Acerca del autor

Mark Monmonier is distinguished professor of geography at Syracuse University's Maxwell School of Citizenship and Public Affairs and the author of many books, including, most recently, From Squaw Tit to Whorehouse Meadow, also published by the University of Chicago Press.

Fragmento. © Reproducción autorizada. Todos los derechos reservados.

COAST LINES

THE UNIVERSITY OF CHICAGO PRESS

Contents

PREFACE AND ACKNOWLEDGMENTS.....................................ix1 DEPICTION AND MEASUREMENT.....................................12 DEFINITIONS AND DELINEATIONS..................................133 NEW WORLDS AND FICTITIOUS ISLANDS.............................284 TRIANGLES AND TOPOGRAPHY......................................425 OVERHEAD IMAGING..............................................586 ELECTRONIC CHARTS AND PRECISE POSITIONING.....................707 GLOBAL SHORELINES.............................................868 BASELINES AND OFFSHORE BORDERS................................1029 CALIBRATING CATASTROPHE.......................................11610 RISING SEAS, ERODING SURGE...................................13111 CLOSE-UPS AND COMPLEXITY.....................................14712 EPILOGUE.....................................................163NOTES...........................................................167BIBLIOGRAPHY....................................................193INDEX...........................................................215

Chapter One

On a top-ten list of mapped features, the coastline is a shoo-in for first place. Because the sea provides food, transportation, and recreation, the shoreline is at once a boundary, an attraction, a source of livelihood, and a hazard. On maps it reproduces the distinctive shapes of Africa and Cape Cod, and in textbooks and science magazines it reconstructs continental drift and dramatizes the changing climate's rising seas. A challenge to mariners and marine scientists, the coastline is raised and lowered twice daily by tides and occasionally realigned by storms, which famously shorten the shelf life of nautical charts, on which its representation demands careful measurement and prudent compromise.

Not all maps have a coastline. Whether a map has one often depends on its scale, which cartographers define as the ratio of map distance to ground distance. At a scale of 1:1 a map would be as large as the territory represented, and hopelessly cumbersome. At 1:87 it would shrink a minuscule part of the world to the detail of an HO-gauge model railroad-smaller than reality but large enough to show sidewalks and sewer lines. Most maps have much smaller scales. At 1:24,000 the U.S. Geological Survey (USGS) topographic maps favored by hikers and earth scientists compress an acre of land into a tiny square barely a tenth of an inch on a side. Even so, mapmakers consider these maps large scale, while a 1:200,000,000 world map a mere eight inches across is unquestionably small scale. Small-scale world maps need coastlines to delineate continents, but less than half of all large-scale topographic maps include a stretch of seacoast.

Scale also affects the intricacy of cartographic coastlines, illustrated in this chapter by several maps for Five Islands, Maine, a tiny fishing village on Sheepscot Bay, thirty miles east of Portland. Marge and I rented a cottage there one summer when our daughter was young, and because we've been back a few times, I know the area well enough to appreciate the 1:24,000 USGS topographic map (fig. 1.1), which shows a settlement of roughly eighty buildings, mostly summer homes. (I enlarged this black-and-white excerpt to clarify symbols better differentiated in color on the original.) One of the tiny black rectangles represents our vacation rental, and another pinpoints the local "lobster pound" down near the dock, where the Thibodeau family sold the best take-out steamers on the Maine coast. Despite the simplified shoreline, I can recognize the narrow, rocky beach where Jo collected shells and flat pebbles. As a scale-model memory prompt, maps make excellent souvenirs.

Fine print at the bottom of the map sheet indicates that USGS mapmakers based their seaward coverage largely on the hydrographic measurements and shoreline features in figure 1.2, extracted from a nautical chart published at 1:15,000 by the U.S. Coast and Geodetic Survey, now the National Ocean Service (NOS) and part of the National Oceanic and Atmospheric Administration (NOAA). Intended for maritime users, charts present a comparatively sketchy view of the land, usually only for a narrow belt along the coast. In this example, topographic detail stops at a coast road a quarter mile inland-no point in cluttering the map with symbols of limited use to sailors. By contrast, the seaward portion of the chart is rich in soundings (depth measurements), bathymetric contours (lines of equal depth), and other symbols that warn of shallow water where a ship could run aground.

On nautical charts the coastline is more than just a line. Pictograms identify rocky areas submerged at high tide, and asterisks locate isolated rocks alternately covered and exposed. South of Malden Island the label "Foul" identifies an area where individual hazards remain uncharted and anchorage is risky. The USGS map (fig. 1.1) includes some of the soundings and most of the depth contours, but the NOS chart highlights unsafe water with a darker blue and ropes off dangerous zones with dotted lines. And while the topographic map names all four larger islands (Crow, Hen, Malden, and Mink), the chart identifies only the two that can be labeled without disrupting hydrographic detail. Trade-offs abound in mapmaking, and on a nautical chart hidden dangers trump feature names.

