11 replies to this topic

[meghan m]

Hi,
Currently I am a student in the Cartographic Specialist program at Fleming College.
We have a coop project to complete in 3.5 weeks time and we experienced a major setback with respect to data and therefore had to start over last week. The final output is a 24''x30'' thematic map with a background which is a detailed base map of Nunavut. It will contain surround information about the differences of wastewater treatment between northern and southern latitudes. The base has been created in ArcMap, and we have explored several DEMs in effort to find proper resolution. ETOPO1 data is what was settled on. We created a custom colour ramp in ArcMap, and it has left the north of Nunavut with a yellow tinge. Can this be considered as cartographically correct? If not, what would be some suggested routes to take in ArcMap or Photoshop to make the North appear a little bit more northern.
Attached is screendump of where we currently are on the project.

Attached Files

•  nunsd.JPG   76.24K   92 downloads

[Hans van der Maarel]

Hi Meghan, and welcome to Cartotalk!

Technically the yellowish tinge is correct, those elevations do occur there, but it does look a bit odd when used as a backdrop for thematic data. I would say there's 2 options:

1: get rid of the color ramp and instead blend a 'regular' shaded relief with land cover data. You've already done that with Greenland, might as well continue for the rest of the map.

2: redo the color ramp so that yellow isn't used anymore.

Tom Patterson has been doing a lot of work on cross-blended hypsometric tints. I would strongly recommend reading his website on this matter.

Hope this helps.

[rudy]

Hi Meghan, and welcome to Cartotalk!

Technically the yellowish tinge is correct, those elevations do occur there, but it does look a bit odd when used as a backdrop for thematic data. I would say there's 2 options:

1: get rid of the color ramp and instead blend a 'regular' shaded relief with land cover data. You've already done that with Greenland, might as well continue for the rest of the map.

2: redo the color ramp so that yellow isn't used anymore.

Tom Patterson has been doing a lot of work on cross-blended hypsometric tints. I would strongly recommend reading his website on this matter.

Hope this helps.

Hi Meghan!
How are things are SSFC? It's been a while since I graduated ('97) but I'm sure a few of the same people are still there (instructors, that is. Hopefully not students!).

I concur with Hans. Considering that you are under an unexpected time constraint I would suggest that you alter your colour ramp. Combining the DEM with landcover might be preferable but it adds to the complexity of the final product . . . . which means more work and possible printing issues. When you are running out of time, keep it simple.

[meghan m]

We downloaded the file from shadedrelief.com, but when we clipped the area we wanted, the colours changed. They turned a lot darker than they were originally.
Also, the image is in WGS 84, and when we project the clipped image into Lambert conformal conic in ArcMAP, the image is unrecognizable. Since we are making our final document in Adobe Illustrator, we tried placing the png file into Adobe and putting the layer into the MapView with a lambert conformal conic projection. Usually, this works perfectly, but for this file it didn't work. We registered it to its world file, and it was very tiny. I don't have images of this at the moment, but I will post them soon.

Thanks!

Fleming is good, I'll be sad to leave. Lindsay, however...I'm not sure. Bruce Petit was probably here back then, and he's still pretty sharp.

[Hans van der Maarel]

We downloaded the file from shadedrelief.com, but when we clipped the area we wanted, the colours changed. They turned a lot darker than they were originally.

Okay... that should not happen... Which file did you use? The "sneak preview" with cross-blended hypsometry? I didn't try that one myself, but I worked extensively with other files produced by Tom and never had any problems. Also re. projection.

[meghan m]

Ah, I figured out why the colours changed. The stretch type in the symbology was set to standard deviation. I reset it to "none" and the colours changed back. All is well...for now. Thanks!

[meghan m]

Have you used the NE2_modis3.jpg from shadedrelief.com? I brought it into ArcMAP and it shows North America in approximately 4000E 2634S. It's in GCS_WGS_84. Is this image supposed to do this?

[Hans van der Maarel]

Have you used the NE2_modis3.jpg from shadedrelief.com? I brought it into ArcMAP and it shows North America in approximately 4000E 2634S. It's in GCS_WGS_84. Is this image supposed to do this?

Well, it's in WGS84 alright, but I seem to recall some issues with the world files Tom supplied for Natural Earth II. You should have a NE2_modis3.jpw file as well, which probabely looks like this:

0.02222222222222 0.00000000000000 0.00000000000000 -0.02222222222222 -180.00000000000000 90.00000000000000

Chance the last 2 lines so the file looks like this:

0.02222222222222 0.00000000000000 0.00000000000000 -0.02222222222222 -179.9888888888888989.98888888888889

The trick is that the last 2 lines are the x and y coordinates of the center of the top left pixel. In the original, the coordinates were of the top left corner of the top left pixel, which, along with the other parameters, causes all sorts of problems with the data going slightly off the earth and creating a gap on the other end.

If you don't have the file yet, simply create it using Notepad or another plain text editor (make sure to save it as plain text) and that should do the trick.

[meghan m]

I managed to georeference the jpeg with a world shapefile to get it in the right place. Then I projected it into Lambert Conformal Conic, so far so good. But at a 1:5,000,000 scale, the resolution is quite poor, even worse than the original DEM we have (the one with the yellow tinge). Final thoughts? I'll try to export the image again at a small cell size, and hopefully ArcMAP won't crash again.

[meghan m]

Here's a sample of what the image looks like so far.

Attached Files

•  nunavut.jpg   530.38K   53 downloads

[Hans van der Maarel]

Here's a sample of what the image looks like so far.

How big (size/resolution) is your map?

[meghan m]

How big (size/resolution) is your map?

The final map will be 1:5,000,000. I tried to reproject the image, but to no avail. I'm in the process of photoshopping the original DEM we were using before, with better results than I expected. The land looks okay, but the water needs a lot of work.