred in the skies over Ankara, as the fighter jets broke the sound barrier. The ANTO, METU and KO-LOD seismic stations recorded the vibrations that occurred both as a result of the sonic booms and the bombing attacks conducted by the F-16s. Places in Ankara that were bombed by the coup plotters during the coup attempt included TÜRKSAT, the Gölbaşı Police Special Operations Centre (PSOC), the Ankara Police Department, the Turkish Grand National Assembly (TGNA), and the Presidential Complex. On the other hand, air units of the TAF bombed the Akıncı Air Base, which was being used as an operations base by the coup plotters. The distances of these locations that were bombed to the ANTO, METU and KO-LOD seismic stations are given in Table 5.

 

Table 5. The distances of seismic stations to the locations that were bombed

TÜRKSAT Gölbaşı PSOC Police Department TGNA Presidential Complex Akıncı
ANTO 25,400 m 8,400 m 9,500 m 6,800 m 6,700 m 30,700 m
METU 25,400 m 8,400 m 9,500 m 6,800 m 6,700 m 30,700 m
KO-LOD 28,000 m 11,800 m 8,700 m 7,700 m 5,500 m 27,300 m

 

An evaluation of the recordings of the seismic stations shows that there are three different types of vibration recorded. The first one is the recording of the seismic and acoustic vibrations resulting from bombing attacks, which were covered and explained in the previous section. By analysing these recordings, the location of each bombing attack was identified, and their timing was calculated at the level of milliseconds (Table 6). The error margin in the calculated timings of the attacks is within ±0.2 seconds. The distance of TÜRKSAT and the Akıncı Air Base to the seismic stations is relatively greater than to the other locations. Thus, since no seismic and acoustic recordings are available for any of the blasts that happened at these two locations, the timing of the explosions at these places could not be calculated. The second type of vibration that was recorded by the seismic stations was the acoustic recording of the sonic booms that occurred when the F-16 fighter jets reached or broke the sound barrier. Explosions caused by TNT, ANFO or IEDs occur as a result of chemical reactions, and the energy release takes place almost immediately, in two or three milliseconds. When an object breaking the sound barrier causes a sonic boom, it is actually a physical phenomenon. Many sonic booms occurred during the coup attempt and in most cases, witnesses confused them with bomb blasts. Records indicate that sonic booms do not generate seismic waves. The third type of vibration recording is the recording of sounds made by F-16 fighter jets while flying at a low altitude and at less than the speed of sound. As with sonic booms, this type of vibration does not result in seismic vibrations that can be recorded by seismic stations. However, the duration of these acoustic recordings is longer than that of sonic booms, and their magnitude is higher at certain frequencies (Figure 9-16).

 

Table 6. The locations and timings of F-16 attacks calculated on seismic and acoustic recordings

Attack No. Target Timing

(hour:minute:second.millisecond)

1 Gölbaşı Police Special Operations Centre 23:17:25.900 .
2 Gölbaşı Police Special Operations Centre 23:59:20.990
3 Ankara Police Department 00:55:38.354
4 Ankara Police Department 01:07:34.814
5 TGNA 02:33:09.005
6 TGNA 03:22:49.665
7 TGNA 03:23:47.567
8 Presidential Complex 06:16:32.913
9 Presidential Complex 06:17:44.982

 

