1

I have been looking for a historic soil map shapefile of Bosnia & Herzegovina for a while and finally found one - a UN FAO project archive had the files available, albeit designed for ArcView. The issue I have however is that no CRS was assigned to the data - this wouldn't be normally a problem, but from that 2000-2002 postwar period of the project, it is not clear what projection was preferred for this data. To make matters worse, the former Yugoslavia used many different Projections from my research so it's given me a bit of a headache.

The one hint I have is from documentation of that specific FAO project ('INVENTORY OF POST-WAR SITUATION OF LAND RESOURCES IN BOSNIA AND HERZEGOVINA (GCP/BIH/002/ITA)'), I found a description of the Ellipsoid and Datum that were used;

Basic topographic data in the Federation of Bosnia and Herzegovina (as well as in most ex Yugoslav countries) was available on topographic maps at various scales produced by the Military Geodetic Institute based in Belgrade. Gauss Krueger [sic] projection (3 degree), (with Bessel - 1841 ellipsoid, and Hermanskogel [sic] datum) was used as the State Coordinate System. Transformation parameters for the World Geodetic System (WGS 84) were also calculated, therefore all data recorded (using GPS) or assumed from other sources can be transferred in State Coordinate System.

I've tried settting a few Layer Projections in QGIS so far based on that information and some educated guesses using EPSG.io, but these are extremely far off and/or badly deformed.

  • Psuedo Mercator EPSG 3857
  • WGS84 EPSG 4326
  • GCS Hermannskogel ESRI 104102
  • Several Gauss-Kruger 3-degree Projections covering Central Europe & Germany
  • MGI / 3-degree Gauss zone 6 EPSG 31266
  • MGI / Balkans zone 6 EPSG 31276
  • MGI 1901 / Balkans zone 6 EPSG 3908/8678
  • Macedonia State Coordinate System truncated EPSG 9945

The best I have found so far are;

  • Pulkovo 1942 / Gauss-Kruger 3N EPSG 28463
  • Pulkovo 1995 GK Zone 3N EPSG 20063
  • IGM95 / UTM zone 33N EPSG 3065
  • MGI (Ferro) / Austria Central Zone EPSG 31292
  • SRB_ETRS89 / UTM zone 34N EPSG 8682
  • Slovenia 1996 / UTM zone 33N EPSG 8687
  • WGS 84 / UTM zone 33N EPSG 32633
  • WGS 84 / UTM zone 34N EPSG 32634

They get very close by being relatively at the same Latitude, but are off horizontally by a few hundred Kilometers. What's strange is the 'better' projections don't use those Datums or Ellipsoids as described, so I'm not certain if the original documentation is even correct. It's also strange that these are for other neighbouring countries, not the 'Balkan Zone' ones that are explicitly for BiH. Here's and example below of a 'good' one.

Image showing the horizontal allignment distance between the best CRS I've found so far (in Red), and the 'correct' alligned WGS84 Administrative outline with the Psuedo Mercator OSM backdrop map.

Can anyone suggest another CRS that I've missed that might correctly allign with the map?

Otherwise I might need to look into creating a custom CRS of '33.5N' to shift the shapefile across which will still not be as accurate as an official Projection.

Improve this question

1 Answer 1

Reset to default
1

Best I found is modified EPSG:8678 MGI 1901 / Balkans zone 6 with different origin and added datum. It seems to work. Checked in QGIS as Custom CRS using Proj.4 notation:

+proj=tmerc +lat_0=0 +lon_0=18 +k=0.9999 +x_0=500000 +y_0=0 +ellps=bessel +towgs84=682,-203,480,0,0,0,0 +datum=hermannskogel +units=m +no_defs

enter image description here

Improve this answer
1
  • This works perfect, thank you! It's also working with some other projectionless files that come alongside I used to verify.
    – Needs_Caffeine
    Commented Jul 16 at 12:45

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.