TransDEM Forum

TransDEM News, Support, Hints and Resources
It is currently 23 Nov 2017 06:12

All times are UTC + 1 hour




Post new topic Reply to topic  [ 13 posts ]  Go to page 1, 2  Next
Author Message
PostPosted: 16 Aug 2016 11:53 
Offline

Joined: 28 Mar 2011 03:35
Posts: 67
I found a good source of historical Yugoslavian maps here:

http://library.mcmaster.ca/maps/ww2/ww2 ... dx100k.htm

The maps are in jp2 format which Transdem doesn't read, but via Irfanview and PSP7 managed to get them down to a reasonable size for Transdem to deal with.

However a bit stuck with the georeferencing as there's not a clear indication of Lat/Lon on the maps and the projection is shown as lambert Conical Orthomorphic with the spheroid as Bessel, which doesn't correspond to any of the views in Transdem. These are great maps for doing the old Yugo NG in particular the Ohrid line which is clearly marked unlike the more recent Russian maps. Any assistance would be appreciated.


Top
 Profile  
 
PostPosted: 17 Aug 2016 04:05 
Offline

Joined: 30 Jan 2011 13:03
Posts: 159
G'day BigVern,

Surely, Vern, you've been using TransDEM long enough to realise that the brown coloured numbers printed at each corner of the map refer to the latitude and longitudes (as Eastings and Northings, in this case) of some point close to the corner? On these maps, this is the intersection of the brown grid lines that appear on the map at the closest point to the corner. With these co-ordinates, it should be an easy task to manually georeference these maps in the normal way. As far as the projection is concerned, I would 'try' using the 'standard' UTM/WGS84 setting in TransDEM (despite the reference on the map to Lambert Conical Orthmorphic), as that's how I perceive the projection to be. If it doesn't work out as expected, then one of the other Lambert projections might do the trick (geophil might like to "enter the fray", at this point - if he's back from Italy - and offer better advice)...

Jerker {:)}


Top
 Profile  
 
PostPosted: 17 Aug 2016 08:07 
Offline

Joined: 28 Mar 2011 03:35
Posts: 67
I'll see if those lines and numbers are there but if you look at the maps from that site (try Kicevo or Ohrid) they are a bit of a mess in terms of actually showing useable reference points. If all else fails guess I could eyeball it from the location in Google Maps to that point on the map but I know from previous experience with the Italian maps off Castaneda, that doesn't always work and you end up with rivers half way up the hillside, etc.


Top
 Profile  
 
PostPosted: 17 Aug 2016 11:30 
Offline

Joined: 05 Jan 2011 16:45
Posts: 1057
I'm afraid, Garry, your suggestion won't work.

We do have quite a bit of information on these map sheets, however, these are not very encouraging because of one specific element.

As you, Garry, have already pointed out, we know the projection and its parameters, we have the ellipsoid and we even have a hint for the datum.

Let's analyse the details. The projection is given as “Lambert Conical Orthomorphic”. That appears to be a synonym for “Lambert Conformal Conic” with one parallel, one of the standard map projections for large scale topographic maps. TransDEM knows the formulas. The projection coordinates are in metres as they are for almost every map projection. (This only exception I have seen so far was a Cassini/Soldner projection for Trinidad where the coordinate unit is a Clark's chain link.)

We also know the true and false origin for the “brown grid”. The true origin is given as 54° 54' N, 14° E. In projection coordinates this equals the false origin of 800,000 E and 601,000 N.

Then we are told the ellipsoid as being “Bessel” which is a precise definition. (It's called spheroid here, which is actually more correct in the mathematical sense, since our ellipsoids always have two equal semi-diameters.)

So far we have all the necessary parameters to run the projection formulas. But this would be in a world of its own. The issue is the so-called geodetic datum, or more precisely, its “Fundamental Point”. Our map says “Greek datum”. Mapping to the Greek datum therefore means to see the world from an Athens perspective, at least coordinate-wise. This is independent of any projection and also applies to geographic coordinates, latitude and longitude. The Greenwich Meridian, from the Greek point of view, will not be the famous metal strip at the Greenwich Royal Observatory, but a few hundred metres off.

With the arrival of the Global Positioning System (GPS), surveyors decided to create a universal ellipsoid, called GRS80, later slightly modified into WGS84, and this now serves as a reference for global and absolute coordinates. For the national geodetic datums, conversions were defined, to transform their lat/long to WGS84 lat/long. A common method is the 7 parameter Helmert transformation, which is also used by TransDEM.

Before GPS times and before absolute coordinates this was a minor issue. It did not really matter in every day usage whether the datum was a Greek, Italian, Austrian or Hungarian one. And that explains why we don't find much information on the “Greek Datum”. Yes, there are sources which declare a fundamental point in Athens, but not how this relates to the WGS84 system. The only clue I have been able to find was a two parameter transformation to the much newer “Greek Geodetic Reference System 1987”, GGRS87. But information on this is vague and not consistent. (GGRS87 would be easy, it uses GRS80 and conversion to WGS84 is well known.)

So here we are. It's the “Greek datum” that causes problems.


