Penrith to Lithgow

[Roland Ziegler]

Bruce,

I start a new thread for your route building plans.

Back to the DEM issue:

Tonight I've found some time to set up a spatial search rectangle for Penrith to Lithogow and examine the ASTER thumbnail images of all the available "granules" within or intersecting with it (total: 89, but many apparently useless due to cloud cover). I ended up in ordering a custom DEM for seven granules, with mostly 0% clouds, except for two with 3% clouds. The granules are overlapping, and the whole area should be covered. I expect reasonable results and usable DEMs.

Queueing for DEM production at the Data Center may take many weeks. They shall email me once processing has been completed. Total download volume for these DEMs will be around 150 MB. However, after merging/cutting them to the area you really need it should be much less.

I assume that by the day these DEMS will be delivered, TransDEM, my new DEM editing tool, will have enough useful functions to deal with that data conveniently.

[Bruce Kennewell]

Roland, I don't know how to thank you for going to this trouble!

That is a terrific thing to do and I just hope that, once the Zusi editors are translated, I have the understanding and capability to create the route.

Thank you very, very much for this.

Regards,
Bruce.

[Roland Ziegler]

No problem.

Keep in mind, though, that these DEMs will be "relative". Their elevations lack an absolute reference which we will have to add by providing "ground control points". One could have solved this by ordering "absolute" DEMs but then you need to provide those control points in quite a complicated way. My approach with TransDEM is much easier but will be less accurate.

So once the DEMs will be available we need some of those "ground control points" for TransDEM by picking a dozen locations and their elevations from topographic maps. These will then be applied to the relative DEM and will adjust the DEM to the absolute reference points.

To do that we will need the coordinates of these ground control points in a coordinate system that TransDEM can understand.

The coordinate system consists of two parts: The "grid" which could be lat/long or, for example, UTM, and a geodetic datum or map datum which identifies the ellipsoid the coordinate grid is based upon, e.g "WGS84".

Your topographic maps should state both coordinate grid and datum.

If your maps are not UTM and/or not WGS84 I need to know which grid and datum they use as TransDEM may then need support for an additional system. I will start with integration of "national grids" in the next "work package".

[Bruce Kennewell]

Roland, I don't have actual topographic maps , yet, for the route. The only references which I have are the Gov. Railway's track plans, gradients, curve data and profiles.

However, obtaining the correct maps is easy and I can do so after January 5th.

I do have a couple of other topo maps here (Western Australia) and and I am looking for 'WGS84' and 'UTM'

The maps are of scale 1:250000 and underneath the scale on one of them is "Contour interval 50 metres" and under that is "Blue numbered grid lines are 10,000 metre intervals of the universal transverse mercator grid, zone 50, Australian national spheroid".
No reference to WGS84 on that one.

The other map has some information under a heading "About This Map", which says:-
Projection....................Map Grid of Australia 1994 (MGA94) UTM on GRS80 ellipsoid.
Horizontal Datum.........Geocentric Datum of Australia 1994 (GDA94) equivalent to WGS84.
Vertical Datum............Australian Height Datum (AHD)
Horizontal Accuracy...... +/- 140 metres
Vertical Accuracy......... +/- 25 metres
Elevation in metres; 50 metre contour interval.

As these maps both came from AUSLIG (Australia's national mapping authority) I am assuming that those for the Blue Mts. will also fall into the same category.

Does all this make some sense to you?!

Regards,
Bruce.

[Roland Ziegler]

Bruce, it does make sense :

universal transverse mercator grid, zone 50

That is UTM. Zone 50 is one of the 60 zones of 6° longitude each which is used by the transverse mercator projection to unwrap Mother Earth to cylinder surfaces. We're here in zone 31, 32 and 33. Zusi routes should lie within one zone only, but that can easily be accomplished by transformation.

Australian national spheroid

which right now is unknown to me. But it shall not be problem to get the parameters for this.

Projection....................Map Grid of Australia 1994 (MGA94) UTM on GRS80 ellipsoid.
Horizontal Datum.........Geocentric Datum of Australia 1994 (GDA94) equivalent to WGS84.

If I understand this correctly, it is more or less UTM on WGS84. I will look up the differences between GRS80, GDA94 and WGS84.

Important thing for me is: no new projection with all its complex and error prone formulas, just a "simple" datum shift, if at all.

I can do so after January 5th

We have a lot of time. The ASTER Data Center is slow. Last time it took 3 months!

[Bruce Kennewell]

Good morning (my time), Roland.

Well, I'm glad that you can understand what that's all about, mate!

No worries about waiting three months.....by that time hopefully the translation will be well underway.

Roland, I can obtain "Grid Digital Elevation Model (DEM), 3 Seconds of Arc" DEM which just might include the area in which I am interested.
Details are on the AUSLIG site, here.....
http://www.ga.gov.au/nmd/products/digidat/dem_3s.htm

From what I can understand, this offers a more accurate "map" than does DEM of 9 Seconds of Arc.....a 100 metre grid instead of 250 metres.

This has me somewhat bamboozled, so please forgive my stumbling about in the dark here!

Regards,
Bruce.