To illustrate more fully the effect of scale, I juxtaposed the shorelines from these two large-scale maps with corresponding delineations from topographic maps at 1:100,000 and 1:250,000. As figure 1.3 shows, the 1:15,000 nautical chart accommodates not only a more detailed shoreline than its 1:24,000 topographic counterpart but a few additional tiny islands as well. By contrast, at scales of 1:100,000 and 1:250,000, islets disappear and the shoreline becomes smoother and more rounded. At the smallest scale (fig. 1.3, lower right) the markedly distorted footprints of the four surviving islands are more nearly similar in size, while Hen and Malden Islands, which are close enough to be connected by a footbridge, were pushed farther apart for clarity.

Displacement of features is common to coastal maps because a line symbol 1/50-inch (0.51 millimeter) wide at 1:250,000 represents a band 417 feet across on the ground-nearly seventeen times its theoretical width at 1:15,000. This leaves the mapmaker with three choices: amalgamate the two islands, shift them apart, or drop the smaller one. The 1:15,000 chart's more generous canvas accommodates numerous twists, turns, protuberances, and indentations, while the 1:250,000 topographic map must shoehorn an area 278 times larger into the same size frame.

Features that survive at smaller scales are often exaggerated. For example, the narrow inlet at the top of the 1:15,000 and 1:24,000 depictions becomes a distinctly wider incursion at 1:250,000. Why the map includes this feature is not obvious. Perhaps the cartographer who compiled the 1:100,000 topographic map thought the indentation dispensable, while the compiler of the 1:250,000 map, aware that appearances matter on reference maps, enlarged it to make the coastline look right. The smaller the scale, the greater the cartographic license.

Verisimilitude is particularly important for Maine's numerous long, narrow embayments, aligned north-south and similar in origin to the deep, steep-sided, glacially eroded fjords of the Norwegian coast. Small-scale maps typically retain at least a few of these drowned valleys, suitably widened and evenly spaced, as classic examples of a generalization process cartographic theorists call exaggeration.

A similar rationale governs portrayal of the California coastline near San Francisco. As excerpts from the National Atlas of the United States show (fig. 1.4), small-scale maps avoid overlapping symbols by widening the inlet linking the Pacific Ocean with San Pablo Bay (to the north) and San Francisco Bay (to the south). At 1:7,500,000 and 1:17,000,000 the mapmaker saw fit to include the broad Sacramento-San Joaquin River Delta (to the east) and the prominent coastline near Point Reyes National Seashore (to the northwest), where narrow, linear Tomales Bay follows the well-known San Andreas Fault. This lesser feature drops out at 1:34,000,000 because a line symbol 1/100 inch wide would consume as much space as a corridor more than five miles across on the ground. At still smaller scales San Francisco's bays either drop out altogether or amalgamate into a tiny, barely noticeable dimple on an otherwise smooth coastline.

Because preferred routes and navigation hazards demand greater detail, chart makers employ a wider variety of scales than their topographic kinfolk, who rely on a comparatively uniform treatment to map a larger area for a more diverse clientele. As geographically tailored navigation instruments, harbor charts at scales between 1:5,000 and 1:20,000 promote safe access to ports, while in-shore navigation charts between 1:40,000 and 1:80,000 describe channels, hazards, and navigation aids for approach channels and longer stretches of coast, and off-shore navigation charts at 1:100,000 and smaller are sufficiently detailed for crossing oceans, seas, and gulfs.

Large-scale charts generally portray smaller, more focused areas than small-scale charts, and the area shown-what I call the map's geographic scope-is often chosen to encompass an entire estuary, port, or channel, usually identified in the chart's name. Because multiple map sheets can be inconvenient, if not confusing, nautical charts are often larger than topographic maps. For example, Chart 238, my source for figure 1.2, measures thirty-three by forty-four inches, whereas the Boothbay Harbor, Maine map sheet from which I extracted figure 1.1 is a mere twenty-two by twenty-seven inches.

Unlike their topographic counterparts, nautical charts vary in shape as well as size in order to fit meaningful chunks of the coastline onto a single sheet of paper. As figure 1.5 illustrates for coastal Maine, ease of use requires occasional overlap as well as supplementary, larger-scale charts for ports, rivers, and other areas where navigation is comparatively complex. By contrast, USGS topographic maps partition the country into nonoverlapping, one-size-fits-all quadrangles defined by an arbitrary grid of evenly spaced meridians and parallels, which all too often divide the area of interest among two or more map sheets. What's more, rigorous adherence to a uniform scale not only shortchanges areas for which fuller detail might be useful but also wastes space on largely featureless territory. An extreme example is the Rozel Point SW, Utah 1:24,000 quadrangle map, which portrays a rectangular chunk of the Great Salt Lake in solid blue. I taped a copy to my office door as an amusing example of cartographic obsession.

