an undulating runway data question

[sightseer]

I think Fr Bill's explanation (in the Feb 20 development video thread pg 21 Feb 21 2020 3:48pm) that 'runways are created from available data and if the data shows undulation then so will that runway in the sim' is the best and simplest explanation and answer to that particular question but it leads me to wonder about that data.

How much change in elevation along a runway is needed before it is included in the data?

I could ask where this data is and it is publicly available but I personally am not interested in those answers yet.  I know that airport charts list threshhold elevations but Im not a real pilot and havent even looked at an airport chart in years so I dont know how more refined the runway data is or can be.

So the question of how much change is needed...what is the resolution of this data and does it differ from location to location?

OK...question(s):  will a bush strip suffer from XPlanes problem where there is so much undulation that your Cessna explodes due to the rough landing or will they just be slopes?

Is this dataset editable by people who care and then available to all or are we still where we are with FSX where only 3PD make proper bush strips?

maybe this is too many questions that are unanswerable at this time but I thought I'd check.

 

[sightseer]

Thanks for that.  I had forgotten about game/sim resolution.

but about pilots --  How would a pilot know that an airport he or she had never been to might have a dip or hump that they would have to be aware of?  this data must be available...should be available...

if it is available then what is its resolution? 10 feet? 1 foot?  I can't imagine it would need to be a resolution that would affect the game performance too much but maybe so.

I also can't imagine that this data would be available for some small bush strip somewhere which means 3PDs will still have much to do.

[Lorby_SI]

35 minutes ago, sightseer said:

Thanks for that.  I had forgotten about game/sim resolution.

but about pilots --  How would a pilot know that an airport he or she had never been to might have a dip or hump that they would have to be aware of?  this data must be available...should be available...

if it is available then what is its resolution? 10 feet? 1 foot?  I can't imagine it would need to be a resolution that would affect the game performance too much but maybe so.

I also can't imagine that this data would be available for some small bush strip somewhere which means 3PDs will still have much to do.

As far as the entire globe is concerned, I suggest searching for DTED, DEM and HRTI.

The SRTM project mapped the earth in approx. 30m resolution. There are newer (private) satellites, mapping the earth in "pixels" of 12x12 meters I believe. I wouldn't expect data with that accuracy to be freely accessible though.

I think that the best/only information you can get about runways is on the plates or in the aerodrome manual for the airport, which will contain something like this (EGCC, since Chock mentioned it):

Quote

AERODROME ELEVATION AND GEOID UNDULATION  
 
 Lat: 532114N Long: 0021630W  Elevation: 257 ft   Mid point of Runway 05L/23R.  Geoid Undulation 167 ft  

3.1. Elevation of Each Threshold and Geoid Undulation  
 
 Elevation Threshold of Runway 05L –212ft  Elevation Threshold of Runway 23R – 249ft  Elevation Threshold of Runway 05R- 186ft  Elevation Threshold of Runway 23L – 227ft 
 
3.2. Elevation of the Runway ends  
 
 Runway 05L end  Elevation 249ft   Runway 23R end Elevation 200ft  Runway 05R end Elevation 235ft  Runway 23L end Elevation 186ft 
 
3.3. Significant High and Low Points along the Runway  
 
Runway 05L/23R has a high point, elevation 257ft situated beam Link Golf. 

Best regards

[Mace]

58 minutes ago, sightseer said:

 

I also can't imagine that this data would be available for some small bush strip somewhere which means 3PDs will still have much to do.

Yes they will.  I saw it posted in THIS subforum the other day, that there will be no need for scenery designers anymore.  I saw a lot of things in the "airports" vid that tell me they will very much be needed. 

And as for elevation data, yes, you scrounge up what you can -- photos showing aprons/taxiways/runways, first-hand accounts from guys like Alan (Chock) etc. etc. and for the parts of the airport in-between known elevation points, you might have to take artistic license as Rob mentions.  Plus I can see taking some DEM data and working that in.  I want to see what sort of tool they have for small-scale mesh editing.  It can't just be platforms.

[sightseer]

Thank ya'll very much.  very helpful indeed.  I admit that I was one of those who thought 3PD airport developers wouldnt be needed nearly as much but in reality they'll have even more to do with people flying to places they never would have gone in the past due to surrounding scenery (lack thereof)

I still need to find out what DAPP is... if anyone knows?  deferred approach?  it was mentioned by Jeff Nielsen in reference to undulating runways and wet weather.

I really hope MS/Asobo decide to do something like XPlane has with the gateway.

[Chock]

3 hours ago, sightseer said:

How would a pilot know that an airport he or she had never been to might have a dip or hump that they would have to be aware of?  this data must be available...should be available...

if it is available then what is its resolution? 10 feet? 1 foot?  I can't imagine it would need to be a resolution that would affect the game performance too much but maybe so.

I also can't imagine that this data would be available for some small bush strip somewhere which means 3PDs will still have much to do.

With major airports, they publish a booklet called the: 'Aerodrome Booklet with MINIMA'. This usually has charts for all the SIDs and STARs, taxiway and airfield maps, pages about various procedures and rules for the apron and runway, and amongst all that, if there is any significant information which is deemed necessary for the safe operation of aircraft, such as a significant dip in the terrain of the runway for example, then it will mentioned in the runway operations pages.

These days, such books are usually available in PDF form from the Airport's website too. Since these books go out of date fairly quickly, this is probably the best way to get them rather than buying the printed versions. But, what that does mean is that at airports which do still sell printed versions, you can often find the out of date Aerodrome Booklets in those airport shops which sell model planes and scanners and all that sort of thing.

