In 1983 Edward R. Tufte published an immensely influential book entitled The Visual Display of Quantitative Information. The gist of the book is that graphing data is (or should be) a sophisticated art. If you want to convey useful information to your reader, you have to think about the presentation, not just slap data on a page any which way. Tufte's main peeve is with graphs that present data inefficiently or ineffectively, so that you look at them and don't really know what they're saying. A related subject is that the same information can be presented graphically in different ways to make different points.
At right, I've shown a piece of the North America chart from the Light Pollution Atlas re-mapped to a conical projection. On looking at it, most people's first reaction is "My goodness, almost half of the U.S. is brilliantly lit." Whether intentional or not, that's a direct result of the colors chosen to represent different levels of light pollution. (From darkest to lightest, they're black, gray, blue, green, yellow, orange, red, and white.) The most prominent color in this map is green. That's not surprising, since the human eye is most sensitive to that color, and this is the greenest green that you computer monitor can produce, with all the green pixels firing at maximum, and all the other pixels turned off.
To my mind, this is somewhat misleading. Logically, the colors ought to go from darkest to brightest. But the orange zone appears distinctly darker than the green zone, belying the fact that skies are in fact 9× brighter in the orange than in the green. Shown at right is a very minor tweak, using darker versions of green and yellow. This makes the colors more of a continuum, and it's also more faithful to the underlying reality. In fact, the typical reaction of urbanites or suburbanites on first visiting the green zone is that skies are darker there than they ever imagined could be possible.
The differences may be more apparent in a closeup view. At right is an excerpt from the original map showing the northeastern U.S. and southeastern Canada. Detroit is on the left; Philadelphia and New York are near the bottom right of center; and Ottawa, Montreal, and Québec City (left to right) are above them near the top of the map. Prominently dark are the Great North Woods of Maine and the far greater woods of Canada. Less dark but also notable are the Adirondacks southwest of Montreal and the Alleghenies west-northwest of New York City. It looks as though New Yorkers have to go a long, long way to escape light pollution.
Here's the same region using my revised colors. This more or less matches my perception of the Northeast. It's now apparent that there's an almost unbroken corridor of reasonably dark sky — green at the worst — running down the backbone of the Appalachian Mountains. It's a long drive to a gray area from Boston or New York, but there are pockets of green — areas where the Milky Way is very attractive indeed — not too far away.
Here's a more drastic representation of the same data, which would probably seem appropriate to a typical backyard astronomer with little experience of dark skies. I've cranked green and yellow way down so that they look almost dark and boosted red to be a shade of white. Now the main contrast is between areas where the Milky Way is prominent (yellow and darker) and areas where the Milky Way is faint or invisible (red and white), with orange as a transition zone. This would be my guide if looking for a weekday observing site, where long drives are out of the question. Ideally, I would head for yellow or green, but in practice, I would just try to get out of red or white.
Which of these versions matches your perception of the light pollution around you?