Wreckage Floating debris was collected in the early morning hours after the accident. Two days after the accident, officials had established a one mile exclusion zone against fishing activities near the accident site, and the AAIB had requested specialized deep water recovery equipment to recover the wreckage and, they hoped, the bodies of the pilot and paramedic. Six days later the
diving support vessel Seaway Osprey arrived on the scene of the accident with
remotely operated vehicles (ROV's) and
saturation divers capable of manually recovering debris. Within twelve hours all major components of the aircraft had been recovered except for the left wingtip, which was discovered on 2 May 2005, in a
trawler's fishing net about from the debris field. The wreckage was brought to the AAIB's hangar at
RAF Machrihanish (on which Campbeltown Airport was located at the time of the accident) for analysis. All of the primary structure was recovered from the seabed. Controls were
trimmed near the neutral position. The flaps were in the "UP" position (i.e., retracted). Investigators found no evidence that the controls were jammed. The left main landing gear leg had become separated during the impact, and the right gear leg was bent aft. The nose gear was bent aft and to the right. Both propellers were recovered and showed similar damage. Both engines were stripped down and examined. They were found to be in similar condition, with valves and other moving parts in good shape. Both engines could be turned when the spark plugs were removed. Both engines, with propellers attached, were found at the northern end of the debris field. The cockpit area of the Islander had not suffered significant damage, and there was sufficient space for the pilot to survive. The pilot's seat belt and shoulder harness were found undamaged and unbuckled, with all stitching intact and the buckle still functional. There was no sign of impact on the controls or the instrument panel. Both fuel tanks contained a significant amount of fuel, which investigators calculated to be approximately 100 US gallons in each of the two wing tanks, and the fuel selector valve was in the correct position. Several instruments were damaged by the impact or by the pressure and corrosion from sitting on the seabed for six days. The aircraft's
horizontal situation indicator, or HSI (a navigational instrument to direct the pilot to or from ground-based navigation aids, including the MAC VOR), was not set properly for the approach. The instrument's course selector was set to 103°, though the approach called for a setting of 115°. A second device on the HSI, called the "heading bug", is a prominent marker for the pilot to set to the desired heading so that when the aircraft is flying that heading, the bug is at the top of the instrument's display and aligned with a vertical white line. The heading bug on the HSI was set to 157°. A second instrument situated below the HSI and capable of the same VOR navigation, called the omni bearing indicator (OBI), was also set incorrectly to 309° when it too should have been set to 115°.
Analysis All wreckage was found along a 209-metre long, 50-metre wide section of the seabed at a depth of 78 metres. Investigators found no evidence of structural failure inflight. The damage to the aircraft was roughly symmetrical left-to-right, indicating the aircraft was close to wings-level flight on impact. The separation of the left landing gear leg compared to the bending of the right leg suggested to investigators that impact with the water happened with the left wing slightly lower than the right, as the left gear was the more damaged, and therefore subject to more force against the water. When the left wingtip contacted the water, investigators conclude that the aircraft cartwheeled, bringing the right wing down into the water with great force, which they report as being consistent with the nose gear being bent to the right as well as aft. The pattern of damage to the nose and the underside of the fuselage suggests the aircraft was in a slightly nose-down attitude, as if in a normal descent. Nearly neutral trim settings indicated the pilot was not trying to compensate for some abnormal flying condition, such as an engine failure or flight control problems. The pitch trim was set to a slight nose-down attitude, which the aircraft manufacturer calculated to indicate stable flight at 110–120 kn. This is consistent with investigators' calculations based on the damage to the aircraft that it had been moving at 90–130 kn at impact. Damage to the propeller blades is consistent with both of them turning and similar engine power being delivered to them. The amount of power being delivered to the propellers could not be determined. Therefore, it remains possible that ice had accumulated in the carburettors and reduced engine power. Although meteorological conditions were favorable for
carburettor icing, and the
carburettor heat levers were in the "OFF" position, loss of engine power is not consistent with the speed and shallow angle of impact calculated from the wreckage. Analysis of the engines verified that carburettor heat was not being used at the time of the accident. The AAIB concluded its analysis of the wreckage with the following statement:
Weather and environment Although clouds extended down as low as above the sea surface and surface visibility had fallen to meters at Campeltown Airport, the wind was from the west at 12 knots, and the aircraft was below the freezing level (reported at ). Conditions at the time of the accident made it possible for ice to form in the carburettors, and carburettor heat was not applied to counteract this possibility, but there was no sign that the engine function was impaired. The switches for pitot and stall warning probe heating, and for propeller de‑icing were selected ON. The switches for airframe de-icing, heated windshield and for the ice inspection lamp were selected OFF.
