Automation Mode Change and an Unstable Approach in a Wellington Nor'wester
Two-crew Dash 8 Q300, Wellington, runway 34, nor'wester 30 kt gusting 45. RNAV approach. During descent, ATC instructs 'descend now to 2,000 ft'. The captain selects a higher rate of descent and the autopilot changes from VNAV path to vertical speed mode. The FMA change is small and neither pilot calls it. At 1,000 ft AGL the FO notices the aircraft is 300 ft below the approach path and sinking at 1,300 ft/min; the captain says 'I have the visual, continue'. At 500 ft AGL the sink is 1,500 ft/min and the speed is 25 kt above target. Company stabilised-approach criteria require the aircraft to be on path and within +10/−5 kt by 500 ft.
1. Identify the threats, the errors, and the automation issue in this sequence, and explain how the mode change contributed to the deviation.
Read the worked answer
Threats: a strong gusty nor'wester, a short runway, and an ATC amended descent that compressed the crew's workload at a busy phase. Errors: the FMA mode change was not called or cross-checked; both pilots became absorbed in acquiring the visual picture; and the approach was continued despite being outside the stabilised criteria. The captain's 'I have the visual, continue' is plan-continuation bias.
The automation issue is mode confusion. In VNAV path the aircraft follows a vertical profile; in vertical speed mode it simply chases the selected rate, with no protection of the path. When ATC asked for a rapid descent, the selected rate was high, and the mode change left the aircraft descending below the path. The only cue was the FMA annunciation, which is small and easy to miss. Mode changes should be announced and confirmed by both crew; here the monitoring loop failed because attention had moved outside to the visual scene before the aircraft was actually stable.
2. What should the FO do and say at 1,000 ft, and why is the wording important?
Read the worked answer
The FO should call the deviation immediately, specifically and with the required action, not a question: 'Go around - we are below path, sink 1,300 ft per minute, speed plus 25.' If the captain does not respond, it should escalate: 'Captain, I am calling go-around' and, if the flight path is unsafe, 'I have control.' The call should be made at 1,000 ft, not 500 ft; the earlier the intervention, the more options remain.
Wording matters because vague hints ('are you happy?', 'we're a bit low') can be answered with 'continue' and do not transfer the seriousness of the situation. A clear statement of the observed deviation, the standard being breached and the required action removes ambiguity and gives the captain a decision point. The company's stabilised-approach criteria are the objective reference; the FO is not expressing an opinion but stating that a gate has been failed.
3. Beyond this crew's actions, which organisational defences should prevent a repeat, and how would you change the approach briefing?
Read the worked answer
Organisational defences: a stabilised-approach policy with a hard gate and a no-blame go-around culture; CRM training that rehearses assertive intervention and the 'stop' call; automation training covering FMA callouts at every mode change; and a procedure for responding to 'descend now' requests - for example, asking ATC for a delay, requesting a revised clearance, or briefing that the request may invalidate the profile. Flight data monitoring can identify the mode-change pattern if the operator has it.
The approach briefing should include the expected automation modes and the points at which they will change, the FMA callouts, the stabilised-approach gates, the go-around triggers and who calls them, and the plan for a late or amended descent clearance. It should be interactive, with the FO confirming the plan, so both crew share the same mental model before the high-workload phase begins.