I am attempting to create a GIS Project with multiple Brazilian cities. I have been collecting, and am interested in, shapefiles that delineate the neighborhood boundaries of said cities.

One city's shape file that I have acquired is for Curitiba, Paraná (PR), Brazil. I obtained the city-neighborhood shape file from here: Institute of Research and Urban Planning: Curitiba. You will find the file under "Divisa de Bairros" (Neighborhood Divisions).

I am using Quantum GIS to overlay these city shape files over Google Maps, because ultimately I may need to incorporate street or highway information in to my analysis. I am ambivalent ultimately about which source I use, and am open to solutions using Open Streets if possible.

I have read the GIS help files and followed the process to view the metadata in ArcMap to see what the coordinate system and/or projection system is for this shapefile. However, on reviewing the metadata it says that the coordinates have "No Z Values" and "No measures" and there is not defined coordinate system.

My question is: How can I fix this coordinate/measurement issue so that I can correctly project the location of the Curitiba city shapefile on either the Google Maps or Open Streets projections so that I may move forward with my project?

Note: I have already contacted the creator(s) of the files and asked if they may provide any additional information. I will post or answer my own question as need be.

  • @PolyGeo the suggested answers for that question all involve "brute force", "guess", "ask the people who gave it to you." I am looking for an actual answer that uses GIS rather than trial-and-error. – DV Hughes Aug 22 '14 at 23:00
  • 2
    There are an infinite number of possible permutations; this not something which can be scripted effectively. – Vince Aug 22 '14 at 23:08
  • 2
    GIS can be used to test guesses, but without projection data there is no tool that can automatically figure it out for you. (Consider UTM, for example: there are dozens of UTM zones around the world, and valid easting/northing values could be located in any of them -- you need to know which UTM zone to accurately place the data.) – Erica Aug 22 '14 at 23:11
  • @Vince alright. I will begin exploring those infinite number of possible permutations. – DV Hughes Aug 22 '14 at 23:12
  • 1
    If you wish to avoid "brute force" and an "[educated] guess" then you are down to "ask the people who gave it to you." If the last option of these yields no answer too then GIS can be used to rubbersheet (Spatial Adjust) the data to fit. – PolyGeo Aug 22 '14 at 23:12

On the page that you linked to, there is a note on the bottom:

DATUM: SAD-69/original


OBS.: Devido ao uso do DATUM SAD-69/original por Curitiba, para converter para SIRGAS-2000, devem-se utilizar os parâmetros locais de transformação de coordenadas descritos abaixo:

SAD-69/original(Curitiba) -> SIRGAS-2000

dX = -66,163 m

dY = 2,028 m

dZ = -33,718 m

In English (via Google Chrome, blame/thank it for the translation):

DATUM: SAD-69 / original

Cartographic Projection: UTM-22 (-51º)

NOTE: Due to the use of the DATUM-SAD 69 / original by Curitiba, to convert to SIRGAS-2000, should be using the parameters of local coordinate transformation described below:

This projection needs to be applied to the data you downloaded; I don't know the exact tool in QGis to accomplish this, but there are doubtless tutorials and/or other GIS.se members who can help in that regard.

  • 2
    Another victory for "ask the people who gave it to you"! – Vince Aug 22 '14 at 23:20
  • @Erica: Alright thank you. That was actually helpful. – DV Hughes Aug 22 '14 at 23:20
  • 1
    Glad to help. @Vince That does tend to be the most reliable :) – Erica Aug 22 '14 at 23:20
  • 1
    @Erica: On a final positive note, you are the bomb.com and totally solved my issue. Have a great day. – DV Hughes Aug 22 '14 at 23:36

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