Visual hierarchy

A lot of ink gets devoted to questions of visual hierarchy but, at its core, it is nothing more than the idea that the important part of a map should be more visible than the less-important parts. Introductory texts tend to illustrate the concept with wildly distorted examples in which the title of a map document, for example, occupies more of the page than the map itself. Beyond obvious cases like that, establishing an appropriate visual hierarchy can be tricky.

Here’s a black and white map of a county park in Kansas City. It’s intended to guide readers to some interesting architectural features in that park and to help the reader understand the historical and cultural significance of those features.

It’s a good start, but the lake is a bit vague. And, while I don’t like to add purely decorative elements to a map, this one needed a border to hold the image together. Borders — more technically, a neatline — are a bit controversial. On one hand, they are a standard design element of many maps. On the other hand, there is a notion in graphic design, advanced emphatically by the revered scholar of design Edward Tufte, that one should strive to maximize the use of graphical elements that communicate information and avoid the use of elements that are decorative. A border communicates very little; the map reader is perfectly capable of telling where the map ends without a box to define that border. Yet there is a small amount of value in marking the full extent of a map, if only to assure the reader that nothing was cut off. As a bit of a compromise, I added a very thin dashed border.

It’s a reasonably effective map. It identifies the places of architectural interest, gives the viewer a thumbnail sketch of the features and provides some indication of where in the park they are located. But there is something just a bit muddy about the design. As a thoughtful colleague, the cartographer Liz Thomas, immediately saw what I did not: there was no meaningful visual hierarchy to the map. Everything was of equal visual impact, so nothing stood out. Instead, it was a bit of a visual jumble. So, following Liz’s advice, I gave more visual weight to the illustrations of areas of architectural interest and less to the background map.

In this edition, everything but the notes on architectural features is grayed slightly, so it recedes, visually, behind the architectural notes. I also redrew the small illustrations, making sure to use the same thickness of pen for all the drawings, so the drawings would all have the same visual weight. Finally, I added small connector lines to link the architectural notes to the exact place at which a feature could be found in the park. Normally, adding these lines to a map that is already full of road and boundary lines would create an unnecessary visual jumble, but, with the roads and borders grayed slightly, there is no confusion between the connector lines and the map features. Also at Liz Thomas’s suggestion, I also made the title a bit bigger and spread it over three lines, so it does not disappear into the top left corner.

The use of a border that is purely decorative still rankled, so I modified the border to turn it into a calibrated measure of distance — denominated in kilometers, meters, miles and feet — essentially a scale bar that continues around the entire map.

The calibrated border is an adaptation of the border used on nautical charts published by the National Oceanic and Atmospheric Administration. These charts use a border marked in degrees and minutes, a centuries-old system of identifying locations anywhere on the Earth’s surface. [It’s the same system that is often referred to as “GPS coordinates,” as though latitude and longitude did not exist prior to the public deployment of GPS in 1993.] Experienced users of these charts often use the distance between the one-minute ticks on the chart border to measure distances on the chart, taking advantage of the fact that one minute of arc of latitude is almost exactly one nautical mile.


Lots of text

All maps tell a story. Sometimes it make sense to give the reader lots of freedom to find the story he or she wants from the map, but sometimes you want to point out one very specific story from the many stories that could be embedded in a map image. That’s the case with this project, a map with a lot of text and illustrations meant to point out one very specific narrative about the park shown in the map that follows.

This map began as a simple illustration to accompany an article about Wyandotte County Lake Park, telling the story of the park’s role in a national effort, during the Great Depression, to broaden the experience of national parks and to make some of the national park experience available to people who did not have the time or money to travel to the great parks of the mountain west, the Appalachians or coastal Maine. After the article was done, and a simple map was inserted, it became clear that the map was such an essential part of the story that it made more sense to turn the article’s text and illustrations into components of the map, rather than leaving the map as a component of the article.

A map like this, with lots of text, invites comparison with maps that deliberately use little or no text. Some of the best examples of such maps come from the National Park Service, the same agency that provided the design themes for Wyandotte County Lake Park. In the Park Service’s efforts to redesign their maps to make them more intuitive [Do a search on “Tom Patterson NPS” to reach some of the most thoughtful writing on map design ever published] they deliberately minimized the amount of text on their maps. That design made the maps simpler, easier to read, and was a service to the legions of foreign visitors to National Parks whose skill in reading English are limited.