It is possible to see that the first target of the coup plotters was the police unit, since they had the capability of fighting back. The first F-16 attack took place at the Gölbaşı Police Special Operations Centre at 23:17:25.90, and 41 minutes 55 seconds after this attack, at 23:59:20.99, a second attack was carried out against the same location. Fifty police officers were martyred at the Gölbaşı Police Special Operations Centre, the place where both attacks were conducted. The third attack of the coup plotters was conducted against the Ankara Police Department at 00:55:38.354. The fourth bomb attack was also conducted against the same location at 01:07:34.814, 11 minutes 56 seconds after the third attack. The fifth bomb attack was conducted against TGNA at 02:33:09.005. Both the media and the parliamentarians who were at the site during the attack stated that the TGNA was bombed at 02.38, 02.42 and 02.49. However, recordings from the seismic stations show that these blasts were not associated with bombing attacks, but that they were instead sonic booms caused by F-16 fighter jets exceeding the speed of sound. The sixth and the seventh F-16 attacks were also conducted against TGNA at 03:22:49.665 and 03:23:47.567 with an interval of 57.9 seconds. The eighth attack was conducted against a location between the General Command of Gendarmerie and the Presidential Complex at 06:16:32.913. Finally, the ninth attack was conducted against a location near Millet Mosque in the campus of the Presidential Complex at 06:17:44.982, 1 minute 12 seconds after the eighth attack. As there was only a 57.9 second time difference between the sixth and seventh attacks against TGNA, there is a strong possibility that the same fighter jet conducted these attacks. The same can possibly be claimed for the eighth and ninth attacks conducted against the Presidential Complex, with a time difference of 1 minute 12 seconds. Since the type of explosive used in the ammunition has a high velocity of detonation, the time lag before their secondary shock is very short. For this reason, the amount of explosives could not be calculated. Analysis of the recordings shows that two types of ammunition were used in the bombing attacks. It is believed that while ammunition with a lower amount of explosive was used in attack number 6, which was conducted against the TGNA at 03:22:49.665, the same type of ammunition with a higher amount of explosives was used in the other attacks. It was also observed that while a single blast occurred during attack number 6 against the TGNA, two consecutive blasts took place in the remaining attacks between the seconds 0.16-0.25. It is believed that these consecutive blasts were due to the technical features of the ammunition used. The Mk84 bomb (850 kg TNT) is the consecutive blast effect ammunition used in the inventories of the Turkish Armed Forces, while the Mk82 (224 kg TNT) is a type of ammunition with a lower amount of explosive and single blast feature. It is therefore believed that the Mk82 bomb was used in attack number 6 that was carried out against the TGNA, while the Mk84 was used in the other attacks.

msi-sekil9

Figure 9. Vertical component of the seismic and acoustic recordings at METU Station concerning the F-16 attacks number 1 and number 2 that were conducted against Gölbaşı PSOC (between 23:00-23:59)

msi-sekil10

Figure 10. Vertical component of the seismic and acoustic recordings of ANTO Station concerning the F-16 attack number 3 that was conducted against Ankara Police Department (between 00:00-00:59)

msi-sekil11

Figure 11. Vertical component of the seismic and acoustic recordings and of the sonic boom acoustic recordings at KO-LOD Station concerning the F-16 attack number 4 that was conducted against Ankara Police Department (between 01:00-01:59)

 

msi-sekil12

Figure 12. Vertical component of the seismic and acoustic recordings and of the sonic boom acoustic recordings at METU Station concerning the F-16 attack number 5 that was conducted against the TGNA (between 02:00-02:59)

 

msi-sekil13

Figure 13. Vertical component of the seismic and acoustic recordings and of the sonic boom acoustic recordings at ANTO Station concerning the F-16 attacks number 6 and 7 that were conducted against the TGNA (between 03:00-03:59)

 

msi-sekil14

Figure 14. Vertical component of the seismic and acoustic recordings of ANTO Station concerning the F-16 attacks number 8 and 9 that were conducted against the Presidential Complex (between 06:15-06:20)

msi-sekil15

Figure 15. Vertical component of the vibration recordings and spectrogram graph at ANTO Station concerning sound waves caused by typical sonic booms

msi-sekil16

Figure 16. METU Station’s vibration recordings of the vertical component and spectrogram graph concerning the sound waves that occurred as a result of the F-16 flights conducted at low altitude

4.Conclusion

In this study, the timing of the terror attacks carried out on October 10, 2015, February 17, 2016, and March 13, 2016 were calculated at a millisecond level, along with the number of blasts that occurred during these attacks, and the types and TNT equivalent weights of the explosives used. Following this, the TNT equivalent weights were converted depending on the type of explosive, and the weights of the explosives used were calculated accordingly. Since no seismic recordings are available for the suicide bomb attack that was performed in front of Ankara Railway Station on October 10, 2015, a video recording was used instead in the analysis. Concerning the coup attempt of July 15, the timing of bombing attacks conducted by F-16 fighter jets against the Gölbaşı Police Special Operations Centre, the Ankara Police Department, the Turkish Grand National Assembly, and the Presidential Complex were calculated at a millisecond level, and the number and locations of the air strikes were determined. In addition, sonic booms were differentiated from the bombing attacks, and the actual number of air strikes other than the ones conducted against TÜRKSAT and the Akıncı Air Base was identified.

