Posted 12 November 2005 - 04:53 PM
So my problem is that I want to present and communicate different characteristics (precipitation, population, conflicts, land cover, etc etc) for only selected land areas (in this case deserts). Not necessarily for comparison between maps, but equal area is of course a big advantage (which is less important in this case, since the distortion in most projections are not so much around the equator). The land area outside is not of so much interest, only for communicating the size of the areas, sea area is not interesting at all.
For projections, platé carre will be used over my dead body (i am too much of a snob to use it!). What I look for is that people should easily interpret and read the map, and I am of the opinion that that shape and orientation is very important in a world map, which makes me less keen on exotic projections such as briesemeister, dymaxion etc. If I'd use an ordinary projection, I would use the equal area Goode's Interrupted Homolosine, which treats shape very well (the projection in the CartoTalk logo!).
What I was thinking of on my idea in cutting things out and moving them around was something like this:
mashup.jpg 7.21K 92 downloads
(quick and dirty mashup of the previous map)
I am not so sure about it after looking at this map though, but perhaps one can tweak it more...
For loads of examples on this presentational issue, see the UNEP-WCMC Mountain Watch report
Posted 13 November 2005 - 09:46 AM
If you choose an equal area but then move the continents closer together (is that what you are doing in the example?), you would be adding area distortion between the areas of data, with additional distortions to distance and angle. I think Goode's sounds like it might work; is there a reason you don't want to use it? It's made for showing those patch-connections you are exploring. Just take out the graticule so your readers don't know it's interrupted.
Is it just the matter of the oceans taking up too much space? Maybe I still don't understand what you are asking. If the distance between patches is the least important information, why show it as a world map in the first place?
p.s. I tried that Mountain Watch link but the link to the map report there won't load in my browser (Mozilla)...
Posted 16 November 2005 - 04:01 PM
It is equal area.
Posted 06 December 2005 - 03:25 AM
map from WRI, more info here
Posted 06 December 2005 - 10:34 AM
Here is another one with a similar presentational problem (only displaying population density for the world's coastlines). In this case, if I was the cartographer, I would have exaggerated the width of the coastal zone.
Hmm, the outline on the coast seems to make the darker colors for higher density less visible. For me the lowest category (2<) really stands out, but it's difficult to pick out the other classes. It could be easier to read if the map were bigger, but if that's the publication size it's a bit difficult.
Maybe they could've extended the bands into the water?
Posted 06 December 2005 - 10:43 AM
Posted 07 December 2005 - 04:34 AM
Variation in brightness, hue and saturation. Too much to bring into cardinal order, too near to differentiate.
Just stick to the variation of one variable, preferrably brightness. Or build meaningful qualitative classes and choose different hues of colour with same brighness and saturation.
I also think the coastlines should be broadened.
0 user(s) are reading this topic
0 members, 0 guests, 0 anonymous users