Analysis Low clouds, deteriorating visibility, and winds that favored Runway 29 in Campbeltown, which did not have an instrument approach procedure necessary for landing in such conditions as were forecast for the expected arrival time, posed serious challenges to making a safe landing in Campbeltown. As G-BOMG neared its destination, conditions were worsening. The pilot chose to follow the instrument approach procedure for Runway 11 until he saw the airport, at which point he would circle the airport to land on Runway 29. To perform this manoeuvre, the pilot was not permitted to descend below an altitude of . However, the weather reports the pilot received when he departed and others he received while en route show the
cloud base was down to ("broken" cloud layer). This was acceptable for an approach to Runway 11, but not for the circling manoeuvre to Runway 29. At the same time, the wind had increased from 8 kn to 15 kn in Campbeltown, making a landing on Runway 11 increasingly difficult, with significant
crosswind and
tailwind components. Four minutes before the accident, the controller in Campbeltown advised the pilot that there were now a "few" clouds at as well as the broken layer at . For the duration of the flight, the pilot likely had very little visual contact with the ground due to being in or over clouds, or out over water in low visibility. As the aircraft descended through the cloud base, the pilot likely had no visual cue that he was below the clouds because the aircraft was about out to sea and visibility was reported as . The pilot of a search and rescue helicopter later commented that the sea surface that night would have provided the pilot with little or no visual cues about the aircraft's dangerously low altitude.
Human factors The pilot awoke at 6:45 AM on the day of the accident. Knowing he was on call duty overnight, he apparently took no rest any time during that day or evening. The AAIB acknowledges it can be difficult for a well rested person to sleep during the daytime, and they note that quiet rest is the best some can achieve, but they also point out that 16–17 hours of continuous wakefulness is associated with diminished performance. Single-pilot IFR operations place a heavy workload on the pilot. That workload increases as the pilot begins the approach phase of the flight. During the approach phase, the pilot announced his intention to land on Runway 25, though he knew the actual runway was 29. The AAIB's
human factors analysis suggests this specific type of error might be a significant indicator that the pilot's stress level had increased to the point where it began to further diminish his ability to cope with the heavy workload. They list several potential yet unknown causes for the increase in stress, such as disorientation or an undeclared emergency. Concerning the pilot's performance, the AAIB summarizes:
Postmortem examinations The body of the pilot was found nine months after the accident by a fishing vessel four miles off the coast in
Machrihanish Bay. No obvious internal or external injuries could be detected, nor were there any
bone fractures. Because of the condition of the body when it was recovered, it was not possible to determine the cause of death. The body of the paramedic was recovered, still restrained in the seat by his undamaged lap belt, during search and rescue operations six days after the accident. The paramedic had suffered a major head injury from a frontal impact with the back of the pilot's seat. The paramedic "almost certainly" lost consciousness instantly, and the head injury was potentially fatal, but changes in the lungs showed the cause of death to be drowning.
Testing Paramedic's seat belt To determine the likely cause of the paramedic's head injury, the AAIB requested that Loganair have a man of the same height as the paramedic sit in the paramedic's seat of another of their Islander air ambulances, fasten his seat belt, and lean fully forward to see what obstructions might be encountered. The subject's head came close to the back of the pilot's seat but did not contact it. However, under rapid deceleration, or had the seat belt been slightly looser, his head would have contacted the back of the pilot's seat.
