Saturday, March 7, 2009

The skill and art of map making

Through my interest in digital mapping resources related to biblical studies, I have had the pleasure of making the virtual acquaintance of A.D. Riddle. He studied ancient Near Eastern languages and history at Trinity Evangelical Divinity School (completed an MDiv and almost an MA as well). He is now at Univ of Wisconsin-Madison, working on a 1-yr graduate certificate in GIS & Cartography. This coming fall, he will return to TEDS for PhD work, again focusing on ANE languages and history. His goal is to teach historical geography and make maps of biblical and ANE history. In the meantime he has provided some resources for Todd Bolen of BiblePlaces.com. (Check out the Jerusalem PPT they made.)
A.D. shared with me his first professional map of which you can see a partial, quality-reduced sample above. If you want a map of “Hieroglyphic Luwian Inscriptions from the Iron Age (ca. 1200-700 B.C.): Find Locations in Northern Mesopotamia and Anatolia,” this is the map you want! More interesting to me, however, is the process of map making he used. Even in the small clip I show below, you can see an incredible amount of detail and subtle shading that gives it its professional look.
Here’s how A.D. describes the map making process:
>>>>>>>>>>>
This map was a final project for one of my cartography classes. I used four software applications to create the map: Google Earth, ArcGIS, Adobe Photoshop and Adobe Illustrator. The main feature of this map is the relief, which is shown using two techniques: changes in color (hypsometric tinting) and shading.

(1) First, using two books on Hieroglyphic Luwian, I located all the sites visually in Google Earth. Then I created a database of sites with the latitudes and longitudes obtained from Google Earth.
(2) I downloaded the other data from various websites. For example, the elevation information (a DEM, or Digital Elevation Model) comes from NASA's SRTM mission (http://www2.jpl.nasa.gov/srtm/). Each pixel in the raster, instead of containing color values, contains an elevation value.
(3) In ArcGIS, I overlaid the different data layers: streams, lakes, site locations, and the DEM. I added the DEM layer twice. One was used to create a hypsometric tint map. This is a map where changes in elevation are indicated by changes in color. I manually entered 13 classes that I wanted the elevation divided into, such as 0-100 meters, 100-200 meters, and so forth. I used smaller increments for lower elevations so that I could show more terrain detail below 1000 meters. The second DEM layer was used later to create the shaded relief in Photoshop. Finally in ArcGIS, all of the layers were projected, and I added the scalebar, north arrow, and inset map. Once it was all laid out on the page the way I wanted it, I exported the raster images (the hypsometric tint and the DEM) separately to Photoshop. The sites, streams, and other layers were exported to Illustrator.
(4) In Photoshop, I adjusted the colors of the hypsometric tinting and created the relief shading using a directional lighting source. Here is where the Mediterranean was colored blue and given a coastal “glow.”
(5) In Illustrator, all of the components were reunited. Labeling was added, streamlines were redrawn where necessary, and lakes were given a blue fill.

The final result was a 19" x 20" poster, printed at 300 dpi on glossy paper. After staring at pixels for so long on a screen, I was not quite sure how the printed version would actually look. But in the end, I was quite pleased with how it all turned out. Of course, I already see several things I would like to change about the map. For example, here are some things I noticed. (1) The projection of the inset map should be different, and maybe country labels should be added. (2) Some of the solitary mountain peaks should probably have a triangle symbol. (3) A legend should be included for the “mountain pass” symbols, the yellow dot symbols, and the mountain peak triangles. (4) Also, some of the yellow dots, even if precisely positioned, appear to sit right on top of rivers—these should be shifted off the river lines. (5) On flatter terrain, the shading produced a “ribbing” effect which needs to be smoothed out. (6) I need to draw a cleaner coastline for the Mediterranean Sea.

The most time-consuming parts of the project were (1) locating the places in Google Earth (most of these are very obscure sites), (2) getting my color scheme for the hypsometric elevation classes straightened out, and having to go back and forth between RGB and CMYK, and (3) creating the labels, adjusting their fill and stroke, and positioning each one just right. There was also a learning curve since I had never used PhotoShop before. I spent several hours with a professional cartographer who showed me a lot of nice techniques. I asked another experienced cartographer to check my map and give me feedback before submitting the final version.
>>>>>>>>>>>
That final version clocked in as a 57Mb file! If you would like to see a reduced PNG version that is only 17Mb, A.D. has graciously posted it HERE for your viewing pleasure. If you would like more info about A.D.'s work, you can contact him at riddle2 wisc edu. (You know how to fill in the blanks.)

My thanks--as well as my considerable awe--to A.D. for sharing this map and his map-making process!

3 comments:

  1. Very cool - except that using Google Earth data as part of the process means that the map has a restricted range of uses. You can't publish it in a book (commercial use), for example. (See the Google Earth Legal Notices page.)

    I'm not mentioning this to bash the guy who made the map, just pointing out that freedom is important - in this case, the freedom to make unrestricted derivative works. Even if we think it's unbalanced, we need to pay attention to copyright law. Perhaps either OpenAerialMap or OpenStreetMap might be a suitable source of alternative data?

    ReplyDelete
  2. @gerv - The data that was used from google earth, as stated in the post, was only lat/lon data.

    Google does not own the copyright on lat/lon data.

    The Google Earth Legal notice pertains to using images, such as a screen capture in a map. That would be a violation of their copyright.

    ReplyDelete
  3. I have attended copyright workshops and have tried to be as careful as possible about what data I use and how I use it. Permissions and restrictions on the use of digital data are notoriously fuzzy. This is especially true with mapmaking, where "one of the first and best rules in cartographic design is to beg, borrow, or steal good design. Many good ideas have already been had, and there's no point in ignoring them" (http://blogs.esri.com/Support/blogs/mappingcenter/archive/2008/10/28/just-and-right-u-s-maps-that-need-to-show-alaska-hawaii-and-more.aspx).

    As noted by mikeb, Google Earth was used only to acquire lat/lon coordinates. In addition to Google Earth, I also made use of the database available from National Geospatial-Intelligence Agency and U.S. Board on Geographic Names http://geonames.nga.mil/ggmagaz/geonames4.asp.

    Here is a helpful tool for working back and forth between kml and csv files (which can then be edited in Excel, for example).
    http://kmlcsv.pbworks.com/

    ReplyDelete