Posted on Leave a comment

Lidar & Overlapping UTM Zones

Second zone DEM imported

by Rob Plowright

Intro by Ken: Other than Tasmania & ACT, all states & territories have 2 zones. NZ has just one zone. I recall struggling with marrying datasets from 2 zones for a map that was close to the edge of 54/55. So Rob’s experience will be valuable for the few who are faced with this issue. The general process will also apply to other types of data.

Coping with Australia’s 4 zones

Now that didn’t work

The other day I was doing some virtual exploring: processing lidar in OCAD to see if I could find some interesting terrain. One area I was interested in just happened to be on the boundary between UTM zones 55 and 56; first time I had experienced this. Getting adjacent lidar tiles from different zones to line up in one OCAD file proved to be more difficult than expected. After trying several different approaches and getting nowhere I finally managed to do it, with the help of a suggestion from Ken.

This involved processing the tiles in one zone then using Map > Transform > Change Coordinate System to change the map into the next zone, then processing the remaining tiles. There was one problem however: while the map (contours) transformation went well, the background images (hillshade, vegetation, etc)  did not.

Ken asked me to write-up the process for his up for his blog. Since there was the problem with the background images not transforming properly I sent the first draft to OCAD support. Gian-Reto got right back to me saying that while the method I described would work, there was a quicker way – just using the DEM Wizard. “That’s odd” I thought, “I tried that and it didn’t work”. Gian-Reto soon figured out there was a bug that meant southern hemisphere UTM zones did not process properly and he promptly fixed it.

So just in case anyone wants to know, here is how to do it.

The works

The lidar coverage

I will use an example from Kanangra in the Blue Mountains (not the area I was looking at – which shall remain secret for now).

  • I have five lidar tiles: two in zone 55 and three to the east of that in 56. They are:
    • Taralga201611-LID2-C3-AHD_7746232_55_0002_0002
    • Taralga201611-LID2-C3-AHD_7766232_55_0002_0002
    • Burragorang201805-LID2-C3-AHD_2226232_56_0002_0002
    • Burragorang201805-LID2-C3-AHD_2246232_56_0002_0002
    • Burragorang201805-LID2-C3-AHD_226232_56_0002_0002

As you can see from the double digit components above, the Taralga tiles are in zone 55 and the others in 56

Each tile is 2km x 2km and the numbers after ‘AHD’ give you the grid reference for the SW corner of the tile. So the  SW corner of the first Taralga tile is at 774000 6232000. and the first Burragorang tile is at 222000 6232000  You will notice that the Burragorang tiles have the same northings (6232000) as the Taralga tiles but different eastings.

1. Select co-ordinate system

The area is mostly in Zone 56 so I have decided to make the OCAD file zone 56. Set up a map in zone 56.  First process the zone 55 (Taralga)  tiles. (You could just as easily do zone 56 first).

Select co-ordinate system
Selecting co-ordinate system

The map now looks like this.

Lidar image
Lidar image first zone

Contours (5 and 25m)  are in grey and vegetation height is in the background (I have not bothered rotating the map to magnetic north for this exercise). The blue border shows the OCAD DEM boundary (which is oriented to zone 56)  but I have drawn in the actual boundary of the Taralga tiles’ data with the purple dashes. As you can see the Taralga tiles are sightly rotated relative to the Burragorang tiles, as you would expect since they are oriented to zone 55.

Looking closely
Rotation artifacts

If you look closely there is a little bit of junk in the gaps between the OCAD DEM and the actual (rotated) Taralga tiles.

2. Processing the other zone dataset

Now process the Burragorang tiles – but this time, in the DEM Wizard, make sure it says zone 56 for the DEM tiles.

DEM import next zone
Setting co-ordinate systems for next zone

Now the map looks like this…

Second zone DEM imported
Second zone DEM imported

I have made the Burragorang contours red for the sake of this exercise.

If you look closely at the above image, you can see that there is a large overlap between the westernmost Burragorang tile and the easternmost Taralga tile; almost a whole 2 x 2km tile. It is not always this much. In the area I was looking at before, it was only about 500-600m.

Also in the close-up below, you can see that the red and grey contours  are not exactly the same. This is because the two lidar sets are from different flights: Taralga 2016 and Burragorang 2018. In the other area I was looking at, both lidar sets (zone 55 and 56) were from the same year, and obviously from the same flight, as they matched exactly.

Difference in contours
Close up of difference in contours

The red contours above are more detailed (noisier) as, being more recent, that lidar is higher resolution. If  this map is just for a base map the slightly different contours is not necessarily a problem. But if you want to eliminate the overlapping contours you can crop one set.

3. Eliminate overlap

Since the Burragorang contours are more recent I am going to crop the Taralga contours. Start by hiding the Burragorang (red) contours and draw an area (light blue in this example)  over the parts you want to crop. Then use the Object > crop objects tool. You can also draw the cropping area so as to remove the junk around the edges.

Cropping the overlap
Cropping the overlap

And I get this…

Cropped overlap
Cropped overlap

Unhide the red contours to get this…

The full picture
The full picture

Up close the border between the zones looks like this…

Close up of the join
Close up of the join

Tips

  1. This process, excluding the DEM aspects, can also be used for non-lidar datasets over 2 zones.
  2. If you are new to lidar processing then download one of my DEM templates for use with the DEM Import Wizard. The template has 15 differentially coloured symbols to cover each of the contour selections in the wizard and to avoid confusion with any brown contours in your main or background map. Get one at ocad.com.au/mapping-resources
  3. If you have distortion in a small area, check out Map > Transform > Local transformation. It makes the adjustment of existing maps to geo-referenced base maps (hillshading, orthophotos etc.) easier and more accurate.
Leave a Reply

Your email address will not be published. Required fields are marked *