Enroute navigation Navigation charts used by pilots in the area warn that, because of terrain, the MAC VOR is unreliable for aircraft below or farther than in the direction from which G-BOMG was navigating toward Campbeltown. The AAIB conducted a test flight in a similar Islander aircraft along the same route flown by G-BOMG. Despite the warnings on the navigation charts, the test aircraft was able to navigate by using the MAC VOR during the entire flight. However, the MAC VOR is also equipped with
distance measuring equipment, or DME, which provides a means for pilots to know the distance between their aircraft and the VOR/DME station. On the test flight, the DME signal was not received until the aircraft was from the MAC VOR.
Approach radar coverage After the enroute navigation test flight was completed, the AAIB's test pilot flew several approaches to Campbeltown Airport with the same aircraft while investigators monitored its position on the radar equipment that was used the night of the accident. When the test aircraft flew the approach as indicated on the navigation chart that was found clipped to the pilot's control wheel, there was a small gap in radar coverage when the aircraft was at its most distant point from the airport, beginning at and ending on the return segment at from the MAC VOR. Other trials were conducted at lower altitudes, and in those trials the coverage gap was widened considerably. Investigators tried approaches at low and descending altitudes involving a steeper left turn to 157° (where the heading bug was set on the HSI), and then flying that heading until intercepting the correct course to the MAC VOR. When this track was flown at , the last radar contact was at from the MAC VOR. The last radar contact with the accident flight was at from the MAC VOR, at an altitude of . Flying this track with the accident flight's known and projected descent profile brought the test aircraft to within of the crash site at an altitude of , which was within the margin of error for the differences in wind on the day of the accident and on the day of the test flight.
Conclusions from the AAIB investigation Selected findings The AAIB published 47 separate findings based on their investigation. Below are some of the findings that have greater historical significance because they bear directly on the causes of the accident or the causes of death of the pilot and paramedic, or because they served as a basis for changes in the air transport industry. • "No evidence of a technical fault was found that could explain the accident, though a malfunction affecting critical flight instruments could not be entirely ruled out." • Although a safe landing at Campbeltown was in doubt, the pilot was authorised to make an approach. Weather at his two designated alternate landing airports (Glasgow and Prestwick) was good in case the plane needed to divert. • "The presence of a second pilot may have prevented the accident." • "Had the aircraft been equipped with a radio altimeter, or other electronic low height warning device, which was correctly set to warn of a low height situation, the accident may not have occurred." • "It is probable that the pilot was suffering, at least to some extent, from the affects [sic] of fatigue." • "The pilot may have been operating under high workload, or even overload, conditions in the latter stages of the flight, which may have degraded his situational awareness." • "The paramedic was probably rendered unconscious in the impact when his head hit the pilot's seat in front due to the lack of upper torso restraint." • "Average survival time in the sea, at a temperature of 9 °C, would have been no more than one hour."
Causal factors The AAIB determined the following three causal factors, which are quoted directly from their report. • "The pilot allowed the aircraft to descend below the minimum altitude for the aircraft's position on the approach procedure, and this descent probably continued unchecked until the aircraft flew into the sea." • "A combination of fatigue, workload and lack of recent flying practise probably contributed to the pilot's reduced performance." • "The pilot may have been subject to an undetermined influence such as disorientation, distraction, or a subtle incapacitation, which affected the pilot's ability to safely control the aircraft's flightpath."
Safety recommendations The AAIB made three safety recommendations to UK and European regulators based on the findings of their investigation of this accident. They are quoted below.
2006-101: "The
European Aviation Safety Agency and
Joint Aviation Authorities should review the
UK Civil Aviation Authority's proposal to mandate the fitment of Upper Torso Restraints on all seats of existing Transport Category (Passenger) aeroplanes below 5,700 kg being operated for public transport, and consider creating regulation to implement the intent of the proposal."
2006-102: "Considering the unique circumstances of air ambulance flights, the Civil Aviation Authority, in conjunction with the Joint Aviation Authorities should review the circumstances in which a second pilot is required for public transport flights operating air ambulance services."
2006-103: "The Civil Aviation Authority, in conjunction with the Joint Aviation Authorities, should consider mandating the carriage of a radio altimeter, or other independent low height warning device, for public transport IFR flights operating with a single pilot." ==Aftermath==