Attacks conducted with a high number of bombs can be recorded by seismic recorders; however, when the seismic sensors are not very close to the location of a small explosion, as was the case with the attack conducted against Ankara Railway Station on October 10, these recorders will not be able record any seismic movements. With such small-scale explosions, information about the type and amount of explosive, as well as the timing of the explosion, can be calculated from the sound track obtained from video recordings. Furthermore, it is believed that placing pressure sensors with continuous measurement features in cities at different locations might be very helpful in identifying the location and timing of such attacks, as well as the type and amount of explosives used.

The analysis of seismic and acoustic recordings allow for the sensitive identification of the timing and location of incidents. There is, undoubtedly, always an error margin involved when calculating the amount of explosives using these recordings. In order to further reduce the error margin when calculating the amount of explosives, various kinds of explosive should be detonated for experimental purposes, with the aim of obtaining and examining different types of blast. By improving the parameters used for evaluating these blasts, or by developing new ones, new equations could be formulated for calculating the amount of explosive.

Sources

  1. ABS Consulting (2015) West Fertilezer Incident Support Services Finalr Report, San Antonio, USA
  2. Ambrosini D., Luccioni B., Danesi R. (2004) “Influence of the Soil Properties on Craters Produced by Explosions on The Soil Surface,Mecanica Compitacional, Vol.XXIII, pp. 571-590
  3. Ambrosini D., Luccioni B., Danesi R. (2003) “Craters Produced by Explosions on The Soil Surface” Mechanica Computacional, Vol.XXII, pp. 678-692
  4. Ambrosini R.D., Luccioni B.M., Danesi R.F., Riera J.D., Rocha M.M. (2002), “Size of Craters Produced by Explosive Charges,” Shock Waves, Vol.12, pp . 69-78
  5. Bonner J., Waxler R., Gitterman Y., Hofstetter R. (2013), “Seismo-Acoustic Energy Partitioning at Near Source and Local Distances from the 2011 Sayarim Explosions in The Negev Desert Israel”, Bulletin of The Seismological Society of America, Vol.103, No:2A, pp.741-758
  6. Evers L.G., Cerenna L., Haak H.W., Le Pichon A., Whitaker R.W. (2007), “A Seismoacoustic Analysis of the Gas-Pipeline Explosion near Ghislenghien in Belgium” Bulletin of the Seismological Society of America, Vol:97, No:2, pp:417-425
  7. FEMA 426 (2003) Reference Manual to Mitigate Potential Terrorist Attacks Against Buildings, USA
  8. Gitterman Y. (2013) “Secondary Shock Features for Large Surface Explosions: Results from the Sayarim Military Range, Israel and other Experiments” Shock Waves, DOI 10.1007/s00193-013-0487-y
  9. Gitterman Y., Hofstetter R. (2012) “GT0 Explosion Sources for IMS Infrasound Calibrations: Charge Design and Yield Estimation from Near-Source Observation,” Pure and Applied Geophysics, Published online, 06 September
  10. Gitterman Y., Given J., Coyne J., Zerbo L., Hofstetter R. (2011), “Large-Scale Explosion Source at Sayarim-Israel for Infrasound Calibration of The International Monitoring System” S&T2011, Vienna, Austria, 8-10 June
