I was trying to draw kml that are (one degree in gap) meredians and parallels for Austrailia. e.g see i tried to draw 142 degree and 11 degree lines using fishnet in arcmap but when imported into Google Earth they shifted a bit. I used WGS84(EPSG:4326) when creating lines.I supposed that google earth uses EPSG:4326 (I also tried EPSG:3857) as described at here, here and here.

enter image description here Now how can i achieve 100% matched kml in arcmap. My extent is 100E, 50S to 155E, 3S.


Below are the steps i did to generate KML as attached above-

  1. Open Create Fishnet Tool in a dataframe set to WGS84(EPSG:4326) and assign the parameters as -

My fishnet setting- setting

  1. From the spatial extent i am generating fishnet on i.e. 100E, 3S to 155E, 50S i counted the number of parallels(50-3=47) and meridians(155-100=55) and used this number as row and column number in the fishnet tool as shown above.

  2. Then i used Densify tool to generate sufficient points in the lines(0.001 degrees distance interval) whereas projection does not shifts the whole feature.

  3. I defined the projection of the layer WGS84(EPSG:4326) using Define Projection tool(I also tried EPSG:3857)

  4. Finally i reprojected the line layer into WGS84(EPSG:4326) using Project tool(I also tried EPSG:3857)

  5. And exported the lines into kml using Layer to KML tool.

Now, Problem is when i zoom in the exported kml in google earth i see a shift as shown in the attached.So i need a kml(with generation method) in the above spatial extent(100E, 50S to 155E, 3S) that matches exactly with the grid of google earth irrespective of zoom level

System datails:

I am in Windows 8 and Arcmap Advanced and by WGS84 i mean as below

WKID: 4326 Authority: EPSG

Angular Unit: Degree (0.0174532925199433)
Prime Meridian: Greenwich (0.0)
Datum: D_WGS_1984
  Spheroid: WGS_1984
    Semimajor Axis: 6378137.0
    Semiminor Axis: 6356752.314245179
    Inverse Flattening: 298.257223563
  • What were the precise parameters that you used for Create Fishnet when you created your grid and defined its projection?
    – PolyGeo
    Commented Jun 4, 2016 at 7:43
  • ... and your projection details?
    – PolyGeo
    Commented Jun 4, 2016 at 10:54
  • @PolyGeo added details..
    – SAZ
    Commented Jun 4, 2016 at 11:20
  • Where? Did you use the Project Define tool on your fishnet? If so, what parameters did you provide it?
    – PolyGeo
    Commented Jun 4, 2016 at 11:29
  • 1
    I gave up trying to help on this one because the question does not describe a procedure step by step in sufficient detail for me to be able to perform a test without having to keep guessing/asking what has been done. I recommend taking the Tour and reviewing meta.gis.stackexchange.com/questions/3349/…
    – PolyGeo
    Commented Jun 5, 2016 at 6:25

2 Answers 2


WGS84 is not a projection, it is a Datum, there is a key difference between the two. You could have WGS84 in a geographic projection, or a UTM projection, or many other projections. You want to make sure that not only are you using WGS 84, but are also using the correct projection. Google Earth uses Simple Cylindrical projection with a WGS84 datum, you are likely using a geographic projection as that is the default for Arc. Google earth attempts to reproject such that other projections fit into it nicely but this does not always happen perfectly. Try different projections and see if they translate more easily into google earth.

  • -1! -Ha! When OP said I used WGS84, he mentioned the details of the projection Google Earth uses, there is no way other than to use GCS WGS84 in ArcMap- so i see there is no single piece of information in this ANSWER wherefrom the OP may be benefitted!OP ,already, not only mentioned Projection details used here even with EPSG code but also wanted a kml that can match with GE what about it? You could even generate and share the KML as the OP shared his KML. No ground of upvoting at all!
    – Learner
    Commented Jun 14, 2016 at 3:15

The meridians in your Google Earth picture look like they are displaying at every 0.0035 degrees. It looks like you are measuring the distance between your data, which is placed correctly at 142E, and the 141.9985E meridian.

Meridian image

The measured distance between 142E and 141.9985E is right around 170 meters.

Your data looks correct, but you are validating against the wrong meridians and parallels.

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.