Revisiting shoreline symbols

An earlier post explored ways to draw (or “symbolize” to use cartographer-speak) a shoreline in a monochrome map. A new style has some across my desk, suggesting it is time to revisit the topic.

The basic problem is how to communicate to a map reader both where a shoreline is located and which side of that line is land or water. One approach is to shade the land or water a medium gray. I’ve talked about this in a couple of earlier articles, exploring techniques that can be used with both GIS or drawing software as well as those that work with hand-painted ink wash. Shading creates some problems, however. Map users do not agree on whether the land or the water should be shaded, so there is a risk of confusing map readers. In addition, gray shading is often not handled well by printers; the gray that looks good on a computer screen can appear as much lighter or darker when printed. As a result, there is a lot to recommend techniques that rely not on shading but only on one-color lines. Some very traditional ways follow.

Stippling, or using little dots, to show the land adjacent to the shoreline. This technique has a certain intuitive clarity. The stippling creates a shaded effect for the land, suggesting that the entire land area is slightly darker than the adjoining water area. In addition, the stipples suggest a sandy shoreline, so hint to the map reader the presence of a shoreline. Of course, any suggestion of sand can also be a problem in places where sandy beaches are not present. For a map reader on the rocky coast of Maine, for example, a suggestion of sand is more confusing than helpful. Here is an example of a pen-and-ink map that uses stippling, plus a few little tree symbols to further indicate which side of the line is land.

Another technique is to use small line marks to indicate the shoreline. One style, popular in pen-and-ink maps a century of more ago, is to use a simple dash-dot-dot symbol along the shoreline. The dashes are drawn horizontally on the page, projecting from the shoreline into the water area. These marks have a nice intuitive appeal: they seem to suggest water getting deeper along the shoreline.

Another line technique is to draw fine lines parallel to the shoreline on the water side. Like the dash-dot technique described above, this was also popular in pen-and-ink mapping a century or more ago, but still has intuitive clarity today. Typically the space between the parallel lines increases as distance from the shore increases. Often the outermost lines are drawn with a thinner pen.

The technique of drawing lines parallel to the shoreline can be done with mapping software as well as pen-and-ink. The technique can be a bit tedious, even with software tools. Typically it involves creating “buffers” on one side of the shoreline, then symbolizing those buffers with a fine border. Here is the output from QGIS software:

The sailors and authors Lynda and David Chidell used a superb combination of techniques in their book Cutting the Dragon’s Tail. They stippled the land side of the shoreline, then used very subtle parallel lines on the water side, particularly around headlands. This subtle use of the parallel lines has even greater intuitive appeal than uniform parallel lines because is suggests waves breaking around headlands. Note as well the typeface that suggests the kind of lettering used on nineteenth century maps. With many thanks to David and Lynda Chidell for permission to share their work, the map follows.

Their technique can be adapted to larger scale maps (i.e., those showing a smaller area), as shown in this map of the same small Maine bay with which this discussion began:

This collection of different ways of showing something as simple as a shoreline lead to one worthwhile closing theme. In the cartography-speak, it is “redundant symbology.” In straightforward language, it is the idea that different people respond to different visual cues, so it’s a good idea to communicate the same information with more than one visual technique. For some people, some kind of shading will jump off the page as the difference between land and water. For others, parallel lines might be so evocative of waves lapping on a shore that they are the obvious clue to which side of a line is land and which is water. So, whenever it is possible to provide more than one visual cue without making the map too busy or distracting from other important information on the map (an important topic for another day), do it.

Prime meridians

We all learned about the equator and prime meridian somewhere in elementary school, and, I suspect, have lived very well without giving either idea much further thought.  In fact, except for map geeks and global navigators, the whole idea of latitude and longitude is of no particular importance.  The advent of GPS technology has given latitude and longitude a bit of a resurgence, as our small machines and various web sites are quick to offer the latitude and longitude of various places, whether of not those numbers make any sense to us.  In fact, GPS technology has even co-opted the language of latitude and longitude: it’s common to see latitude and longitude referred to as “GPS coordinates” as though they did not exist prior to the invention of satellite-based navigation systems.

But if GPS gets us thinking about latitude and longitude again, that’s for the best.  Like any system of angular measurement, latitude and longitude values only make sense if we know what zero means.  Obviously, when used with latitude and longitude, zero does not mean “nothing.”  A location at 0 degrees of latitude and 0 degrees of longitude still exists, it is not noplace.  So consider, as a bit of a refresher, what those zero values mean.

