Low vis over water
 

Good decision for a go around. Low VI's on a non precision approach with both pilots looking for the runway, along with a lack of depth perception over the water with little horizon, it's a rather classic setup for such a mistake. Any of us who have spent time over water have fought this, especially in our young single pilot days. Same issues as black hole approaches. Makes me think of Kalitta at Gitmo.

Hotel Tango, well here is your chance to shine!
Quite.
So why does it keep happening? Still seems like a reasonable question.

I think we have flogged this one to death with a big stick.

There again, most pilots have been trained to fly at 50ft or lower, for hours on end..
.

You do that at MDA. If you can’t see it you go-around. If you lose sight of it later on, you go-around. What you don’t do is carry on lower and lower *hoping* to see something: many have died doing this, which is why we have minima.

It would be interesting to know what guidance the PF was following before they went around that got them in the position they ended up in...

16024 So, you're suggesting that "it keeps happening" at SXM do you? Enlighten us further by all means. Do you have any factual knowledge about this airport?

As for recorded accidents, there was an "arriving" accident at SXM in 1971 and that was a DC-3 which crashed 20 miles short due to fuel starvation. In 1972 an DHC-6 crashed on a night approach. In 2014 a Shorts 360 crashed on departure.

Looking at its safety record I would say it is no more an accident waiting to happen than hundreds of other airports around the world.

If the photo is real, then altimetry error would explain the low altitude and the ADSB reports of being on profile as the ADSB would report GPS 2d position and baro altitude. Given how many non precision approach accidents have been caused by misset altimeters, it seems a plausible explanation, especially if the local QNH was below standard.

16024 posted:
Hotel Tango replied:
Guys, is it just me?
I was just asking.
For perspective last 2 flights for me were NPA into coastal runways, and one of those requires a 45 degree turn after break-off to line up, and the other one has PAPIs that are known to be as good as useless. Neither airport makes it into the hysterical "scariest landings" posted all over the interweb.

Ok, I tried.

Alum shuffler:

With all the publicity this incident has received, it's likely that the experts at Transport Canada will take a good look at the DFDR.

I thought the baro altitude is referenced to standard then the adjustment made for QNH elsewhere, like a transponder. Stand to be corrected...

I'm puzzled. Surely they had some indication of distance-to-run, even if they just set an extended centreline manually several miles from the threshold. Altimetry error should be discounted on this one, since a Rad Alt readout will give a good clue, It's a sea-level airfield

Correct.

In this case, add approximately 150' to the Mode S/ADS-B altitudes to get true AMSL values.

I share your suspicion about a misset altimeter but it appears that the QNH was 1019 hPa according to a weather sequence attached to one of the YouTube videos.

From the approach plate it lists the MDA (H) (CONDITIONAL) as 500 feet. It's one of those 'find the faces in the picture' exercises they used to love in airline interviews but it looks to me like you can go to 500 feet if you can stay at 205 knots or below on the miss until you are on the 180 degree track. It's note 1 on the bottom, not to be confused with note 1 on the top of the plate about DME required.

If you can't stay at or below 205 knots in the turn, your MDA is 770 feet it appears to me. I'm guessing the B-738 on a straight in approach is Cat C for the right side mins table.

I agree that it seems that the 2.98 degree path puts you higher than 500 feet at the 2 DME fix. I get about 601 feet for the charted 1.9 nm to the threshold, plus 50 feet for the TCH and 14 feet elevation for about 665 feet at D2.0.

So, you would indeed go missed at the D2.0 [MA10] point on a CDA before reaching 500 feet. I haven't trained for or done a dive and drive non-precision approach for many years and I thought they were pretty much extinct in airliner ops by now.

But in the past, there were options to do a CDA on a path in VNAV, speed intervene (to avoid throttle surge on some Boeings), a CDA on a path with V/S or a dive to mins early, level off and then land or go missed at the MAP depending on what you saw. On the CDA's you would go missed at MDA without visual contact and not level off even if you hadn't reached the charted MAP.

Is it possible that the WestJet crew saw that they could go just a little lower with an early descent below path to 500 feet before D2.0 and lost track of the altitude before breaking out so low? Does the B-738 give an audio callout at minimums on path that they were expecting but may not have got due to the setup in the box? Obviously, altitude needs to be closely monitored whether the automatic callouts work or not.

Depends on customer specification.

I would really like to know how they got themselves to almost sea level short of the runway so I can avoid the trap they fell into next time I’m in that position.

