I have created a grid, which I want to use in an ArcGIS Online application. To create the grid (fishnet) I have used the local projection, which is in meters. I avoided using WGS84 as it is in degrees and is not so accurate. As you can see in the attachment the grid looks ok but as expected the squares are not in the same line. Since I will use this map for reporting purposes, the tilted squares don't look so good.

Do you have any ideas how I could "fix" this in order to look more neat?

enter image description here

  • 3
    Reproject your map to match the grid? – Ian Turton Sep 27 at 19:02
  • The map is published in arcgis online where the base layer is in WGS84 and can not switch it. I tried to reproject the grid layer but it doesn't give the expected result either. – user1919 Sep 27 at 19:06
  • How did you do the reprojections? I agree with @IanTurton in that you should reproject the map dataframe. I would use a Transverse Mercator projection such as UTM. – Aaron Sep 27 at 20:06
  • 2
    Just to clarify, WGS84 is a datum which doesn't necessarily mean geographic, some very good projected coordinate systems are also WGS84, Google uses EPSG:3857 which is based on WGS84 epsg.io/3857. Geographic data doesn't have to be inaccurate, sure the boxes get distorted when displayed in geographic CRS but that doesn't affect the accuracy of the vertices, it's only the way you're looking at the same data that changes, from that point of view distances and areas do not calculate correctly but you can use calculate geometry in a projected CRS to get properly calculated length or area. – Michael Stimson Sep 28 at 5:44

To get a grid or fishnet that has truly horizontal and vertical lines, you have to create in the same coordinate reference system (CRS) that you're going to display it in. In some cases, you can build it in a different CRS and it will still be 'aligned'. That usually requires that both CRS use a cylindrical projection.

In this case, it sounds like you created the fishnet using the area's usual CRS which could be UTM, Stereographic, or Lambert conformal conic. All of these, projected to WGS84 AKA pseudo-Plate Carree or more like Web Mercator will show a 'tilt' to the fishnet.

The 'tilt' seems rather small compared to some that I've seen. The only way to fully remove it would be to build the fishnet in the same CRS that you're going to display the data in.

| improve this answer | |

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.