[Roland Ziegler]

Roland, I can obtain "Grid Digital Elevation Model (DEM), 3 Seconds of Arc" DEM which just might include the area in which I am interested.

3 arc sec equal about 90m spacing on latitudes, a bit less on longitudes - they say 80 m.

ASTER should give us 30m which means about 9 times as many grid points as with the 3 arc sec data. Particulary for mountainous areas the higher the resolution the better.

So I would suggest to see this product as a fall back in case ASTER will not give us what we want.

They say that data cover the "coastal strip" but I could not find more details and they also say that it's only 57MB total, which can't cover a lot as the format xyz is space consuming, even with compression.

I downloaded one of the available ASTER DEMs for the greater Sydney area to check my tools against southern hemisphere challenges. This DEM is a granule north of the city comprising part of the Hawkesbury River mouth. The DEM has the usual problems of relative elevations. Apart from that you can even recognize the freeway bridges over the river - the railway bridge unfortunately is just outside the DEM boundary. (freeway bridge just SW of -33° 32' 151° 12')

Regarding the topo maps I would expect that for an area of many outdoor activities like the Blue Mountains there should be maps of a larger scale than 1:250000.

[Bruce Kennewell]

Regarding the topo maps I would expect that for an area of many outdoor activities like the Blue Mountains there should be maps of a larger scale than 1:250000.

Yes, I have one here of the Lithgow area (includes the Zig-Zag, which was the reason I bought it), that is 1:25000.
It may be possible to go even finer, according to some info on the map....maybe down to 1:10000.

I'll check these out after January 5th., when businesses re-open.

Thanks for clarifying the DEM resolution, too....I didn't realise that the ASTER data was of that scale.

Regards,
Bruce.

[Roland Ziegler]

To wake up this thread again:

The first of the ASTER DEMs has been produced today (with a big hole in the center)

It's available here:
ftp://152.61.128.25/pub/asterdem/relati ... 134314.tif

I hope the other ones ordered will follow in the upcoming days.

[Bruce Kennewell]

Roland,
I have downloaded the file but it will not open in any of my programs capable of viewing TIF format.

Paint Shop Pro says that the file type is not supported!

Bruce.

[Oliver Lamm]

Bruce,

you need a program capable of viewing 'geotiff-files'.

E.g. microdem -> Microdem

Oli[/url]

[Roland Ziegler]

Yes, MicroDEM, conveniently available on the Zusi-CD.

Or my new TransDEM, though still without an English translation.

For a first look at the DEM MicroDEM will suffice.

Once all the DEMs on order are avalilable, TransDEM will be the suitable tool for cropping and merging. However, my simple test last night discovered a possible TransDEM problem either with merge or resampling, presumably a southern hemisphere issue.

[Roland Ziegler]

Here's #2 out of 7:

ftp://152.61.128.25/pub/asterdem/relati ... 111318.tif

[Bruce Kennewell]

Ahh! Microdem!
Okay, thanks Ollie and Roland for that.

I'll download #2 file right now.

Bruce.

[Bruce Kennewell]

However, my simple test last night discovered a possible TransDEM problem either with merge or resampling, presumably a southern hemisphere issue.

I've viewed both maps in MicroDEM and really have no idea of where I am looking - that is, where on the map I might be in relation to a latitude/longitude.

But I see what you mean about "black holes", Roland.

Bruce.

[Roland Ziegler]

Here's #3 out of 7:

ftp://152.61.128.25/pub/asterdem/relati ... 100757.tif

Seems, they're processing about one of our files per day.

Currently these DEMs look like pieces of a mosaic. Roughly, Penrith is located at [S 33°45', E 150° 42'] and Lithgow at [S 33° 30', E 150° 09'], not yet covered by the new DEMs.

[Roland Ziegler]

A DEM per day, and here's #4 out of 7:

ftp://152.61.128.25/pub/asterdem/relati ... 103649.tif

Virtually no holes, perfect terrain and, most important, we seem to have hit Lithgow.

However, also some bad news. Something is wrong with the GeoTiff header. If you open the DEM you will see negative northings. For some reason the DEM misses the 10000 km false northing offset for the southern hemisphere.

I've sent an email to the ASTER people and wait for their response how to deal with it.

Bruce, I would recommend to download the file anyway. If we need to fix this ourselves, I will provide an appropriate patch for TransDEM.

[Bruce Kennewell]

A DEM per day, and .......Bruce, I would recommend to download the file anyway. If we need to fix this ourselves, I will provide an appropriate patch for TransDEM.

Okay, Roland...no worries.

I've downloaded the first three and will get this one tonight (my time...it's now 4:40 PM Friday afternoon).

Regards,
Bruce.

[Roland Ziegler]

Two more: #5 and #6:

ftp://152.61.128.25/pub/asterdem/relati ... 161942.tif
ftp://152.61.128.25/pub/asterdem/relati ... 171910.tif

#6 covers Penrith. The Hawkesbury River is clearly visible - as a hole.

We still have a gap between # 5 and #6. Let's wait and see what #7 will bring.

[Bruce Kennewell]

Okay, Roland.
Have downloaded five and the sixth is coming down the pipe now.

Bruce.