Cartographer's Toolkit

The revision to City Maps: A coloring book for adults went live in all outlets the week of July 11, 2016. Did you know it can be ordered from your local bookstore as well as most major online retailers? You don’t have to order from the Amazon links here on this page, though I do earn some extra spare change if you do.

For this revision, almost all the maps were updated in some way or other. In fact, out of the 44 maps in the book, 36 were revised! A few of the maps had slight scale adjustments but for the most part they depict the exact same locations but with better line work. I’m thinking that taking a close look at what was accomplished in the revision could be super constructive to other cartographers facing similar issues.

TINY SPACES There were some tiny squares in some of the original book’s maps. Occasional complaints about the spaces being too small to color surfaced. To address this, I did eliminate many of the small polygons. For example, in the Venice maps, many of these small polygons represent courtyards within larger polygons representing buildings. Most of the smallest of these were eliminated as shown.

Venice Before

Venice After

In other cases I did leave in small polygons. For example, I was particularly enamored with the visual contrast between Tokyo’s Yoyogi park and the area directly surrounding it. I felt that not only were the small polygons necessary for the understanding of the location, they were also not too hard to color if you chose to make them all a single hue and colored them with broad strokes.

Tokyo’s Yoyogi Park, Small Buildings Not Removed In Order To Preserve Visual Contrast

CRISSCROSSING ROAD CASING (say that 3 times fast) One of the main differences between the original coloring book and this revision is the correction of cased road merge rules. QGIS aficionados should be alerted that this is resolved via the properties>style>advanced>symbol levels>enable symbol levels check box. This was an embarrassing error on my part as the original version did have a lot of clutter from this problem. My error was not in neglecting to notice it (oh boy did I ever notice it) but in not knowing that this was an easy fix. Live and learn and hopefully teach others how to do it better, that’s what I’m aiming for here!

Crisscrossing Casing

Crisscrossing Casing Corrected

NEIGHBORHOODS OSM contributors sometimes delineate the insides of city blocks with polygons that conflict slightly with the borders indicated by street casings. A similar issue is that in some cases polygons represent entire neighborhoods, which also can conflict with road and building lines and polygons by being slightly off geometrically. In most cases I eliminated these conflicts in the revised coloring book.

In this example, I show a before and after of London’s Kew Gardens area. They aren’t a 1:1 scale comparison because I felt it best to zoom into the area a bit for the revised map. But you can see how there are many line conflicts in the before image where overlapping polygons ended up looking visually like abnormally thick lines. There were many other issues fixed in this before/after slice as well.

Kew Catastrophe

Kew Corrected

Vancouver’s Granville Island map was a great example of this problem. The map was a mess due to the overlapping issue. This was resolved in the revision.

Granville Egad

Granville Great

Linewidths were also adjusted in the Granville revision. Not every issue was eliminated. As I write this I see some slight problems there in the upper-middle that weren’t resolved. Dang it! But overall this map in particular is about 100% better than before.

BRIDGES Bridges are represented in many different ways, it seems, in OSM data. I imagine that they are a problem especially due to the fact that OSM editors would naturally choose waterways as places to leave off their digitizing for the day, or to leave off their digitizing entirely. So in some places bridge lines don’t exist at all (e.g., the Manila port area) and in others the bridges don’t connect to the surrounding roadways in a seamless way. Where there were bridge problems, I corrected them by hand to the best of my ability as shown.

Botched Bridge

Better Bridge

Google streetview (keep in mind Google is in Mercator while the coloring book Paris map is in Lambert NTF EPSG 27561 so there is a slight skew in what shows up here vs. the maps above.)

CANALS The canals in Venice were “cut off” from the Grand Canal even though we intuitively think that a map should show them as “connected.” The OSM data has a single enclosed polygon representing the Grand Canal. This makes it easy for a geoanalyst to, for example, select just the Grand Canal for whatever purpose. However, a coloring book user has a different interest, namely, to color all the canals the same color. Therefore, I felt it was in the colorer’s best interest to change this representation in the revised book.

Cutoff Canals

Connected Canals

You might also notice some other issues fixed in the second image as well. Docks that previously appeared to be floating in the Grand Canal were connected to the land and that extraneous centerline at the lower left was eliminated (sometimes river centerlines are present in OSM data but aren’t at all interesting to a colorer).

FANCY ADDITIONS

I did add in some buildings and other features of interest as time permitted. For example, I drew in the Notre Dame Cathedral interior, whereas the exterior lines were all that showed before. I also drew in a better representation of the Sydney Opera House (which actually doesn’t look very geometric from the bird’s eye view regardless) and as shown in the below image, I drew in a dock or two where I thought it would be interesting and where it hadn’t been shown in the previous version.

Sydney Before

Sydney After

Okay, in this example, a lot of things actually changed. For one thing, I added cased road lines to the Sydney map instead of just single lines. For another, the dock is added in the lower left corner. A few buildings were added, the map location was adjusted slightly, and overlapping neighborhood and road lines were eliminated.

PARK PATHS Many park paths were changed from simple lines to cased path symbology, as in this instance. Only the main path lines were shown and the water features were added as well.

Central Park Paths Before

Central Park Paths After

IN CONCLUSION I’m hoping that this post is instructive to beginning cartographers in becoming familiar with some of the detailed symbology work that needs to be done to make a map pleasant and in fitting with the audience for the work. Things like whether or not paths and roads should be depicted as single lines or as cased lines with widths. Things like whether or not certain features should be visible or not. Things like fixing the data where needed either by adding in buildings, or taking out neighborhood boundaries, or connecting bridges to the land banks.

THOUGHT FOR THE DAY

The ability to make a map is great. The fortitude to correct it? Even better!

——————————————————————————————-

All the details you ever wanted regarding the revision to the City Maps coloring book, and probably more! https://t.co/o2jLNWePh8

— Gretchen Peterson (@PetersonGIS) July 25, 2016

Awesome work! These are the painful little cartographic details that separate good maps from great ones. https://t.co/aBYrLBVNZp

— Stephen Smith (@TheMapSmith) July 25, 2016