Like any good project, the transit trail map idea originated in a hut. Obviously, I’d seen people taking liberties with transit maps (notably at the Transit Maps Tumblr) and the idea may have been aided by the utterly silly National Parks, the Appalachian Trail transit map, the uber-meta transit map of transit systems and pages compiling them. I was up with fellow geography nerd Andrew Riely visiting Lakes of the Clouds where the croo had a t-shirt showing, well, showing something German. All of the sudden, gears in my head slowly turned and I exclaimed to Andrew: “what if we did this for the trails in the Whites?!” I think he had the gist of the idea before I’d finished the sentence.
We copped a pencil and paper and scribbled a little at the hut, and then started exchanging hand-drawn maps by email. (By which I mean we sketched maps and then photographed them and sent them back and forth.) After some valiant pencil-and-paper work, it became clear that graphite and 8.5×11 paper was not going to cut it for this project, and that we’d probably have to use some sort of fancy computer program. Then I got busy with work, Andrew traveled abroad and then went back to teaching (geography, natch) and the project fell by the wayside.
With winter came the Birkie, a big ski race, and with it Birkieguide.com updates. I decided that making a transit-style map of the Birkie would be interesting, fun and a good test to see whether I could do something similar for the much more complicated trail network in the Whites. The result looks pretty decent (but still needs to be updated in the NYC Subway style), and almost as importantly, I realized that using Inkscape (sorry, Adobe, I’m not shelling out for Illustrator) it wasn’t all that hard to do! Quick, Mapman, to the nerd-mobile!
The original idea—the holy grail, if you will—was to create a trail map for the Presidential Range. That seemed like diving straight in to the deep end. The Presidentials range has two issues which make it particularly hard for this type of not-to-scale informational map. The first is the dramatically different densities between the Northern Presidentials and the area to the southeast of Mount Washington. The Northern Presidentials rise steeply from Randolph to Madison and Adams, gaining about 4000 feet in about 4000 trail miles, and the summer residents of Randolph, around the turn of the 20th century, scattered dozens of trails across the slopes. The south section has only a few trails in the long valleys of Rocky Branch and the Dry River—maybe half a dozen trails in the same area that, in the northern part of the range, there are 50.
A quick aside: Many of the trail names in the Northern Presidentials reflect the turn-of-the-century era during which they were built. For instance, the Airline trail is so-named because it follows a straight line from Adams to the valley; railroads which followed such routes were similarly named. In King Ravine, the trail splits in to a scramble over rocks called The Elevated and a scramble under them called The Subway. These trails were likely named by Bostonians and New Yorkers at a time when their cities’ transit systems were relatively new. (The Els are mostly gone from both cities today.) And many Randolph Mountain Club signs use “for” instead of “to” in directional signage. I believe there are signs which use both to and for, as in “to Brookside for Mount Madison.” This too comes straight from transit signage—old transit signs in Boston often used “for” in place of “to.” In fact, to often implied a direct connection, while “for” implied the final destination. (An RMC contact updates: “to goes directly there, for heads in that direction.”) Signs for the Red Line to Harvard and bus/streetcar connections beyond might be signed as “to Cambridge, for Watertown, Belmont and Arlington.” In other words, [take this train] to Cambridge, [transfer there] for [other destinations]. There’s a great photostream from MIT Libraries which includes some of this syntax, here are five good examples. The fact that the RMC signs emulate subway directional signs did not make me any less eager to present the trails in this manner.
An additional note on some of the trail names. As far as I know, the Northern Presidentials trails are some of the only ones which are “ways” as opposed to trails or (less frequently), paths. The Valley Way is probably the most well-traveled of the Ways, but they also include the Sylvan Way, the Beechwood Way, the Fallsway, Monaway and Cliffway. They were likely named at the same time as the roads along the Emerald Necklace were built in Boston—the Fenway, Riverway, Jamaicaway and Arborway (and the Lynnway and others)—for carriages going from the city to Olmsted’s parks. (Good old Frederick Law also designed the roadways.) Outside of these trails in and near Appalachia, I know of no other hiking trails which go by “way.”
In any case, the second issue is that the Northern Presidentials are so gosh-darned dense! It’s daunting to look at a map and try to parse out all the different trails when they intersect with each other, at times, every couple hundred yards. I knew it would make for an interesting transit-style map. It just didn’t seem like the place to start. So I took baby steps. I started in the Franconia-Pemigewasset area. No part of it is as dense as the Northern Presis, and, for the most part, its trail density is uniform.
I started with the Old Bridle Path, since it is the path I knew most about (working at Greenleaf Hut I hiked it several times a week for two months, sometimes logging multiple round trips in a day) and it would have more landmarks than others (hut pack trails have many named features as the croos trek up and down the same path dozens of times, and I was glad to throw a couple easter eggs in the map). I pretty quickly hit on some graphic guidelines. Trails should only go at right angles or at 45 degree angles. All trail intersections should be marked with a “station” so as to not worry about hierarchy of lines (i.e. which line crosses over or under). I settled on using two dot types, black-with-a-white-border (local stops on the New York City subway map) for most features and intersections and white-with-a-black-border for developed facilities (campgrounds, shelters, huts). And I began to sketch out the trails.
