Doors to Automatic;
If crews are used to crossing the threshold, closing the throttles within 2-3 seconds and touching down a second or two later then surely they must have known something was a miss. 4000ft would equate to 20 seconds or so after crossing the threshold.
Exactly right.
To your excellent point:
Touchdown from an altitude of 50' above the runway, (our FOQA event uses time from 30'), nominally occurs within 3 to 5 seconds. All other factors being equal, (threshold crossing height, approach speed, rate of descent all nominal), a touchdown longer than six seconds from 50' indicates the landing is "going long". Timings greater than about 10 seconds correlate to a touchdown beyond the TDZ and heighten the risk of an over-run accident.
While a number of factors likely were in play in this accident such as a higher groundspeed through the contributions of a tailwind and a slightly higher TAS due to slightly higher-than-standard SAT and possibly higher airspeed than Vref +5, varying visibility in heavy rainshowers (+SHRA), night-time with possible "black hole" phenomenon in the transition from "inside to outside" leading to lack of depth perception and subsequent long (floated) landing, wet, possibly flooded runway, possible late/no spoiler deployment, choice of landing flap and possibly other, less likely factors, FOQA data still has a role to play in examining the historical record of the fleet type's approach and landing performance.
Examining it from a safety investigation pov, if the record indicates that the performance of this aircraft during the approach and landing rollout was a relative outlier, a different causal chain and airline response may be indicated.
If on the other hand the FOQA data indicates that long landings with higher than normal approach speeds with reduced flap settings and idle thrust, (not the case here, we must assume!), both of which are being widely implemented as fuel saving techniques, then examination of this accident will necessarily take a different approach. The ATSB Report of the QANTAS B744 over-run accident at Bangkok made just such an observation. We will have to wait for the final report before we know which is the case.
Along with ASAP, (which has been reinstated I believe, after the APA withdrew their cooperation/participation, accusing AA management of using such safety information to go after individual pilots - anyone?), American also runs a FOQA Program. Such a Program will have a number of landing events which would monitor approach and landing performance.
Depending upon available parameters for the NG, (like the 320, they should number in the thousands), LFL design/installation and other FOQA Program features and protocols, such events, with associated levels of severity, may include:
Glideslope & Localizer deviation below 500',
Heading Changes Below 500',
Rate of Descent Below 500',
Airspeed Deviation from Vref+5 at 50', (greater than 20kts),
Thrust/N1 levels near/at Idle thrust,
Landing distance from 50' AGL,
Time from 30' to touchdown,
Time from main gear compression to nosewheel compression
High Vertical 'g' at touchdown
High Longitudinal 'g' during rollout
High Lateral 'g' at touchdown
Heading Changes During Rollout
Excessive Aileron, Rudder Deflection During Rollout
Non-standard/Reduced Flap Setting for Landing
and other events, again, depending upon the FOQA Program design and how interested the airline is in actually knowing what its fleet is doing on a daily basis. There are carriers around who don't want to know, mainly to avoid the higher costs of "knowing", a point raised numerous times in other threads.
Each of these events would have nominal performance numbers, the exceedence of which would produce a FOQA Event, the greater the exceedence, the greater the severity of the event.
This information would be trended, the outliers examined and so on. Where the outliers are signficant, a call to the crew to further understand the event can be made.
So, for example, if the 737NG fleet was showing a trend to long landings including long hold-off times greater than the times you mention, (eight seconds is about the upper limit from 50' to t/d and will likely produce touchdowns outside the TDZ which is normally the first 3000' for 1/3 of the runway whichever is less, but practically speaking anything longer than beyond the first 2000'), then questions will arise about what was done, if anything, about the trends in terms of flight crew awareness programs, during recurrent training highlighting the importance of landing within the TDZ, etc.
If the landing was indeed as reported, about 4000ft past the threshold, the question clearly becomes why and was it inevitable and if so, from what point might it have been rescued. Most pilots are reluctant to initiate a go-around after touchdown and before reverse has been selected but it can be done, (certainly NOT after reverse is selected, the associated risk being far higher than the risk of a lower-speed over-run). Both the aircraft's DFDR/CVR and the broader "upstream" safety tools described herein will assist in answering this question. This is the second serious overrun accident of this specific kind in a decade for American so establishing these facts,
including an examination of the operational culture, is important to either rule out or deal with any related factors.
PJ2