Giving the benefit of the doubt and assuming that they didn’t just bust minima and get woken up by the 100R call, what could have led to this incident? Visual at MDA but going slightly under then a confusing picture? Following an inappropriately programmed flight director in a visual segment? There’s no extended approach lighting so you’ve got to be able to (continuously) see the runway and/or slope guidance in order to continue.

What was so convincing that it almost put them in the sea following it?

Was there a Windshear event? Microburst event?

Wait for the FDR amd official report

Airbubba, misset alts is the most plausible explanation to cover the actual and ADSB profiles. My comment about lower than standard pressure would have made the error even easier to make, simply forgetting to select QNH at transition, but the error can still occur from a misheard or mal-transmitted QNH. Whatever the cause for altimeter error, it's still the leading cause for non-precision app accidents and fits well here as something to be investigated. It may, of course, have nothing to do with this event...

Airbubba
That would put the aircraft 180ft lower than the 500ft MDH so 320ft. Is that enough error to show in that picture?

Note: The number of altimeter setting errors reported to ASRS is extremely high.

So, you would indeed go missed at the D2.0 [MA10] point on a CDA before reaching 500 feet.

Interesting. Because all CDA SOP's I've used are to GA at MDA if no contact. In Europe I've not come across this situation where 'M' is reached before MDA. From passed Ops 'M' was the decision point on a Dive & Drive. True about Cat C for B737/8. If you didn't use 500' MDA you would be in real trouble at 'M'. If the vis was only 3500m. i.e. coordinated with 500' & 1.9nm you would be very high if stopping at 770' and unstable to land.
The comment about the RA shouting at the crew about their true height is valid. I wonder if Westjet set MAA after leaving platform alt, or set 500' on MCP. If the latter the A/P or FD would level off. If MAA was set then the "+100" or "approaching minimums" & "minimums" would also shout at the crew. They would be looking out of the window. I wonder if much earlier they had been sucked into "sea contact" vertically, i.e. knowing there was nothing to hit except a ship's mast. Were they manual or A/P CMD?
Back to my earlier question: I would be interested to know what they did differently the 2nd time.
Regarding the 'photo shopping' of the first photo; why would she? Why take the 1st & 2nd photos, doctor the 1st, publish them with accusations and set of a manure storm that can easily be answered by the Canadian CAA. They have the crew and technicians to solve that problem should it be true. After all, they did make a GA.

Possibly, though you’d have thought they’d have said something afterwards and/or performed the WEM at the time...
Fair enough but we don’t know if it was reported.

Yesterday, here in YYC where West Jet is based, the local media aired the video of the infamous approach. West Jet finally acknowledged that the approach was in fact too low. They claimed that due to rapidly changing weather conditions on the approach, the crew defended below the glide path but quickly corrected and executed a missed approach as they are so thoroughly trained to do! Of course they downplayed the entire event, and suggested that the event will further enhance safety at WJ as the event will be reviewed and shared amongst crews. I'm assuming that this will be done through their SMS, but herein lies the disparity. Realizing that an airline has to maintain a positive public image, they chose to ignore the event or downplay its seriousness. The premise of an SMS is full disclosure. Come clean with what occurred, identify the weaknesses, develop changes to procedures so it won't re occur and share the process. ( at least that's how TC here in Canada wants the system to work). When the company denies the significance of an occurrence does it not promote a culture within to not embrace SMS. Here in Canada, that means back to fulll regulatory enforcement, and disciplinary action against the airline, crews and management.

It seems like a bigger deal is being made of this than needed. They were low, yes, but it seems like a matter of getting visual with the runway then either gusting winds/windshear or illusions / misjudgement due to the weather. But the key thing is that they realized their mistake, admitted it, and went around as they should have.

No one is free from making mistakes, the key is to recognize and react. Seems like they did that here.

About the descending below MDA, they likely did so and did so correctly: being visual with the runway and continuing as normal. We have a clear view of the aircraft in the photos and videos from various vantage points. While there was some weather and rain, the pilots must have been visual with the runway and lights.

VDP distance in St.Maarten
 

Some interesting points about the VOR approach into SXM.

If descending to an MDA of 770 feet the VDP distance is 2.2 DME.
If descending to an MDA of 500 feet the VDP distance is 1.5 DME.

So if WestJet was descending to 770 feet, the prevailing visiblity might not have enabled the runway to be visible.

Clearly though, the aircraft was below 770 feet on the go-around.

Probably descended below MDA, broke out below the 3 degree path, and executed a go-around.