At first, the colors I used were easy, but hideous. To change colors, I’d simply grab the color sliders and slide them all the way one way or another, so the colors I used were red, green, blue, cyan, magenta and yellow. And black. (RGB, CMYK for you graphically-inclined people out there.) It was easy, and it made your eyes burn. But that didn’t deter me. I decided on symbology for the Appalachian Trail (a black outline) and that it would be the only trail with a different hierarchy; trying to decide which trails were more and less important and varying widths accordingly seemed arbitrary.
Trails were colored arbitrarily. I thought of using a few colors and coloring the trails by blaze color (but then there would just be a ton of blue trails) or by trail maintenance organization (but then it would be mostly whatever colors I picked for the AMC and the USFS). This would, in a sense, emulate the early maps of the unified New York Subway, when each operating division (the IRT, BMT and IND) was colored separately. (This type of map might work to show each trail separately and allow the colors to mean something, but it’s possible that the colors would be mostly confusing, and the short trails would be hard to show and label).
However, there are major issues with both these schemes. There is no compelling reason to color the trails by trail maintenance organization. Most hikers don’t particularly care if they are on an RMC trail or an AMC trail or a Forest Service trail or a trail that is a Forest Service trail but where maintenance has been contracted out to the AMC. Not only do most hikers not care, most likely don’t know. Some of us pay attention to sign colors and to the small emblems for each organization. Most people ignore that and pay attention to the distance left to the hut. As for blazes, this is slightly more useful information, but has a couple of issues. First, trails in the wilderness are unblazed, so how do you show them? Second, trails above treeline are not really blazed, but rather cairned (and when blazed trails ascend above treeline, they change from blazes to cairns). Finally, the AT is blazed white, side trails are blazed in blue, and a few other trails which don’t connect to the AT are blazed in yellow. You therefore use only three colors, and most trails are colored blue. Add to the fact that it would be really hard to differentiate between different trails when they crossed; in the current iteration there is no instance where two trails of the same color intersect. It would certainly be fun to create a version of the map with trails colored by maintenance organization and by blaze color, but the main version needs more than three colors, and needs to be differentiated by something more than who built the trail, and what color they slapped on the trees.
I settled on using the six colors—there is no junction in the Whites between more than six trails, it turns out. (In case you are wondering, the one instance of this occurs at Madison Spring Hut, where the Gulfside, Star Lake and Osgood trails intersect the Airline Cutoff, Valley Way and Pine Link. Which explains why, when I worked there, confused guests and passersby frequently inquired as to which trail began in which direction.) Because of the number of trails, and particularly because trails which were obvious continuances of each other change names, the other schemes would make it difficult to denote where one trail ended and another began. And, yes, I began the whole Presidential Range project mapping Madison Spring Hut and going out from there.
A further question: how much of a sense of place and direction to include. Should I throw direction to the wind and have north-south trails appear to head east-west? Should I fully compress sparsely-trailed areas to minimize wasted space? Or should I attempt to mimic the actual lay of the land, but with the stylized transit map? I settled somewhere near the former. I’d do my best to show the network first, but if there was room to maneuver, I’d certainly draw trails to best show how they actually lie on the ground.
Then I had to decide what non-trail features to include. Mountains would be included as features. Should I put in rivers? (No—hard to decide which to include and which not to, and I could show major bridges and fords as features.) Roads? Yes, as white (originally gray; white mimics the NY Transit map) lines much narrower than trails, but only “relevant” roads which led to trails and trailheads. Towns? Yes, with a different symbol (I settled on a white diamond). Then came labels. Towns were capitalized and horizontal, roads were italicized and placed along the roadway. Features were labeled at 45° angles (this, too, would change), with facilities (express station symbology) capitalized. Then I added in the wilderness under the map. Tah-dah!
Well, not really tah-dah. I still had to label the trails. My idea, which wound up working the first time, was to use New York-style bullets for trail names, but with two or three letters in each to denote the trail name, and a key below the map for each. (I’d done this, in a very small scale, with the aforementioned Birkie map.) So, if you looked at the map and saw a green “GL”, you could match it with the GL in the key and find out it was the Greenleaf Trail. Or if you found the Greenleaf Trail in the key, you could find the green “GL” on the map. I could have multiple trails with the same color, and even multiple letter pairs as long as the trails were different colors (i.e. SR is for both Sawyer River and Signal Ridge). And, when necessary, I could use three letters for clarity (or trails where the hut croo refers to them solely by three-letter acronyms, like the OBP and the GRT) and still (sort-of) fit them in the “bullet.”
This worked quite well. Shorter trails would have one bullet, longer ones would have it repeated along the trail. I had thought about a Vignelli Map (a separate-line-for-each-trail would be fun but might be very confusing), but it works well because of the trunk-and-branch New York Subway network, which doesn’t exist in the Whites. (The newer Kick Map is another potential template.) And I didn’t use the London Underground template because it seems best for a lot of intermediate stations (I would later change this), while the circle scheme of the New York map seems best for many intersection trails. I went with New York, with a nod to the 45 degree angles of Paris. For now.
