Investigation
The National Transportation Safety Board (NTSB) announced that they would send a team to the crash site on February 13 to begin the investigation. NTSB spokesman Steve Chealander said that 14 investigators were assigned to the crash of Continental Connection Flight 3407. Both the Flight data recorder (FDR) and the Cockpit voice recorder (CVR) were retrieved and analyzed in Washington, D.C.
After initial FDR and CVR analysis, it was determined that the aircraft went through severe pitch and roll oscillations after positioning its flaps and landing gear for landing. Until that time, the Dash 8 had been maneuvering normally. The de-icing system was reported to be turned on. During descent, the crew reported about 3 miles (4.8 km) of visibility with snow and mist. Preceding the crash, the aircraft's stall-protection systems had activated. Instead of the aircraft's diving straight into the house as was initially thought, it was found that the aircraft fell 800 feet (240 m) before crashing pointing northeast, away from the destination airport. The passengers were given no warning of any trouble by the pilots. Occupants aboard the Dash 8 experienced an estimated force two times that of gravity just before impact. Chealander said information from the aircraft's flight data recorder indicates that the plane pitched up at an angle of 31 degrees, then down at 45 degrees. The Dash 8 rolled to the left at 46 degrees, then snapped back to the right at 105 degrees, before crashing into the house, and erupting in flames on impact.
At the crash scene, an area 2 square miles (5.2 km2) in size was cordoned off, despite the small footprint of the actual damage. Investigators stated it would take three or four days to remove all human remains and a few weeks to positively identify them. As the recovery efforts proceeded, Chealander remarked that freezing temperatures as well as difficulty accessing debris were slowing the investigation. Portable heaters had to be brought to the site to melt ice left in the wake of the firefighting efforts. Initial analysis of the aircraft's remains revealed the cockpit had sustained the greatest impact force, while the main cabin was mostly destroyed by the ensuing fireball. Towards the rear of the aircraft, passengers were found still strapped in their seats.
On February 15, more information on the crash was released by the NTSB saying it appeared the plane had been on autopilot when it went down. The investigators did not find evidence of the severe icing conditions that would have required the pilots to fly manually. Colgan Air recommends pilots fly manually in icing conditions, and requires they do so in severe icing conditions. The NTSB had issued a safety alert about the use of autopilot in icing conditions in December 2008. Without flying manually, pilots may be unable to feel changes in the handling characteristics of the airplane, which is a warning sign of ice buildup. The NTSB also revealed that the plane crashed a mere 26 seconds after trouble was first registered on the flight data recorder.
More details emerged on February 18. It was reported that a re-creation of events leading up to the crash indicated that the stick pusher had activated, which pushes the nose down when it determines a stall is imminent in order to maintain airspeed so the wings continue to generate lift and keep the aircraft aloft. The crew, concerned about a nose-down attitude so close to the ground, may have responded by pulling the nose upward and increasing power, but over-corrected, causing a stall or even a spin. Bill Voss, president of Flight Safety Foundation, told USA Today that it sounded like the plane was in "a deep stall situation".
On March 25, 2009, NTSB investigators said that icing probably did not contribute greatly to the accident. On May 11, 2009, new information came out that Captain Renslow had failed three "check rides" - the flying equivalent of driver proficiency tests, including some at Gulfstream International with its controversial pay-to-fly program, and it was suggested that he may not have been adequately trained to respond to the emergency that led to the airplane's fatal descent. Crew fatigue was also suspected, as both pilots appear to have been at Newark airport overnight and all day prior to the 9:18 pm departure. In response to questioning from National Transportation Safety Board members, Colgan Air officials acknowledged that both pilots apparently were not paying close attention to the aircraft's instruments and failed to follow the airline's procedures for handling an impending stall in the final minutes of the flight. 'I believe Capt. Renslow did have intentions of landing safely at Buffalo, as well as first officer Shaw, but obviously in those last few moments ... the flight instruments were not being monitored, and that's an indication of a lack of situational awareness,' said John Barrett, Colgan's director of flight standards. The official transcript of the crew's communication, obtained from the cockpit voice recorder, as well as an animated depiction of the crash, constructed using data from the flight data recorder were made available to the public on May 12, 2009, the first day of the public hearing. Some of the crew's communication violated federal rules banning nonessential conversation.
On June 3, 2009, the New York Times published an article detailing complaints about Colgan's operations from an FAA inspector who observed test flights in January 2008. As in a previous FAA incident handling other inspectors' complaints, the Colgan inspector's complaints were deferred and the inspector was demoted. The incident is under investigation by the Office of Special Counsel, the agency responsible for U.S. Government federal whistle-blower complaints.
Safety issues examined during the accident investigation process included pilot training, hiring, and fatigue problems, leading the FAA to issue a "Call to Action" for improvements in the practices of regional carriers.
On February 2, 2010, the NTSB adopted its final report into the accident. This was the first time in 15 years that a report had been adopted by the NTSB in less than a year from the date of the accident. It concluded that the cause of the accident was pilot error.
The captain failed to react in the proper manner, by decreasing the angle-of-attack, when the stick shaker activated. Instead, following the activation of both the stick shaker and the stick pusher, he countermanded by pulling back on the stick, which greatly exacerbated the situation. "...his improper flight control inputs were inconsistent with his training and were instead consistent with startle and confusion. It is unlikely that the captain was deliberately attempting to perform a tailplane stall recovery."
The NTSB was unable to determine why the first officer retracted the flaps and also suggested that the landing gear should be retracted. Her actions were also inconsistent with company stall recovery procedures and training. The actions of both pilots led to the aircraft entering an accelerated stall.
The method by which civil aircraft pilots can obtain their licenses was also criticized by the NTSB. The report was published on February 25, 2010.
The NTSB determined that in addition to Renslow's inadequate response to the stick shaker activation, there were key contributing factors. Primary among these were the flight crew's failure to monitor airspeed in relation to the rising position of the lowspeed cue and adhere to sterile cockpit procedures, Renslow's failure to effectively manage the flight, and Colgan Air’s inadequate airspeed selection and management procedures for approaches in icing conditions.
The board further found that: "The pilots' performance was likely impaired because of fatigue, but the extent of their impairment and the degree to which it contributed to the performance deficiencies that occurred during the flight cannot be conclusively determined." NTSB Chairman Deborah Hersman, while concurring, made it clear she considered fatigue a contributing factor. She compared the twenty years that fatigue has remained on the NTSB's Most Wanted List of Transportation Safety Improvements (without getting substantial action on the matter from regulators) to the changes in tolerance for alcohol over the same time period, noting that the performance impacts of fatigue and alcohol were similar.
Read more about this topic: Colgan Air Flight 3407