[IMG]http://i779.photobucket.com/albums/y...pssq3fpqno.png[/IMG]
[IMG]http://i779.photobucket.com/albums/y...psxrbkngoq.png[/IMG]

If they forgot to set the local altimeter and left things at 1013 hPa, wouldn't they actually be higher than the 500 feet indicated baro altitude?

Those low transition levels (FL 65 in this case) are spring loaded to mess up. I've sure made the mistake myself but had the luxury of catching it on an ILS crossing altitude.

There may never be an official report is my guess. Is there some official Canada government site that might list this as an incident to be investigated? Or, will the Dutch authorities possibly take a look?

I agree that it looks odd, am I perhaps misreading that chart? Also, somebody check my math(s) on the D2.0 crossing height for a 2.98 degree path. I figured 1.9 nm equals 11,545 feet multiplied by the tangent of 2.98 degrees gives 601 feet of altitude change plus 64 feet for the 50 foot TCH on the chart. In the past, non-precision approach calculations did not include the TCH in some cases and you had to manually add it to the FMS to get a good advisory descent path.

Also, on a non-precision approach, even if a nice path is depicted in the FMS, it doesn't guarantee that path will continue to the touchdown zone in my experience. Years ago the LOC Rwy 27 at SAN (San Diego, California for that feller who couldn't figure out how to Google 'prosumer' ;)) had a nice path in the box to the runway threshold but since the actual touchdown zone was displaced, you had to add power and level off momentarily to get on the PAPI's for the rest of the descent.

The "baro altitude" that the OP is referring to is the data sent by the transponder (which is the only information we have access to at this time).

That data isn't affected by whether or not the crew set the correct QNH on the altimeter subscale.

I believe Ian W. was talking about the misset altimeter scenario causing a faulty indication in the cockpit from the context of his quote of my earlier post.

See: http://www.pprune.org/9706787-post99.html

It's easy to get busy down low and miss the transition level, especially when it's below FL100 dodging weather in the Caribbean. And, with the accents, 1019 might certainly be heard as 1009.

A misset altimeter is definitely one of the things that can ruin your whole day on a non-precision approach down to minimums.

Another thing that can get you on the NP approach is a bad setup in the FMS giving you a path that is not the right one. Still, it shouldn't get you below minimums without the runway in sight if you do the call outs and adhere to stable approach criteria.

Christine Negroni has a picture of the UPS BHM A300 crash in her article linked on the first post on this thread. In that mishap, the crew had an incorrect FMC setup with a bogus path that was above them. For some reason, the captain V/S'ed them into the ground with 1500 feet per minute down at 1000 AGL, which should have been an automatic go around.

Descended?

Shurely shome mishtake...

What I find odd is that there are no DME/ALT checks after 1600' at 4.8nm. EFIS a/c do LNAV/VNAV descents to MDA. (the chart shown in #17 is not an RNAV approach. Is there one?) However, older, perhaps more private a/c are steam driven. This would invite, especially over the obstacle free sea, a Dive & Drive profile. Either way, the crew procedures should be a level off at 500' or 770' until 2nm and a decision if the runway is in sight yes/no. Then a land or GA decision.

I have to admit, if was in my (dream on) private twin prop on a murky day with a hot date waiting, I might be tempted to shave a bit off (if you get my drift) and nibble below the belt to see if I could get in. However, I would not like a crowd to join me, so I think in a B738 case I'd be respecting my master and his rules.

Rat, the "M" is the missed app point. It's not a great name for it. It's not the point at which you should commence a missed app; that should happen at the minima on the glide path. It is the earliest point that lateral navigation should switch from the approach path to the missed app path, preventing early turns on intercept headings and not relevant (and hence rarely shown) where the missed app tracks a radial or is flown straight ahead.

The minimums callout, automatic or from PNF, would, on a non-precision app or Cat 1 ILS be baro alt related, so again, incorrectly high subscale settings would create an excessively low callout. More evidence that this could be the cause, but, again, far from conclusive.

The 737NG does not have GPS related altimetry, though the side profile view on the NAV display on aircraft with this customer option will show a GPS derived profile to a database display of terrain and airfield elevation. While there is no warning generated, a close look at that profile wile show a mismatch between baro and GPS profile by having the runway floating above or buried under the terrain, depending on which direction the altimeter setting error occurs. The rad alt displays from 2500' on the PFD. Custome spec may have a number of automatic callouts. Any crosscheck against the baro alt while over the sea should have revealed misset altimeters if that was the case.

You may be right.