So I had a perfectly functional map that was a horror show to look at. I sent it out to some friends and one response was: “It is a little painful right now to see all the variation of high octane neons.” Well, yeah. So I downloaded a color picker broser add-in and opened the New York Subway map and started sampling. I wound up with six trail colors out of a possible eight, deciding not to use light green (the G train) and brown (the J and Z trains). Then I sampled Central Park (the Wilderness) and the background. Oh, and to pull out the Appalachian Trail a bit more I gave it a white background extending beyond the black edges, then further refined it by giving it a white outline surrounded by a black outline.
I thought about making the AT white. It is white-blazed, after all, for over 2000 miles. However, in the Whites, it follows historic, existing trails, and it seemed uncouth to lose the richness of them in order to have a single-colored path snaking across the map. (Nearly all of these trails were built decades before the AT was conceived.) In the Pemigewasset, the AT follow five trails, in the Presidentials, it traverses nine! But since it’s really the only trail which can be objectively placed in a hierarchy, I had to highlight it somehow. (Regarding the hierarchy, I thought about making certain trails narrower and wider, but there was no good way to decide which would be higher and lower. For instance, the Valley Way would be wider since it’s the main pack trail to Madison. But what about the Airline, which is sometimes used for packing? Or the Crawford Path, which is part of two pack trails (Mizpah and Lakes) but only for portions of the path? And then what of a trail like the Randolph Path, which is heavily used in some sections and sees light usage in others? So I settled on the only hierarchy involving the AT.
And I then took what I’d learned in the Pemi and ran it over to the Presidentials. This was more daunting and there were a few times I had to explode portions of the map and start from scratch. And for a while most of the map was a hodgepodge of sketched-in black lines waiting to be colored, or removed, or rerouted. But using the Pemi template, the map was not too daunting, and once I had the Northern Presidentials filled in the rest fell pretty well in to place.
Another question was whether to use straight lines or curved ones. I am very much in the learning process for vector programs, and Inkscape in particular. So to start, I used Inkscape’s Bézier Curve function (no, I didn’t really know what it was, either) I changed most of the lines to curves (with minimal modification, and I left non-trail features as straight lines). And—I don’t really like it. It’s certainly different, but I sort of like the simplicity of the straight-line version. While the curves are nicer than the staid angles, they seem a bit too artsy for the scope of the project. (Another option would be to keep the straight lines, but smooth the angles. This would be the next iteration.)
After some busy times, I launched back in to the project. Actually, I began over with the Birkie map, for two reasons. First, it is much simpler than this one. And second, they have a t-shirt contest that ends on June 10, so if I want to submit for that (and I’m not entirely sure this is what they’re looking for, but it’s worth a shot, right?) I need to finish it up. So I spent some time with the Googles and figured out more of how the curves and nodes and handles work. It turns out that if I move the nodes by the same amount each time, I can get pretty-darned-good curved angles without losing the integrity of the map as above. Now, too bad there isn’t a function which lets you duplicate the handle for every node; I had to do this all by hand.
The next challenge is to deal with the labels. In the past iterations, I haphazardly had all trail-feature-related labels at 45 degree angles. This is, well, bad. I’ve since learned, from the Transit Maps guru’s blog, that I should have all labels horizontal if possible (probably not) and if not, have one angle, preferably 45 degrees rising left to right. So that is a bit of a challenge, which necessitates a few minor changes to the trails, and a bunch of changes to the text (bonus: it’s a good chance to get the text all well-aligned with the graphics, which is important!).
Oh, I counted the trails, too. On the Presidential Map alone, there are 99. This brings to mind any number of Jay-Z references. (It has increased to 100 or 101 as I’ve added small trail segments. Oh, well.)
Another question: I kept all non-trail symbology as straight lines with angular corners. I like it because it is in keeping with the hierarchy: those symbols look different because they are. Is this a good thing or a bad one?
There are a few other, minor tweaks. Like slightly shrinking three-letter text so it better fits in bullet points. And deciding if I want to keep overnight facilities capitalized or if the different symbol is enough hierarchy. And the very important step of changing all the fonts to Helvetica.
One other major change, recently, has been to go for a London-New York hybrid “station” symbology. For any trail intersection, I’ve kept the black circle. Which, in the scheme of a trail map, is a lot of the nodes. However, for on-trail features between intersections, I’ve gone with the London-esque pip symbology, which makes the map a bit cleaner.
The other major change has been to start creating trail diagrams for the back of the map. Many transit maps have line-by-line information (New York, Boston and Chicago, for example) and I wanted to do the same. So I am in the process (the painstaking process) of going through each line, creating a diagram, and adding information such as cross trails, features, elevations and distances. The product is great, but I’m expecting a dozen hours of work at least before it’s finished.
The plan is to refine these further, and then see about publication (and maybe start a Kickstarter to drum up funding). And, perhaps, expand the concept to other trail systems. Acadia? Dacks? Smokies? Grand Canyon? Yosemite? We’ll see.