Jump to content

 
Photo

Showing the centrality of Hawaii in the Pacific

- - - - - Pacific Ocean Hawaii

  • Please log in to reply
8 replies to this topic

#1
woneil

woneil

    Will O'Neil

  • Validated Member
  • PipPipPipPip
  • 129 posts
  • Gender:Male
  • Location:Falls Church, Virginia
  • United States

This is a map for use in presentations and a book on Peal Harbor. It's intended to emphasize the central position of Hawaii in the Pacific. The islands suffer economically from their great distance from land, but occupy a powerful strategic position. Constraints included that it be legible when printed on a 6"x9" (152x229mm) book page and that it cover approximately the longitude range shown.

 

The red circles are intended to give a sense of scale. Each represents an increment of 5 days' steaming from the Pearl Harbor sea buoy at 14 knots speed of advance, which would be typical of a fast naval task force.

 

I'd love to hear comments and suggestions about how to make it tell the story more clearly and attractively.

 

For background, the basic map was drawn in Versamap, an ancient vector mapper that came already provided with (inter alia) a cleaned up version of the still more ancient CIA World Databank in five resolution steps. It's very well suited to this task because, being vector rather than polygon oriented, it is entirely unconcerned about the antiprime meridian. Of course one can patch up world-coverage polygon sets to join at 180 deg, splitting at 0 deg, but it's a pain in the neck that I prefer to avoid when not otherwise too inconvenient. 

 

The layers generated in Versamap were output as high resolution PDFs for compositing and annotation in Photoshop. In the process i edited the modern boundary representation in the underlying data to approximate 1941 boundaries. I left off most of the Pacific islands because the seemed to clutter up the map too much, and did not add much that was of significance. The U.S.-owned islands scattered to westward from Hawaii are highlighted because of their role in screen it from easy approach.

 

I selected the Robinson as providing a remarkable balance between reasonably good shapes and reasonably good areas, even over this huge range of latitudes and longitudes.Again, to make it easy, I'd like to stick to projections that Versamap supports, but that covers a pretty wide range. 

Attached Thumbnails

  • Pacific.png

Will O'Neil
Author and amateur cartographer

http://analysis.williamdoneil.com/w.d.oneil@pobox.com

#2
David Medeiros

David Medeiros

    Hall of Fame

  • Moderator
  • PipPipPipPipPipPipPip
  • 1,312 posts
  • Gender:Male
  • Location:Redwood City CA
  • Interests:Cartography, wood working, wooden boats, fishing, camping, overland travel, exploring.
  • United States

I have a few thoughts on the general style:

 

1) There's more detail than you need in the coastlines for a map of this scale. Take a look at the Natural Earth downloads data for 1:50k map scales for simpler line work.

2) I don't think showing major rivers adds any thing to the map (unless specifically part of the text). They show too much detail and the color is a bit harsh. I'd remove them. If you need to keep them again look at the NE 1:50k data and then color match them with the oceans (same blue or darker blue in same general hue).

3) make sure your buffers (distance rings) are geodesic and not euclidian for true distances. 

4) You should have enough room to spell out Island in most cases.

5) Use true parentheses not brackets for your nationality sub labels and consider stacking them under the main label.

6) place your distance labels on the distance rings as path type, not horizontal type.  This is probably easier to do in Illustrator than Photoshop. Also, unify the distance label style, they say different things in diff styles at the various ranges. As much as possible avoid abbreviations.

7) It may be preferable for book publication to use a projection that allows you to fill a rectangular space with the background map as opposed to creating the less familiar shape (to most readers) than what is shown here.

8) coordinate labels. If this isn't a requirement I might drop the regular labels for something like just the equator. If you keep them be sure to label them the same way (so all in or all out of the map neatline

 

See Hawaii inset map here for somparison of NE 1:50k land data. Also example of differebnt projection that allows you to show the area as a rectangle:

https://en.wikipedia...i_State_Map.jpg


GIS Reference and Instruction Specialist, Stanford Geospatial Center.

www.mapbliss.com
@mapbliss

#3
woneil

woneil

    Will O'Neil

  • Validated Member
  • PipPipPipPip
  • 129 posts
  • Gender:Male
  • Location:Falls Church, Virginia
  • United States

Thanks very much for the suggestions. I'll be back with more specific responses after I have some time for a bit of experimentation.

 

One comment does mystify me, however. You say "3) make sure your buffers (distance rings) are geodesic and not euclidian [sic] for true distances." Surely you cannot mean to suggest that geodesics could possibly be suitable for the purpose of representing various distances on the surface, since the distance from the center to the geodesic line is by definition a quarter of the Earth's circumference. The circles shown are true small circles on the globe's surface, projected into the Robinson projection.That is they each represent a locus of constant distance from their common center. It is a virtue of this projection that small circles centered at a moderate distance from the Equator should appear so nearly circular to the uncritical eye when seen projected.

 

It is the kind of book that presupposes a reader who is not wholly unsophisticated.


Will O'Neil
Author and amateur cartographer

http://analysis.williamdoneil.com/w.d.oneil@pobox.com

#4
tangnar

tangnar

    Master Contributor

  • Validated Member
  • PipPipPipPip
  • 149 posts
  • Gender:Male
  • Location:South Carolina
  • Interests:Rivers, Urban Design, Geomorphology, Bicycling, Outdoors
  • United States

I like the brackets instead of the parenthesis... maintains a sort of 'military' look to me. If there's no reason other than aesthetics to change, I like it. 



#5
David Medeiros

David Medeiros

    Hall of Fame

  • Moderator
  • PipPipPipPipPipPipPip
  • 1,312 posts
  • Gender:Male
  • Location:Redwood City CA
  • Interests:Cartography, wood working, wooden boats, fishing, camping, overland travel, exploring.
  • United States

Thanks very much for the suggestions. I'll be back with more specific responses after I have some time for a bit of experimentation.

