aterpster, don't confuse probability with overall accuracy. GPS navigation is perfectly fine for terrain avoidance as long as one gives a wide-enough margin to allow for the inherent inaccuracy. One does not buzz a peak in IMC with a 50ft clearance if one can avoid it. On the other hand, that 50 feet will make the difference between green and grey on a landing.

I use GPS waypoints in the Sahara regularly, and it is common occurrence to find marked points be off the actual position by 5-10 metres on a return years later, and that is without the altitude component.

That being said, me too would welcome a return to the thread topic.

Whatever the accuracy of the RNP-AR, it is still a requirement to fly the part below minimum manual AND be visual with the runway(environment). If not, Go Around.
The approach is designed to be safe and get you to a position near the runway if you stay within the designed limits.
If the .3 or .15 allowable deveation does not bring you to an acceptable, but safe, position to land, then don't land.
If it looks good at MDA continue, but keep in mind that any weatherdeterioration after that also mandates a Go-Around.

andasz ; I too welcomed a return to topic but my post was removed. Mods like this stuff here rather than in the Tech Log pages. All of us know what happened at KTM, it was always waiting to happen. It will happen again. No amount of willy waving will make a cloudbreak procedure a safe Cat111.

sleeper:

Good point and perhaps "on point" for this accident. But, keep in mind this VNKT RNP AR approach, and all RNP AR approaches have a DA, not an MDA.

Gordomac:

I don't think all of us know what happened at KTM. No doubt RNP AR is not a precision approach for purposes of landing. Such procedures are better than, "cloud break" (I like that colorful phrase) but they are not non-precision IAPs, rather more like LNAV/VNAV at DA.

Perhaps this is a too strong statement to agree with, let's just say neither the airline nor the airfield come as a complete surprise in the context of this accident. I think the only ones who really know what happened are the ones who were sitting in the front row, and they are still busy figuring out exactly what to remember.

The whole discussion of RNAV RNP is pointless as the end of the approach is VISUAL, because it is a NON-PRECISION APPROACH.

If indeed they were flying an NPA, they reached MDA, they made the call to land, and the rest is history.

As for the rest of it...I'm waiting for the official report as always.

Aterpster,
DA/MDA same principle. Below it is visual and manual.

Good point Mach!
Exactly what I thought when I read they touched with one gear outside the runway.
This one is shouting visual illusions...

Well there ya go! Here's me thinking that all the (Cat 1) ILS's I've flown also had a visual bit at the end, just like all the NPAs I've flown, and just like this RNP-AR! How wrong I was. No need to look out the window to land visually at all, apparently, given an ILS is a PRECISION approach.

The fact of the matter is that, within a bulls-roar, this RNP-AR would have placed the aircraft in exactly the same spot (horizontally and vertically) as if it had been an ILS. The actions required by the crew at the Decision Altitude (yes, on this non-ILS approach) would have been exactly the same.

@Capn Bloggs: You're right. NPA isn't the only type of approach where you land visual at the end. Whilst your sarcasm is duly noted, I hope the essence of my post wasn't lost upon you.

Allow me to re-phrase: regardless of how the aircraft got to MDA, the pilots decided to land. Let's wait for the hopefully accurate analysis of the FDR/CVR for the rest.

Heard from someone on the plane that it took almost 10 minutes before they opened the doors to let them out

sleeper:

Yes indeed. Except at DA you are in the slot, so to speak, and presumably the aircraft is in final configuration for landing. No so on approaches with MDA.

Semantics again. For both you need enough visibility to make a safe landing.
And yes for a nonprecision it is more.
If some-one lands on the runway-edge lights he/she clearly did not see enough.

Last technical stuff on RNP
 

For those that want to learn something here : just back from a meeting with some experts on RNP present . (i.e. EGNOS and Airbus ) I asked a few questions and now know a little bit more about RNP APPs and this is how it works ( on Airbus aircraft at least ) in simple terms.

The 0.3nm accuracy is only for the last part of the App segment not for the whole APP . 2 minutes before FAF , the RAIM box inside the aircraft will verify if all the satellites are aligned and if the position they received is correct and will be guaranteed for the next 5 minutes . The guaranteed precision you will get is still 0.3,NM either side of track , but in practice it is better . The correct definition is “ The system will guarantee that in the next 5 minutes your deviation will not be superior to your RNP value “

If the RAIM cannot guarantee the integrity it will issue a warning that should result in a go around (if IMC ),but this accuracy is only guaranteed for 95 % of the time.( although in practice it is better ) Only with an augmentation system ( such as WAAS or EGNOS ) you will get more than 99% . But Never 100%, mainly because of solar activity . 2014 had a very strong solar peak activity and the lowest was 99,4 . Better availability is expected in 2015.
EGNOS is already available today for regional’s and business jets , for large airliners the first aircraft fully factory equipped is the A350 . Retrofit for older types is expensive. But remember EGNOS is only for Europe .

What is more interesting is he obligation to follow certain rules to actually perform an RNAV-AR APP :you need certified aircraft, certified and trained crews , a certified airport and trained air traffic controllers . It is not because it is in the aircraft data base that you can use it . And there lies our first problem.
.
The other is when an airport changes things in the ground or in the runway and do not tell the people that make the data bases . Remember that one .

Next : some airports have issued visual RNAV APP plates on their own and there are currently no official procedures for them , hence the (correct) IFALPA warning.

Finally all strongly ay that RNP is not and never will be a substitute for ILS,. It is still a non precision APP and should be treated as such.
Now I put my willy back where it belongs and leave the discussion over to the "real" experts here.

Thank you for taking the time to post that. Interesting.

Thanks ATC Watcher. Interesting. I think the first thing that needs to happen is understanding the changing, blurring concepts of approaches. In the old days, we had ILS (THE precision approach) and all the others (NPAs). That's changed because now some GNSS approaches are also classed as precision approaches eg GLS.

The new descriptions are 2D and 3D approaches; 2D being lateral guidance only (VORs, NDBs, RNP APCH eg RNAV (GNSS) ) and 3D, being lateral and vertical guidance eg ILS, MLS, GLS, RNP APCH LNAV/VNAV, LPV and RNP-AR.

As I mentioned before, for all intents and purposes, an RNP-AR is virtually a precision approach because you are required to follow the vertical guidance/path, just as you have to on an ILS. It puts you at 300ft on final 1nm out. A GLS puts you on final at 200ft and 2/3nm out. We should therefore be talking about 2D and 3D approaches now, not Non-Precision and Precision.

As for the idea that a crew would try to use an RNP-AR when they hadn't been trained or approved for it, that equates to gross negligence on the part of the company for not ensuring that it's pilots knew that, as well as the regulator for letting it happen (and of course the crew for being so ignorant).

last position on the plane
 

You can find new picture from the Airbus 330
“GÖBEKLÝTEPE” BU HALE GELDÝ

Expensive fix ? Will she even fly out ?

Less than a year old so more than likely she's worth saving