Here is a quick snap I took of a page from an old (May 2005 - Issue No 122) edition of EGCC's Aerodrome Booklet. You can tell it is old because it refers to Runway 06/24, which it hasn't been designated as for a number of years owing to magnetic shift changing its headings to 05/23. However, you can see at the bottom of the page that it has cautionary notes concerning varying terrain elevations as you approach the threshold of the runway over the Bollin Valley:

 

[jpc55]

The first thing that pop into my mind was NOTAMs.  

[Noodle]

Many European airports have vertical obstruction charts in the AIP which include the runway elevation profile.

I'd post a screengrab, but uploading images seems to be broken on mobile.

FAA should have a similar product through their GIS portal.

[Dominique_K]

Runways do not undulate, the underlying  terrain does. Having chart technical references of the runway is not enough. The said runway is stuck in a digital elevation map. If the latter is less precise than the runway specs, it is going to be a bit awkward. We have a taste of that (in reverse) with trenches and plateaus in P3D when the  elevation data are more precise than the runway specifications.

To have undulating runways everywhere, we need to have DEM at less than a 5m sampling for everywhere in the world. I might be wrong but that does not exist, to my knowledge. For civil use at least. 

So the solution is to manually terraform according to photos. I don’t know of any developer who is very fond of that in the present state of affairs (P3D).

 

[RALF9636]

20 minutes ago, domkle said:

So the solution is to manually terraform according to photos. 

 

As far as I understood the latest video they already do a manual terraforming on all the 37k airports. He said the runway slope is applied from chart references. To my understanding that means that at least the two runway thresholds (their elevation is included in all charts) will be at the correct elevation.

That already goes beyond the resolution of the underlying mesh, thus requires manual terraforming.

Anf if they can do it with two points, I see no reason why they or 3PD shouldn't basically be able to add more reference points. It might get more complex, but in principle it should be the same procedure.

 

Edit: Actually he said the runway slope is created automatically by their algorithm from the chart references (at 4:05). So by adding more reference points, there might also be an algorithm that creates an undulation.

[Dominique_K]

10 minutes ago, RALF9636 said:

As Matthew (in his inimitable way 😋) rightly pointed out in another thread undulations (humps and vertical final curvatures) and slopes are two different things. 

As far as I understood the latest video they already do a manual terraforming on all the 37k airports. He said the runway slope is applied from chart references. To my understanding that means that at least the two runway thresholds (their elevation is included in all charts) will be at the correct elevation.  

That already goes beyond the resolution of the underlying mesh, thus requires manual terraforming.

I am not sure that I follow you here. It is not beyond the mesh resolution, it is inferior to the precision of a mesh. The two thresholds are hundred of meters apart, far more that a mesh sampling.  

Anf if they can do it with two points, I see no reason why they or 3PD shouldn't basically be able to add more reference points. It might get more complex, but in principle it should be the same procedure.

I have no idea how FS20 processes the terrain so it is difficult to answer.I've done a little terraforming (mesh manipulation) in P3D. It is not a piece of cake, I'll tell you, specially when you have only visual references.

Edit: Actually he said the runway slope is created automatically by their algorithm from the chart references (at 4:05). So by adding more reference points, there might also be an algorithm that creates an undulation.

Where can the reference points come from if they don't exist in the charts but from the surrounding mesh ?

 

[RALF9636]

29 minutes ago, domkle said:

 

What I meant was the lateral position of each threshold will be pretty much exact, much more than the mesh resolution. So the elevation of the thresholds cannot be taken from the mesh (because it would be an interpolation).

 

But anyway, let's move on. It's all speculation. Let's wait and see what they can do.

[Mark1616]

No matter what resolution, polygons will never result in a smooth ride. The airplane's ground model must handle curvatures. I expect that runways will be internally represented using some sort of polynomial spline fitting the underlying terrain, and then the curvature is exactly defined at any point of the runway. Only the visual representation must use polygons.

The Train simulator, for example, works in this way, i.e. the curved rails are graphically representated with very short lines, while the physcal model of the train interacting with the rails uses curvatures and like this achieves a constant centripetal force acting on the train.

[Mace]

10 hours ago, domkle said:

piece of cake

You know this expression "piece of cake"!   Does it exist in French?

I too have played around with editing DEM data and you're right, it's definitely no "piece of cake".  That's what I meant earlier in this thread when I wondered about having good tools to do this.  I can see something like that not being in the SDK, since probably most people will just want to put in some parking and be done with it -- but it should be in the SDK.  We don't just want to be able to manipulate the mask polys (like they showed in the video) -- that is good, and necessary, but we also need a way to (relatively) easily adjust the mesh.

[Dominique_K]

3 minutes ago, Mace said:

You know this expression "piece of cake"!   Does it exist in French?

I too have played around with editing DEM data and you're right, it's definitely no "piece of cake".  That's what I meant earlier in this thread when I wondered about having good tools to do this.  I can see something like that not being in the SDK, since probably most people will just want to put in some parking and be done with it -- but it should be in the SDK.  We don't just want to be able to manipulate the mask polys (like they showed in the video) -- that is good, and necessary, but we also need a way to (relatively) easily adjust the mesh.

Actually we say indifferently "C'est pas du gâteau" (gâteau=cake) or 'c'est pas d'la tarte" (tarte=pie)