 

One comment does mystify me, however. You say "3) make sure your buffers (distance rings) are geodesic and not euclidian [sic] for true distances." Surely you cannot mean to suggest that geodesics could possibly be suitable for the purpose of representing various distances on the surface, since the distance from the center to the geodesic line is by definition a quarter of the Earth's circumference. The circles shown are true small circles on the globe's surface, projected into the Robinson projection.That is they each represent a locus of constant distance from their common center. It is a virtue of this projection that small circles centered at a moderate distance from the Equator should appear so nearly circular to the uncritical eye when seen projected.

 

It is the kind of book that presupposes a reader who is not wholly unsophisticated.

 

The difference between a geodesic and euclidean buffer is one of curvature. It is common for many GIS users to create a euclidean buffer, a flat distance ring away from a center point when they intended to create a geodesic buffer, one that accounts for the earths curvature when modeling distance. I wasn't suggesting that your buffers were incorrect (I didn't inspect them that closely). It was a general error checking comment not a critique, and as much for other readers as for you. The point was to call attention to the requirement that the rings be created as geodesic buffers to be accurate at this scale.

 

http://www.esri.com/...1/geodesic.html


GIS Reference and Instruction Specialist, Stanford Geospatial Center.

www.mapbliss.com
@mapbliss

#6
woneil

woneil

    Will O'Neil

  • Validated Member
  • PipPipPipPip
  • 129 posts
  • Gender:Male
  • Location:Falls Church, Virginia
  • United States

Thanks. I was confused because it appears that terminology in GIS diverges from that in spherical trigonometry and navigation, in which a locus of true spherical equidistance has for centuries been called a small circle rather than a buffer. Geodesic is a term reserved for a segment of a great circle in these fields. As a mathematician and sometime navigator I tend to gravitate to their terms.

 

 

 

Thanks very much for the suggestions. I'll be back with more specific responses after I have some time for a bit of experimentation.

 

One comment does mystify me, however. You say "3) make sure your buffers (distance rings) are geodesic and not euclidian [sic] for true distances." Surely you cannot mean to suggest that geodesics could possibly be suitable for the purpose of representing various distances on the surface, since the distance from the center to the geodesic line is by definition a quarter of the Earth's circumference. The circles shown are true small circles on the globe's surface, projected into the Robinson projection.That is they each represent a locus of constant distance from their common center. It is a virtue of this projection that small circles centered at a moderate distance from the Equator should appear so nearly circular to the uncritical eye when seen projected.

 

It is the kind of book that presupposes a reader who is not wholly unsophisticated.

 

The difference between a geodesic and euclidean buffer is one of curvature. It is common for many GIS users to create a euclidean buffer, a flat distance ring away from a center point when they intended to create a geodesic buffer, one that accounts for the earths curvature when modeling distance. I wasn't suggesting that your buffers were incorrect (I didn't inspect them that closely). It was a general error checking comment not a critique, and as much for other readers as for you. The point was to call attention to the requirement that the rings be created as geodesic buffers to be accurate at this scale.

 

http://www.esri.com/...1/geodesic.html

 


Will O'Neil
Author and amateur cartographer

http://analysis.williamdoneil.com/w.d.oneil@pobox.com

#7
woneil

woneil

    Will O'Neil

  • Validated Member
  • PipPipPipPip
  • 129 posts
  • Gender:Male
  • Location:Falls Church, Virginia
  • United States

As promised, I've been experimenting, starting of course with the projection issue. I'm made up the attached cartoons, showing various options scaled for the text block (outlined in red) of a 6" x 9" page. (Shown at smaller than full size.) One, the Behrman, is based on a rectilinear graticule while the others are cropped to fit a rectangular frame. I'd be interested in others' views. 

 

It's apparent from examination of the Tissot indicatrices that the Robinson does come closest to preserving both true shape and area overall, but the differences are not stark.

 

Note that I need space within the text block to add a caption, an essential for the internal economy of the book's organization.

Attached Thumbnails

  • Behrman-Cylindrical.png
  • Hammer-crop.png
  • Mollweide-crop.png
  • Orthographic-crop.png
  • Robinson-crop.png

Will O'Neil
Author and amateur cartographer

http://analysis.williamdoneil.com/w.d.oneil@pobox.com

#8
Dennis McClendon

Dennis McClendon

    Hall of Fame

  • Validated Member
  • PipPipPipPipPipPipPip
  • 1,172 posts
  • Gender:Male
  • Location:Chicago
  • Interests:map design, large-scale maps of cities
  • United States

I'm drawn to the orthographic because of how it instantly communicates the vastness of the Pacific.  So long as the Japanese Home Islands and California are clearly in sight, I don't think it's any great loss that the rest of Asia and North America are less visible.

 

And I agree that the rivers are irrelevant here.


Dennis McClendon, Chicago CartoGraphics
chicagocarto.com

#9
sitesatlas

sitesatlas

    Master Contributor

  • Validated Member
  • PipPipPipPip
  • 195 posts
  • Gender:Male
  • Location:Madrid, Spain
  • Spain

You might want to consider the azimuthal equidistant projection for a simple reason: if you center the projection on Pearl Harbor, the distance rings will actually appear as concentric circles and not ellipses. And, similar to the orthographic projection (which is also azimuthal), you'd get a very good sense of the vastness of the Pacific Ocean.

http://www.progonos..../shapePres.html


Michael Borop
World Sites Atlas
http://www.sitesatlas.com





Also tagged with one or more of these keywords: Pacific Ocean, Hawaii

0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users

-->