Thoughts from an observer
Please forgive in advance intrusion by ex-amateur pilot and frequent passenger.
Two things stand out for me after having followed this discussion from the beginning. The first is that no pilot of the experience of the AF crew would willingly end up in the weather situation they were in. And yet they did. That suggests either a failure of the weather radar, a failure of interpretation, or a weakness in the pilot/radar software interface (which could lead to a failure of interpretation of the weather radar). Other similarly-experienced crews were successfully dealing with the same weather system that night. So what was different about AF 447?
It is striking that crews in this part of the ITCZ have no real-time tools to deal with convective activity other than the onboard radar. PIREPs seem few and far between (and hampered by problems with HF), there’s no ground-based radar of any kind (ATC or dedicated weather radar) to assist, and there appears no current way of getting any detailed information from the weather satellites, which at least would provide a bigger picture of what’s behind whatever the onboard radar is painting. It appears that, by comparison with crews dealing with the same kind of weather in a continental context, crews traversing the ITCZ over the oceans are flying with the equivalent of at least one eye shut.
The other thing I found striking is trying to imagine what that cockpit must have been like with all the alarms going off, trying to make sense of multiple failures happening virtually simultaneously, trying to reset computer systems, and all while trying to hand-fly an aircraft to maintain a very delicate balance between stalling and overspeed without accurate airspeed information – while in the kind of turbulence that’s not a common experience even for the most experienced crews – and at night.
Not getting accurate airspeed information is bad enough in that situation, and would have presented a very difficult challenge all by itself (harder, perhaps, that the one the AeroPeru crew faced – at least they didn’t have to deal with a very narrow margin between stall and overspeed). But having to be computer operators at the same time?
That’s not to say that a Boeing crew would have had things any easier. But there are downsides to automation in any context. Airbus normal law would undoubtedly have saved the Buffalo Dash-8 crew from stalling. But being surrounded by computers, all screaming failure at the same time, couldn’t have helped the AF crew deal with what was already a horrible situation.