By convention, we measure latitude north or south from the equator, so the equator is the line of zero latitude.  And, also by convention, we measure longitude east or west from the prime meridian, a line of zero degrees of longitude that runs through Greenwich, England and is, therefore, sometimes called the Greenwich meridian.  Is it coming back from elementary school?  But the equator and the prime meridian are very different things.

The equator has some physical reality.  It’s the line around the “middle” of the earth, equidistant from the north and south poles.  And the north and south poles have physical reality; they are the places where the earth’s axis of rotation intersects the earth’s surface, the place where, were you to stand there, you would rotate like a top rather than rush forward as the earth spins.  Now, we need to equivocate a bit: the earth is not a perfect sphere, and the axis wobbles a bit, so defining the equator isn’t quite so unambiguous.  But the point remains: the idea of the equator is grounded in the geometry of the earth.

The prime meridian, however, has no similar geometrical reality.  A treaty among western industrial nations in the late 1800’s established an agreement to use a prime meridian that runs through Greenwich, England (where the Royal Observatory is located), but that location has no significance relative to the earth’s spheriod.  The prime meridian could have been drawn anywhere.  Prior to the treaty, it was drawn in lots of different places.  By way of example, there is a map of Europe in the lobby of an Austrian restaurant here in Kansas City.  When you look closely, you see that the prime meridian is not Greenwich, nor is it any European location, as you might expect for a map of Europe.  It’s Washington, DC.

So if the prime meridian can be anywhere, here is an alternative to consider.  For years, I made my living teaching geography at Haskell Indian Nations University, where issues of the sovereignty of indigenous nations is taken very seriously.  This example comes from some of those conversations.  It’s a map of the desert southwest of the United States, focusing on the Navajo Nation (and, inevitably, the Hopi as well, because their lands are surrounded by the Navajo lands).  The prime meridian, instead of being at Greenwich, England, is drawn to run through the Shiprock, a stunning geological feature and a place of significance in Navajo creation stories.

navajo national full grid illustration 150dpi


The Shiprock is a volcanic neck surrounded by dikes, channels that magma followed on its way to the surface of the earth, that stand, something like the mast and sails of a sailing ship, over the surrounding land.  The city of Shiprock, a center of administration for the Navajo Nation government, is nearby.  This map draws the prime meridian through the geologic and cultural feature, not the nearby city.

Illustrating Harlan Hubbard’s Shantyboat Travels

setting out map with text over topo 150dpi.png

setting out map over hillshade with text 150dpi.png



Just after the end of World War II, Harlan and Anna Hubbard built a shantyboat — a small, unpowered houseboat — on the banks of the Ohio River, then spent six years floating down the Ohio and Mississippi Rivers and exploring the bayou country west of New Orleans.  Harlan Hubbard wrote two books about their trips, Shantyboat, A River Way of Life and Shantyboat on the Bayous, both published by the University Press of Kentucky.  The books are revered for their description of river travel and as classics of the literature of simple living.

Both books are illustrated by a single map showing the entire span of the voyage.  However, the books provide no detailed maps of the specific locations where critical parts of the narrative take place, even though Hubbard makes clear his attention to the importance of specific places and the power of the land itself to shape the conduct of the people who inhabit it.  To address Hubbards concern for the characteristics of specific places, and to help me understand the geography that is important to the narrative, I made two maps of the locations around Brent, Kentucky that were critical to the development of Harlan and Anna Hubbards shantyboat life.

The Hubbards began their married life in a home and studio, built by Harlan, in the town of Fort Thomas, Kentucky.  But when they made the decision to build and live on a shantyboat, they chose a patch of Ohio River bottomland just outside Brent.  They built a rough cabin then, using lumber salvaged from a building in nearby Covington, spent about a year building their boat.  They lived aboard the boat for another year, finishing preparations for their trip.  Finally, just before Christmas in 1946, they cast off from Brent and began their trip downstream.

The map shows the locations where they salvaged the lumber, built the boat, and made camp on the first few days of their six-year trip.  The descriptions of those places come directly from Harlan Hubbard’s writing, and are used with the kind permission of the University Press of Kentucky.  More information about the books is available from the University Press of Kentucky at

Two versions of the map are shown.  The colorful version shows the Hubbard’s travels superimposed on U.S. Geological Survey topographic maps that were made while the Hubbards were traveling.  They show the cultural features — the cities, towns, railroads, ferries, quarries — that made the Ohio River a vibrant community in the middle of the twentieth century.  The gray map shows only the terrain and the river, so emphasizes the natural rather than cultural landscape through which the Hubbards traveled.