It's probably worth clarifying what we can deduce (based on the transponder data) to have been the minimum height to which the aircraft descended on its first approach (at approximately 1.1 nm from the threshold):

Baro height: 100' (to nearest 25')
Actual height AMSL: 250' (approx - based on baro height, adjusted for METAR QNH)
Altimeter indicated height: we don't know (but likely to have been close to one of the two above values unless a completely spurious subscale setting was used)

There is a new RNAV (GNSS) Rwy 10 posted earlier:

http://www.pprune.org/9702902-post48.html

What makes me think that WestJet shot the VOR Z is that the missed approach on the WS 2562 ADS-B track looks more like the 180 degree course of the published miss on the VOR approach rather than the direct ONBED missed approach on the RNAV plate. Also, with weather at minimums, MDA is 500 feet on the VOR approach, 700 feet for the RNAV.

Of course, the tower may have given them turn right heading 180, maintain 4000 feet after the RNAV approach.

A lot of altimeter discussion. VDP is approx 486'/1.6 nm. The picture is at approx 50' and .3-.5 nm. An altimeter setting has nothing to do with getting that low while in visual conditions.

Allegedly report of wind shear which is why there was a delay before subsequent approaches were attempted.

For all this talk about wrong altimeter setting...it's a sea level airport. The pressure differential (between std and actual) is not that high, and the altitude error resulting from a missed (with a D not a T) altimeter setting is way less than the altitude at which they ended up at.

Some other fellow posters have commented about some sort of distraction after breaking out at minima. This seems the explanation that satisfies me the most. Murky weather, breaking out at or near minima, both pilots looking out for the runway, PM forgetting to monitor instruments, altitude vs horizon difficult to assess due to weather plus it being a beach/sea airport certainly doesn't help with the optical illusions.

For the sake of the argument, yes they did what they're supposed to, i.e. go around, but I certainly would like to know how they ended up in such scenario, so we all (or me at least!) can avoid or at least identify such scenario before getting to such little error margins...

Oh and btw, to me at least, the microburst/windshear scenario doesn't quite work with me, given what I've seen both on video and also the metar I saw from one of the articles I've seen regarding this hmm, occurrence.

Regards, your Caribbean/Latin American jet pilot

Are you seriously suggesting that the difference between standard 1013.2 mb and on the day QNH 1019 mb (about 150') isn't significant on approach ?

misset: to set, adjust or calibrate something incorrectly :ugh:

Maybe someone can set up a approach to landing rating website.

Enjoy the link, pretty cool.

Maho Beach Cam - SXM Airport Cam on St Maarten

For the sake of the argument, yes they did what they're supposed to, i.e. go around,

Given that, I presume, they had 500' set in Baro and that there would have been GPWS call out at +100 & minimums, I wonder what caused them to make a GA so low? If they went below MDA, that's one thing. An altimeter mis-setting might be 150'; that still leaves them 350' above the waves, but as someone said the error would likely put them high. So what instigated the GA? Some guys say they did what they should have, correctly; therefore there is no case to answer. Hm? They should have done it much earlier!
So what else happened? They sure as heck didn't bust out at minimums, see the water so close and think this was not a good idea and then GA. Something happened between the Baro call out of "minimums" and the decision to GA. From the photo that seemed to be quite a longtime. Meanwhile the RA was reading very low. How do WJ fly an NPA? Was PM looking for the runway and PF inside? Were they flying manually and both pilots looking outside wth no-one 'managing the office'? Were they in V/S & automatics with both pilots looking outside?
In most of my Ops PF looks inside & PM looks outside nearing minimums. At minimums if PM does not see the visual reference and call visual PF makes a GA: never looks outside. No hesitation. Both pilots scratching around at minima, both looking outside, no-one managing the office, manual flight or V/S, going pass "minimums not visual" recipe for...............

Could be similar to Lion Air in Bali, where they didn't go around soon enough to avoid landing on the water........

Escape path, altimeter error is not limited to forgetting to set QNH on transition. They could have been passed/misheard an incorrect QNH and set 1029 instead of 1019, for example. I've had that happen a handful of times, and was fortunate to remember the pressure in the weather brief or have got a Volmet or ATIS that said something different, catching the mistake. Perhaps this crew didn't have the same luxury of a cross check. It fits with why they went below real altitude minima heads-in, why the ADSB shows no major deviation, and fits with the most common cause for such incidents. It's all circumstantial, but don't shoot the idea down without evidence to the contrary.

The ADS-B does show the deviation, as it is designed to do..

No matter what QNH you set the altimeter to, it has no effect on what's sent by the transponder.