Top
 Profile  
 
PostPosted: 17 Aug 2016 15:23 
Offline

Joined: 28 Mar 2011 03:35
Posts: 67
In summary then Roland, the best use for these maps is either to use to try and pick up the course taken by the lines in Google Earth to plot a kml line (which can then be referenced against more modern mapping such as OSM) or just try as I noted to GeoRef via approximate references from Google Maps. However at 1:100000 you only need to be off by a few pixels to translate into significant displacement in Trainz.

Shame, as I was so pleased to discover this particular map series.


Top
 Profile  
 
PostPosted: 18 Aug 2016 00:52 
Offline

Joined: 30 Jan 2011 13:03
Posts: 159
G'day geophil & BigVern,

...so, I stand corrected (or to be more precise, I sit), which doesn't bother me. However, there may be a workaround available to us, IF we can match the projection. If any of the available projections utilises the same formula required to correctly project the maps in question, then the offset created by the geodetic datum MAY be overcome by using the "Shift DEM" option (by means of appropriate Control Points), to move the DEM so that it 'sits' properly under the raster maps (although I would much rather be shifting/transforming the raster maps so that they 'sit' properly on top of the DEM. This is a 'modification/upgrade' to TransDEM that I suggested quite a while back). Again, correct me if I've got it wrong but if we're MANUALLY georeferencing the map clipping, using co-ordinate references provided on the map does it matter which projection the map is viewed with, especially since, once that process is completed, we then transform the map into the UTM/WGS84 projection?

Jerker {:)}


Top
 Profile  
 
PostPosted: 18 Aug 2016 07:59 
Offline

Joined: 28 Mar 2011 03:35
Posts: 67
Quote:
Again, correct me if I've got it wrong but if we're MANUALLY georeferencing the map clipping, using co-ordinate references provided on the map does it matter which projection the map is viewed with, especially since, once that process is completed, we then transform the map into the UTM/WGS84 projection?


I guess that then comes down to how accurately the map was drawn. From what I can see, these maps were created either before or during WWII so they are certainly not based on any form of satellite or other modern surveying technique. Definitely on balance I think the best use is as an assist in GE to find or guess the course taken by these old routes and use that to build a detailed kml path. What's annoying is I did have the Ohrid and Dubrovnik Lines in kml's on the old laptop but botched the data transfer and lost them (along with a load of other stuff) in the process.


Top
 Profile  
 
PostPosted: 18 Aug 2016 18:29 
Offline

Joined: 05 Jan 2011 16:45
Posts: 1057
Some approximation may work, but it will probably only be good for a small area. Wouldn't Open Street Map have the railways? Could be easier to find reference points than ortho-imagery.

Anyway, I have put it on my ToDo list as a wish. I would try to use the available datum information. All steps in TransDEM will have to be implemented before we can say anything. That's unfortunate because in the end, it may still come to nothing.


Top
 Profile  
 
PostPosted: 19 Aug 2016 08:27 
Offline

Joined: 28 Mar 2011 03:35
Posts: 67
Quote:
Wouldn't Open Street Map have the railways?


Unfortunately not. To the best of my knowledge the course taken by the old Ohrid line was never shown on OSM and in other parts of (former) Yugoslavia where the Steinbeis and Dubrovnik routes were previously clearly shown, these appear to have been removed, for reasons unknown. The pitfalls of Open source, as I notice abandoned railways have also been removed where previously shown in other countries - as diverse as the UK and Australia.


Top
 Profile  
 
PostPosted: 19 Aug 2016 21:29 
Offline

Joined: 05 Jan 2011 16:45
Posts: 1057
BigVern wrote:
Quote:
Wouldn't Open Street Map have the railways?


Unfortunately not. To the best of my knowledge the course taken by the old Ohrid line was never shown on OSM and in other parts of (former) Yugoslavia where the Steinbeis and Dubrovnik routes were previously clearly shown, these appear to have been removed, for reasons unknown. The pitfalls of Open source, as I notice abandoned railways have also been removed where previously shown in other countries - as diverse as the UK and Australia.
I see. With OSM it's always down to the contributors, of course. You may find fascinating details, like accurate track plans of tram depots, to the last piece of rail, for a place strictly off-limits to the public. I wonder...

Abandoned lines, if attributed correctly, should still be rendered - OSM will never throw away valuable data - but may not be very easy to detect.


For the "Greek Datum", I have discovered further information which looks more promising. It will need some extension to my datum shift class but I think it's worth a try.

Have you noticed that there are two longitudes marked on the Yugoslavian maps, black and blue? The original black is east of the Paris Meridian, while blue is east of Greenwich. That's a most interesting piece of history. On the European continent, under influence of the French revolution, land cartography was based on a longitude origin in Paris, originally 20 degrees east of Ferro (El Hierro), later somewhat adjusted, the western end of the antique world. There was an international conference in the late 19th century where the Greenwich meridian was agreed upon worldwide but it still took decades before it became the mapping standard everywhere. Germany switched from Ferro to Greenwich (and from Cassini/Soldner to Gauss/Krüger) in the late 1920s.

A few years ago BBC 4 showed a 3 part documentary series on the history of maps, very informative :idea:. There is also a short article on the Paris Meridian in wikipedia.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 13 posts ]  Go to page 1, 2  Next

All times are UTC + 1 hour


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum

Search for:
Jump to:  
cron

Imprint

Powered by phpBB® Forum Software © phpBB Group