Junk rig sailboats in eastern Maine




The junk rig is a sail design generally associated with China and southeast Asia.  It’s characterized by a sail that extends in front of the mast, with stiffeners, called battens, that run the full width of the sail. The intricacies are well documented by the sailing community. See, for more detail, see the Junk Rig Association. The junk rig has been adopted by recreational sailors, who admire the inherent safety of such a sail. It is very easy to reef, or to reduce the area of the sail when increasing  winds make a large sail unsafe. The junk is also safer than many other rigs in a jibe, when the stern of a sailboat is pointed toward the wind and the sail quickly shifts from one side of the boat to the other.

These images are from a small bay in eastern Maine, where where junk rig aficionados from across the eastern half of the US have been gathering each fall. Both boats were originally built with more conventional rigs, then converted to the junk sail. The boat on the left is an American design intended for overnight cruising.  The one on the right is a Mirror Dinghy, a British design intended for day trips, and designed to be easy for inexperienced boat builders to assemble from plywood.

Digression to block prints, continued: Small workboats of eastern Maine

hard chine 100 dpisoft chine 100 dpi rotated.png

As the waters of the North Atlantic Ocean warm, the lobster industry of the east coast is declining. The fishery has contracted substantially off Long Island and the Massachusetts coast. However, in eastern Maine, it continues to thrive, though harvest levels have varied dramatically from year to year. Lobsters are trapped from coastal workboats, typically of between 30 and 50 feet, crewed by two or three people. Those crew use small skiffs to ferry themselves to and from their boats.

These prints show two typical skiffs.  One is flat bottomed so it is easy to pull up on a float or to just leave on a tide flat when not in use.  It is also convenient for transporting a bunch of supplies, because the load can rest of a flat surface.  The construction technique is simple: a flat bottom of plywood or planks, then wooden planks for the sides, meeting at a sharp angle. That angle is the chine; when the angle is sharp, it’s called a hard chine.  The other has a rounded bottom, so it is more difficult to build with hand tools and a bit less convenient for transporting supplies.  As fiberglass and molded plastic displaced wood as a construction material, inexpensive soft chine skiffs have become common tenders.


Propelling a small skiff

two lobstermen paddling no text 100 dpitwo lobstermen paddling corea 100dpi with text.png


Some crews use outboard motors to propel those skiffs, particularly in harbors that are open to strong winds. In sheltered harbors, however, crews typically paddle out to their boats. Instead of using a a traditional rowing position, in which the rower is seated facing the back of the boat to propel it forward, these lobster boat crew members typically stand near the bow of the boat and use a full-sized oar as a paddle to propel the boat stern first.
The skiffs are small and many lobstermen are big, so one person per skiff is a comfortable load. Once, however, I saw two big lobstermen paddling a skiff, backwards, in the harbor at Corea, Maine, as shown in this print.  Backwards rowing is a technique that has been used for many decades. Photographs dating back to at least the 1940s show individual lobstermen using this technique.   Drawings by Harlan Hubbard, who explored the Ohio and Mississippi Rivers in a home-built shantyboat, show Ohio River boatmen using the stand-up technique in 1946.


A further digression to block printing. Eastport, Lubec, and the sardine industry.

This is going to be a bit of a stretch, and a continuing digression from a site that is dedicated to map-making.  But there is a connection, however tenuous.  Back when I taught cartography to college students, I told my students that, in our class, we would work on making maps that:

  1. Speak the truth
  2. Tell an interesting story
  3. Are beautiful.

Ignoring number one and three for the moment, we can recognize that map making is a form of story telling.  Specifically, it’s a way of telling stories, graphically, about places.  The material that follows is also about telling stories graphically about places, but using a different set of tools: block prints and a bit of text.


Eastport, Lubec, and the sardine industry




For much of the late 19th and early 20th centuries, the American sardine industry was centered in the small Maine towns of Eastport and Lubec.  However, there was one big problem: there are no sardines in the western North Atlantic Ocean.  So the domestic sardine industry caught the plentiful local herring, put the small ones in cans and called them sardines, then smoked the larger fish and sold them in boxes as smoked herring.  French canneries, who used fish of the genus Sardina, were not amused, and tried to prevent American producers from using the word “sardine” on their product.   This block print illustrates the ambiguity of the industry, with the same fish carrying two separate names.

