1 Topic by burkmarr

  • burkmarr
  • burkmarr
  • FSC BioLinks Digital Development Officer
  • Offline
  • From: Preston Montford
  • Registered: 14-05-2006
  • Posts: 105

Topic: WMS WGS-84 projection appears incorrect

The images below are both screen shots of MapInfo Professional using the NBN Gateway WMS to display a tetrad distribution map for a single species.

The first image shows the WMS configured to project with EPSG:27700 (BNG):

[img=bng]http://www.gilbert21.org.uk/temp/wms-epsg27700.jpg[/img]

The second image shows the WMS configured to project with EPSG:4326 (WGS-84):

[img=wgs84]http://www.gilbert21.org.uk/temp/wms-epsg4326.jpg[/img]

(I get the same result as above if I use the third available projection - CRS:84)

The scale of the errors reminds me of a time when I was trying to convert from BNG to WGS-84 but was actually converting to OSGB36 by mistake. Could that be the problem?

This is important for me to resolve because I want to use the WMS to project distribution maps over BING and Google maps, so I have to be able to project acurately in WGS-84. I'm getting exactly the same scale of errors when I'm projecting over a BING map (haven't tried Google yet).

Rich

Richard Burkmar
Biodiversity Project Officer
Field Studies Council

burkmarr's Website

2 Reply by burkmarr

  • burkmarr
  • burkmarr
  • FSC BioLinks Digital Development Officer
  • Offline
  • From: Preston Montford
  • Registered: 14-05-2006
  • Posts: 105

Re: WMS WGS-84 projection appears incorrect

I did some more work on this and I believe that I have confirmed my suspicion that the WMS is converting coordinates to the OSG36 lat/lon coordinate system instead of WGS84 lat/lon, but projecting these as WGS84.

This is what I did:

1. I used MapInfo to save a copy of my tetrad squares vector layer (white squares) with WGS84 lat/lon coordinate system.

2.  I also used MapInfo to save another copy of my tetrad squares vector layer with OSGB36 lat/lon coordinate system.

3. I exported both these layers to MapInfo MIF files. I then 'mashed up' the files by taking the coordinates from the OSGB36 MIF file and putting these with the headers describing the coordinate system/projection from the WGS84 MIF file.

4. I imported this file into MapInfo. As expected the tetrad squares were displaced from their correct positions.

5. I opened the NBN WMS again and displayed the distribution tetrads again (green) using the EPSG:4326 projection. Hey presto - they exactly matched the incorrectly positioned vector tetrads.

[img=mashup]http://www.gilbert21.org.uk/temp/wms-mashup.jpg[/img]

Can someone take a look at the Web Mapping Service and correct this? It must be a simple configuration problem.

Rich

Richard Burkmar
Biodiversity Project Officer
Field Studies Council

burkmarr's Website

3 Reply by Graham French (edited by Graham French 24-09-2012 11:52:11)

  • Graham French
  • Administrator
  • Offline
  • Registered: 23-05-2006
  • Posts: 430

Re: WMS WGS-84 projection appears incorrect

Hello Richard

I will ask the NBN Gateway team to have a look into this further to see if there is an issue with converting gridreferences to the WGS84 coordinate system

Best wishes

Graham

Graham French's Website

4 Reply by Graham French

  • Graham French
  • Administrator
  • Offline
  • Registered: 23-05-2006
  • Posts: 430

Re: WMS WGS-84 projection appears incorrect

Hello Rich

The WMS uses the default transformation done by ESRI ArcGIS Server and it looks like this is causing the reprojection issue with WGS84.

A new NBN Gateway release is planned in December. In this release there will be a change in the underlying technology that powers the WMS so that this projection issue should be resolved with this new release

Best wishes

Graham

Graham French's Website

5 Reply by burkmarr

  • burkmarr
  • burkmarr
  • FSC BioLinks Digital Development Officer
  • Offline
  • From: Preston Montford
  • Registered: 14-05-2006
  • Posts: 105

Re: WMS WGS-84 projection appears incorrect

Wow - that's surprising.

It's not a show-stopper for me. For large scale 10 km distribution maps it's not noticeable at all. But when you zoom into a tetrad, or 1 km distibution maps it is noticeable if you overlay on a backdrop with a grid - for example Bing Maps OS backdrop. 

And of course six-figure grid references - 100 m squares - are shown in completely the wrong place - can be a couple of hundred metres out here in Merseyside.

Here's to the new release then!

Many thanks,

Rich

Richard Burkmar
Biodiversity Project Officer
Field Studies Council

burkmarr's Website