Political map of Japan & its territories
Posted 22 August 2008 - 01:20 AM
Even with somewhat greater resolution in the publication output a few islands will not be discernable -- Minamitorishima (Marcus I.) and Takeshima (Dokdo) are two examples. I could cheat, I suppose, and enlarge them by a factor of 2x or 3x.
I used World Vector Shoreline in Manifold, labeling and titling in Photoshop. I experimented with several projections before going back to Robinson. I discovered that Manifold's scale bar was wrong -- it needed to be stretched. I'll have to go back and check the others. Something to watch out for. A little worrisome that it would get something so fundamental wrong.
Not so many options in a map like this, but perhaps someone will have suggestions.
Posted 22 August 2008 - 02:36 AM
Aside from that, I'm not too keen on the red/magenta color use. I would keep red for Japan, as it's apparently associated with heroism in Japan and it's used in the flag. But the magenta circles for the runways clash with it.
Email: firstname.lastname@example.org / Twitter: @redgeographics
Posted 22 August 2008 - 06:28 AM
I would keep red for Japan, as it's apparently associated with heroism in Japan and it's used in the flag. But the magenta circles for the runways clash with it.
Well, red is the favorite color in Asia... I agree with Hans, the two colors are too alike to be discernible on the map and it clashed too...
Have you tried with a symbol, perhaps?
As for the scale, Robinson's projection is not equidistant... "Robinson, was essentially contracted to develop a map projection that did not maintain angle, direction, or limit distortion, but was sanctioned to produce a map projection that "looked good" for books and atlases" (source).
Projections used for world maps (like Robinson's) are not often equidistant (they could, but I think our public is not accustomed to seen them so they're not used that often). I have in front of me my NG maps of the world and the scale is 1:36,384,000 at the equator... Your Japan centered map was equidistant (or at least you azimuthal projection used to draw the circles), but only from Japan... Calculating the distance between Montreal and Paris on that map would be wrong... All that to say that you could use an conic equidistant projection centered on Japan and your scale should be more accurate
Posted 22 August 2008 - 08:33 AM
Posted 22 August 2008 - 08:55 AM
For distant islands, aviation facilities are of great importance. Availability of a runway compatible with jet airliners (as these are, at least up to Boeing 737 size) implies at least a potentially good conneciton with the rest of the world. Absence means deep isolation.
In response to the question of why magenta circles, the answer is that this is a standard symbology in aeronautical charts which will be immediately recognized by anyone with any aviation background. It is certainly true that unless one is seriously color blind, they definitely clash with the red masses!
I do think I want to stick with the red. I need a strong contrast with the background to maximize visibility of the small islands, and the red will not be a political problem in this case. (Colors always have political implications for varous groups.) I could use green, but that really would not do much to improve harmony with magenta and it would reduce contrast.
I imagine that magenta was adopted in the first place to improve visibility with red night lighting. It is also used for critical symbology (such as lights and buoys) in nautical charts. Even navigators recognize it as ugly!
I am thinking I will try dark blue for the circles. I don't think this would affect recognition too much and it should blend better. I'm reluctant to change to another symbol because the use of the circle is so well established. Communication first, aesthetics a very distant second is the rule in a case like this.
The scale issue is one which needs to be viewed in proper scale. I can't think of any projection which maintains constancy of scale across any substantial area. Indeed, I don't think it possible, even in principle.
It is possible to have scale be locally constant in all directions at any given point. This is a property of the Lambert Conformal Conic, which is why this is used for small-scale aviation charts such as the Jet Navigation Chart (JNC) series. I did try the LCC but it emphasizes the southern areas to a degree which seems undesirable for my purposes here.
The Robinson is not isotropic in distance at any point, but it is nearly so at every point in a sufficiently small region about the center of the projection, as here. Moreover, the distance is nearly constant over a region large enough to meet the needs of this map. So my scale will be within 5% or so of true throughout most of this map, which is sufficient for the uses to which it will be put.
Sophisticated users of course can judge scale and distortion from examination of the graticule, and use it rather than a scale bar as the gauge of distance. (In nautical charts, indeed, no scale bar is provided.) They will see immediately that the increasing non-rectilinearity of the graticule in the upper corners of the map implies that the scale is distorted in these regions. But they will also see that this does not affect the regions of most interest very much.
Posted 22 August 2008 - 09:01 PM
I did decide to blow up a few of the very small islands a little to make them at least slightly visible.
Thanks for the suggestions and comments.
0 user(s) are reading this topic
0 members, 0 guests, 0 anonymous users