I need to convert (lat,long,height)points into a cartesian system for an area as small as 40kmx40km or even 10kmx10km. I'm currently using the UTM projection, but there are some issues regarding accuracy and distortion. The good thing with UTM is that GDAL library easily do the math via python coding.

My professor told me 'plano topográfico local' (in Portuguese) would be a good option. The bad thing is that I have no idea of the EPSG code for this system in order to use GDAL. Any ideas?

  • 2
    I'm guessing that 'plano topográfico local' will translate to 'local coordinate system' which I think means that you will not find an EPSG code for it because it may have been created as a "one-off". – PolyGeo Jun 12 '15 at 3:18
  • 1
    It depends what you're trying to do with the data in the 'Cartesian' coordinate system. Different projections have merits and detriments.. so UTM doesn't work - why? what's wrong with it? How can you tell it's not accurate? According to the Heisenberg's uncertainty principle everything is inaccurate. – Michael Stimson Jun 12 '15 at 3:37
  • 2
    Although you write about "some issues," exactly what are they? What forms of distortion are of concern? What are your actual accuracy requirements? Depending on what they might be, you might find UTM acceptable, or you could apply a very simple adjustment (change the scale factor), or you might need a more customized projection. – whuber Jun 12 '15 at 16:44
  • Thanks @PolyGeo. I guess I will have to create my own projection. In order to do that automatically, I'm possibly implementing some python code. – Evandrojs Jun 15 '15 at 12:40
  • @MichaelMiles-Stimson, thank you for the comment. I have searched for literature comparing the amount of distortion provided by different projections, but never found it. I was working with UTM when I found a brazilian standard NBR 14166:1998 that sets "plano topográfico local" as the system to used in areas as small as 50km x 50km. That system ignores earth curvature, but considers terrain heights. On the other hand, as I understood, UTM considers the ellipsoid , but tackles altitude as and attribute. – Evandrojs Jun 15 '15 at 12:54

I cannot imagine that UTM is not suitable for your purposes. Are you sure you picked the correct zone for the location of your dataset? Example: http://spatialreference.org/ref/epsg/32632/

I don't know much about P.T.L. but it seems to work similarly as UTM. It is unlikely that it is included in common GIS coordinate reference system (CRS) definitions, though. Probably you have to define it as new CRS and enter the parameters manually (or get the reference from an existing dataset in P.T.V)

This may help as well: https://metrica.zendesk.com/hc/pt-br/articles/204619335-Como-transformar-coordenadas-UTM-para-o-Plano-Topogr%C3%A1fico-Local-

  • Just because a spatial reference doesn't have an EPSG number doesn't make it any less valid or usable, quite often you can get them (or make) as a prj file and use -a_srs ESRI:FULL\PATH\TO\PRJ.prj to supply the parameters - provided that the parameters are known. – Michael Stimson Jun 12 '15 at 5:24
  • Actually I would like to generate a projection automatically, given a range of lat/longs . That is possible with UTM due to python calling of GDAL libraries. I guess the easisest way for my project purpose is implementing the code by myself. Using any software would lead to less flexibility. – Evandrojs Jun 15 '15 at 13:18
  • UTM has a drawback with the scale factor of 0.9996 which can lead to wrong length/area results if software is projection-aware. – user30184 Jul 16 '15 at 20:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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