The industry, incidentally, brought great prosperity to several towns along the coast of Maine.  Eastport and Lubec enjoyed much of that prosperity, with dozens of canneries as well as associated industries such as can-making plants, cat food factories, mustard mills, and even cosmetics factories that used the tiny glistening herring scales to give makeup a sparkle.  The industry collapsed in the decades following the second world war as access to refrigeration made it easier for people to consume fresh fish and as tastes shifted away from stong-tasting fish like herring to mild fish like cod and pollack.  A remnant of the fishery remains, devoted primarily to catching adult herring for use as bait with which to catch the star of the eastern Maine maritime economy, the lobster.

For more of the story of Eastport and the sardine industry, see this article from Historical Geography.

Confluence – Tributaries of the Kansas River

perry kansas confluence.jpg

The Delaware River flows into the Kansas River, from the north, just below the small town of Perry, Kansas.  The confluence is primarily of interest to recreational boaters, for whom the trip down the Delaware to the Kansas then down the Kansas to an impoundment at Lawrence makes a pleasant day trip.  It is the arrival of the Delaware from the north that is  geologically noteworthy, and illustrative of an interesting pattern of the Kansas River watershed.

The Kansas river lies along the southern edge of its watershed.  All of its large tributaries, particularly along the lower stretch of the river, flow in from the north.  The only meaningful exception is the Wakarusa River, which flows parallel to the Kansas before making a short jog to join the larger river.  The Wakarusa flows so close to the Kansas that it is thought to have been an alternate channel when the more northerly channel was blocked by glacial ice.

The asymmetrical pattern of tributaries to the Kansas River results from the history of glaciation of the region.  At the height of the Pleistocene glaciers, the southern edge of the ice sheets stopped immediately adjacent to the course of the modern Kansas River.  The stopping point can be inferred by deposits of coarse glacial material and out-of-place boulders, called “erratics” along the Kansas River valley.  As those same glaciers melted, that meltwater formed streams that flowed south from the leading edge of the retreating glaciers toward the Kansas River.  Those streams were the ancestors to the Delaware and the other south-flowing modern tributaries to the Kansas River.


Confluence – Ohio and Erie Canal


miami and erie canal.jpg


The Miami and Erie Canal no longer exists. It was built between 1825 and 1845 to provide a route for barge traffic between Lake Erie near Toledo, Ohio and the Ohio River near Cincinnati. Today, some of the locks remain in place and isolated sections of the canal still hold water.

The surface of Lake Erie is slightly over 100 feet above the elevation of the Ohio River. The canal ascended the relatively gentle terrain of northwest Ohio until it reached St. Marys, where it climbed rapidly through a series of locks to the canal’s peak, the Loramie Summit, near the town of Bremen. Almost immediately, the canal dropped 67 feet through five locks at the town of Lockington. From there, it descended gradually to the Ohio River, a hundred miles to the south.

The canal rises above the elevation of the two largest water sources in the region — Lake Erie and the Ohio River — to a summit in the middle of the state, so it had no reliable source of water throughout the navigation season. Canal builders had to construct a series of reservoirs to provide water to the canal.

The Miami and Erie Canal was part of a wave of canal-building that swept the eastern half of the country in the first half of the nineteenth century, inspired in part by the success of the Erie Canal. In Ohio, the Miami and Erie Canal connected communities in the western half of the state, while the Ohio and Erie Canal was to connect the eastern part of the state with a water route from Lake Erie and Cleveland to the Ohio River at Portsmouth. In conjunction with the Erie Canal, these canals provided a water route from the middle of Ohio to New York City and, from there, to any seaport on Earth. The capacity of the Miami and Erie canal was low, however; the canal was designed with a working depth of four feet, so it could accommodate only small canal boats. It was built with a towpath for draft animals and was never modernized for use by motorized vessels. Instead, it passed into obsolescence and was abandoned as railroads were built across the state in the 1850s and 1860s.

Because is was built through sediment, rather than cut through bedrock, the canal had to be lined with clay or stone to keep water from leaking out. It also required constant maintenance, particularly dredging of accumulated loose sediment, to keep it usable. The outlet at Cincinnati was never easy to navigate, so much cargo was offloaded to horse-drawn wagons at that city and reloaded to Ohio River barges. After freight traffic began to move by rail rather than barge, the canal ceased to be commercially useful. Many segments were abandoned and are dry today, but some retain water and are used recreationally.