Map scale is not the only index of cartographic detail. A new measure emerged in the 1970s, when geographers began to study land cover with satellite imagery. Called "resolution" because it reflects an electronic sensor's ability to detect, or "resolve," features on the ground, this new index refers to the pixels (picture elements) of a digital image captured from a satellite or aircraft. Because pixels are usually square, resolution is expressed as the ground distance, in meters, along one side. Although an image can be displayed at any map scale, its resolution limits the level of detail, as illustrated in figure 1.6 by two overhead views of Five Islands. With a resolution of one meter, the aerial image on the left not only shows roads and clearings but also provides a sharper picture of the tidal flats than the satellite image to its right. Open water stands out in dark gray or solid black on the 28.5-meter (93.5-foot) satellite picture, while its coarse pixel grid blurs the shoreline like the fuzzed-out faces of confidential informants in a television documentary. At a much smaller scale, pixels would not be noticeable and the shoreline would seem well defined.

Though the two images in figure 1.6 resemble photographs, they are merely graphic renderings of numbers recorded for grid cells arranged in rows and columns. The left-hand image looks like an air photo because its grid is not obvious and its pixels portray the intensity of reflected sunlight largely as the human eye would see it. Look closely and you can see the wake of a speedboat cutting across parallel waves into deeper water. By contrast, the satellite image to its right records the relative intensity of infrared light reflected from the sea, the beach, clearings, and treetops. Invisible to human eyes, infrared distinctions would go unnoticed without infrared film or electronic sensors. Because water absorbs infrared radiation, pixels wholly offshore look black while light tones pinpoint healthy vegetation or bare sand, both highly efficient in reflecting infrared light.

A similar framework is used for digital elevation models (DEMs), which record elevations estimated for the centers of grid cells, typically thirty meters on a side. Like satellite images, gridded elevation data represent the shoreline as a zigzag boundary between onshore cells with a positive elevation and offshore cells, with elevations set to zero. When the resolution is a meter or less, or the map scale relatively small, a DEM provides a useful portrait of the coastline. But when the map scale is large relative to resolution, the display will look blocky, as in the right half of figure 1.6.

Map scale poses a serious limitation on measurements like the "95,000 miles of shoreline" that the Coast Guard cites when appealing for money or patience. A NexisLexis query of major newspapers and wire services found this estimate in over fifty news articles, 93 percent in post-9/11 stories on national security. Although 95,000 miles of coastline seems an intimidating responsibility, this figure is only one of several available estimates, which vary widely depending upon which maps one measures.

I traced the 95,000-mile figure to Aaron Shalowitz's Shore and Sea Boundaries, a massive compendium of legal and technical details published in 1964. Shalowitz was the Coast and Geodetic Survey's boundary guru in the postwar years, and even though he never mentioned the 95,000-mile total, readers can add together the 88,633 miles of "tidal shoreline (detailed)" tabulated by state, the 1,338 miles reported (but not totaled) in a table called "Areas Over Which the United States Exercises Sovereignty," and the 4,678 miles cited in a single paragraph labeled "Shoreline Along the Great Lakes." The resulting sum, 94,649 miles, rounds to 95,000.

The U.S. Coast and Geodetic Survey measured the country's 88,633-mile maritime boundary in 1939-40 using a "recording measure"-a small hand-held device that counts the revolutions of a tiny wheel run carefully along a line-"on the largest scale maps then available." Although the technique is not otherwise described, it was common practice in college map-analysis classes to make three measurements and take their average. If one measurement seemed an obvious blunder, it was thrown out and another taken. Nowadays we measure a curved line once, by summing up the individual lengths of its many short, straight segments, captured directly from aerial imagery or by carefully tracing delineations on existing maps and charts. The data and technology are available for revising the 1939-40 estimates, but replication is not a high priority. And because several federal agencies use the data to allocate funds among the states, re-measurement could be highly political were the numbers to change radically.

When NOAA rises to the task, as it surely will, noteworthy revisions seem likely insofar current charts are generally more detailed than their predecessors. And even though electronic calculation does not sidestep arbitrary decisions about how far inland to follow an estuary, modern technology allows experimentation with multiple definitions, impractical for prewar coastal cartographers intimidated less by political fallout than the tedium of manual measurement and the need for consistency. As Shalowitz described the process, the "shoreline of bays, sounds, and other bodies of water was included to the head of tidewater, or to a point where such waters narrowed to a width of 100 feet." At that point, the technician added a hundred feet to the running total and resumed measuring on the other side. A further qualification addressed long, narrow tidal rivers, which could inflate the results. "Both shores of a stream were measured if over 200 yards wide, but streams between 30 yards (100 feet) and 200 yards in width were measured as a single line through the middle of the stream." However straightforward, these round-number thresholds, in feet and yards, are blatantly arbitrary. If Thomas Jefferson had persuaded Congress, a century and a half earlier, to replace traditional English weights and measures with a decimal system, the criteria and the resulting measured lengths would surely be different. (Continues...)

Excerpted from COAST LINESby MARK MONMONIER Copyright © 2008 by Mark Monmonier. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.