4 replies to this topic

[rmcculley]

Well, these aren't my first illustrator maps, but they are my first illustrator maps since about 1995. My brother is moving from Spruce Grove to Kitimat in a couple of weeks, so I put together a quick map for him.


It is also one of my first attempts at using hillshades on imagery. I'm not completely satisfied with the colours, but not bad for the time I spent on it.

I didn't spend a lot of time on the maps at all, but I did like the control on label placement in illustrator.

 Kitimat_Summer.jpg   136.74K   288 downloads

This map is using Modis 32 day composite imagery from August 2005. Elevation data is Canadian Digital Elevation Data 1:250,000 dem's.

 Kitimat_Winter.jpg   157.68K   252 downloads

This map is using Modis 32 day composite imagery from December 2004. Elevation data is Canadian Digital Elevation Data 1:250,000 dem's.

[BEAVER]

What are the black labels? Are those town names? The image quality you posted is really low to see the terrain. Looks very blurry to me.

[Hans van der Maarel]

I agree, could you post a higher-resolution sample? Just a detail of a small area would be enough already.

[frax]

what is it that you are trying to present? The travel route, and what the nature would look like on the trip... ?

[Martin Gamache]

A few suggestions and comments (in order of complexity and time investment)

Combining satellite image data and relief can sometimes work and sometimes lead to very muddy results. I think you ended up in the latter. You would be better off using just the MODIS imagery in my opinion. You may be able to rotate the image and and have the map not be North up and get plenty of relief to show up from natural cast shadows. Or you may just want to use the relief data combined with a judiciously chosen hypsometic tinting scheme (I've posted ahout this recently in another thread). Either of these solutions would be preferable to the current approach.

Another option would be to use the MODIS derived landcover product or the MODIS Vegetation Continuous Fields. Combined with the relief you could get a very nice product. Or you may want to save yourself some work and just use Tom Patterson's Natural Earth dataset. You could download just the landcover data and combine it with your Canadian 250m relief image. That might loook very nice.

Another option if you dont like the MODIS team's classification approach is to classify your own MODIS imagery. MODIS' approach must take into consideration the entire globe and thus can overlook many local biomes/landcover type. A quick ISODATA classification of your scene with local knowledge of the area might lead to a perfectly acceptable landcover dataset.

As it is now I can neither see the relief nor the imagery and I would argue that this is clearly a case where less would be better. since I suspect this is a quick map product for a friend you probably dont want to invest time into option 3, all the other choices are likely as easy to implement as your original work and will yield a much nicer product.

mg