  11. Gitterman Y., Pinsky V., Hofstetter A. (2007), Seismic Energy Generation and Partitioning into Various front Different Seismic Sources in the Middle East Region”, Israel Air Force Researh Laboratory, Final Report, September
  12. Gitterman Y., Pinsky V., Hofstetter A. (2005), Using Ground Truth Explosion for Studying Seismic Energy Generation and Partitioning into Various Regional Phases,” 27th Seismic Research Review: Ground-Based Nuclear Explosion Monitoring Technologies, Ranco Mirage, California-USA, 20-22 September
  13. Koper K.D., Wallace T.C., Aster R.C. (2003), “Seismic Recordings of the Carlsbad, New Mexico Pipeline Explosion of 19 August 2000” Bullettin of the Seismological Society of America Vol.93, No:4,pp.1427-1432
  14. Koper K.D.,Wallace T.C., Reinke R.E., Leverette J.A. (2002) “Empirical Scaling Laws for Truck Bomb Explosions Based on Seismic and Acoustic Data” Bulletin of the Seismological Society of America, Vol:92,No:2, pp:527-542
  15. Koper K.D., Wallace T.C., Hollnack D. (1999), “Seismic Analysis of the 7 August 1998 Truck Bomb Blast at the American Embassy in Nairobi-Kenya” Sesimological Research Letters, Vol.70, No:5,pp.512-521
  16. Kristekova M., Moczo P., Labak P., Cipciar A., Fojtikova L., Madaras J., Kristek J. (2008), “Time-Frequency Analysis of Explosions in the Ammunition Factory in Novaky Slovakia”, Bulletin of The Seismological Society of America, Vol.98, No:5, pp.2507-2516
  17. Kuran F. and Polat S. (2015) ‘5 Eylül 2012 Afyonkarahisar Mühimmat Deposu Patlamasının Sismik Kayıtlar Kullanılarak İrdelenmesi’ The 3rd National Conference on Earthquake Engineering and Seismology, October 14-16, 2015 – DEÜ (Dokuz Eylül University) – İzmir
  18. Kuran F. (2014-a) “Three Massive Explosions and Seismology” 2nd European Conference on Earthquake Engineering and Seismology (2nd ECEES), 25-29 August 2014, Istanbul
  19. Kuran F. (2014-b) ‘Suriye İç Savaşının İstihkâmcılık Açısından İrdelenmesi’ MSI Turkish Defence Review, Issue: 113, pages 58-69
  20. Mihelic B. (2013) “Unplanned Explosions at Munitions Sites (UEMS),” The 6th Workshop of the South East Europe Regional Approach to Stockpile Reduction (RASR) of Conventional Weapons and Munition Initiative”, 3-5 March, Bled, Slovenia
  21. Ottemöller L. and Evers L.G. (2008), Seismo-Acoustic Analysis of the Buncefield Oil Depot Explosion in The UK, 2005 December 11,” Geophysics Journal, 172, pp. 1123-1134
  22. Reynolds J.M. (2011), An Introduction to Applied and Environmental Geophsics”, Second Edition, Wiley-Blackwell Press
  23. Rigby S.,E. And Sielicki P.(2014) “An Investigation of TNT Equivelance of Hemispherical PE4 Charges” Engineering Transactions, Vol:62-4, pp.423-435
  24. Salzano E., Landucci G., Reniers G. and Cozzani V.(2014) “Domino Effects Related to Home-Made Explosives, Chemical Engineering Transactions, Vo:36, pp. 349-354
  25. Toksöz, M.N., Kuleli, S., Gürbüz, C., Kalafat, D., Bekler, T., Zor, E., Yılmazer, M., Öğütçü, Z., Schultz, C.A., and Harris, D.B. 2003. Calibration of regional seismic stations in the Middle East with shots in Turkey. Proceedings of the 25th Annual Seismic Research Review, Volume I, 162-171.
  26. Weber P.W., Millage K.K., Crepeau J.E., Happ H.J., Gitterman Y., Needham C.E. (2014) “Numerical Simulation of a 100 ton ANFO Detonation”, Shock Waves, DOI 10.1007/s00193-014-0547-y
  27. http://www.orfeus-eu.org/eida
  28. http://www.deprem.gov.tr

 

800 total views, 2 views today