Cartographer's Toolkit

Map Making Tips, Tricks, and Inspiration

GIS Maps in the Wild

August 15th, 2014

For the most part, GIS map apps are inflicted on the public without user testing. User experience with the map app is also not studied after the fact, as a general rule. These things, which have the potential to affect thousands of people, are essentially designed and deployed in a black-hole environment where the best one can hope for is to get a few bits of feedback from colleagues on twitter or from a desk-mate who might simply state, “that’s cool.”

A stellar-mass black hole in orbit with a companion star located about 6,000 light years from Earth.

This is why I was excited to see the mention of a map application on a totally non-map related forum the other day and even more excited to see that there were a lot of responses regarding not only the content of the map but also regarding the usability of the map. What follows is a list I’ve made of what those responses consisted of, inasmuch as they provide instructive commentary for cartographers to apply to our mapping applications in general.

Our Case Study:

The map app

The forum discussion (non-mappers)


STATE THE DATA ORIGINS The mapping application has a 5 sentence paragraph defining how the mapped variable was calculated. This paragraph was also quoted in the original forum post. Even though the paragraph explained that the data was from a “Survey of Consumer Finances,” most people seemed to assume that the data was purely census data, based on the fact that census blocks were used as the geographic unit of display. A forum poster pointed out the source of the data later on in the discussion as people wondered how the census data was used to calculate net worth (it wasn’t.) Lesson: be clear about where the data comes from; people do want to know so don’t bury the information in a long paragraph.

SCRUTINIZE EVERY WORD The instruction to “click on the block” to see details about census data for that block was interpreted by at least one reader as “block” as in a 3d square object. Some readers were therefore not familiar with the census block concept. The lesson is that every word should be examined for alternate meanings, geographic differences, and level of reader knowledge about the subject. You don’t have to dumb it down to kindergarten reading level, but you do have to be cognizant of not using industry-specific terms for a general-public map.

ANTICIPATE AREAL UNIT SKEPTICISM Geographic units will be questioned by smart people. In this case, some people were unfamiliar with census block geography and wondered if the block polygons were always square, leading them to immediately wonder if there was a spatial bias. Of course geographers know there is always some kind of bias with spatial units. (They imply that things are more similar within than without as if there were hard edges between two adjacent units when in reality most data such as soils or census data is continuously distributed, is just one example of the problems inherent in geographic units.) But non-geographers don’t know about the modifiable areal unit problem. I think that there isn’t much a geo map app can do to acknowledge this bias other than with wordy disclaimers, but it remains a point to consider mentioning when you present the map in person.

CLICK-FOR-DATA NOT ALWAYS OBVIOUS A few people seemed to not realize that you could get at detailed information about the blocks by clicking on them. To describe how to do it one responder simply explained, “it shows up when you click on the colors.” The obvious to you is not always obvious to others. When explaining how to do something on a map app, you can make the wording quite simple, as was done here.

MAPS ARE FOR PONDERING Starting a conversation with the map and getting people to think about the data is a good end in itself. In this case there was a good discussion about whether the data reflected the net worth of renters or owners and how people with multiple homes were represented.

THEY ALWAYS WANT MORE DATA People will always want more specificity with data. In our example people were eager to know more specifically what the net worth was in places where it was listed as “$500,001,” which was actually just the top of the data range. While there may be very few census blocks with net worths higher than $500,000, finding those few census blocks (the extreme outliers) might be the most interesting thing about the map. The way in which a choropleth is broken down is therefore extremely important to think through thoroughly. For example, for this data and this use case, where outliers are important, the quantile method of class breaks is especially problematic.

ZOOM LEVEL AGGREGATION IS GOOD Several users noted their appreciation for zoom level based geographic aggregation. For example, census tract boundaries are shown at lower zooms while smaller census block boundaries are shown at higher zooms.

ETHICS MUST BE CONSIDERED The notion of data being scary when it isn’t aggregated–when it’s available at a household level, for example–was brought up. All map makers need to be aware of ethical practices. Map app philosophy, however, needs a Socrates for the profession, and until the time when we have an overriding philosophy we have to make conclusions on ethical practices on a case by case basis.

DISCLAIMER FOR THIS LIST Lastly, while reading through the forum comments about the map, be mindful that the majority of people who used the map might have had zero trouble using it, zero trouble mentally thinking through and overcoming geographic unit bias, and zero trouble with data privacy concerns. To adequately understand the spread of user reactions we’d have to do a real study. Thus the disclaimer: empirical data is instructive but use it with caution.


Friday Roundup*

August 10th, 2014

*Written on Sunday. Because I can.

  • Jennifer Davey’s Fort Collins Art Map is on display at the Fort Collins Museum of Art until September 28. It’s an aerial painting of Fort Collins, painted on tiles (yes, literally, tiles). There’s a nifty “making of” video here. I was fortunate to be out on the town with Angel Kwiatkowski–Cohere owner and map-tile idea maker herself–the other night and got to see it in person:FCmaptiles
  • Seeing and experiencing new works of art is always inspiring for map makers, and if you’re into Chihuli, you’ll love the Denver Botanic Gardens’ exhibition that goes on from now until November 30, 2014. The gardens are of course amazing, and adding in bold-colored Chihuli glass just makes them amazing-er. Interspersed with the grape vines laden with fruit, the xeriscape path, and the Japanese Bonsai are: a boat filled to the brim with blue and purple party shapes, a stream boasting light blue glass bubbles, a pink and white polka-dotted “tree” at the end of an allée, and red non-menacing spears peaking out from afar, among others. Lessons for map making? Color groups please viewers, never underestimate the value of a sensational centerpiece idea, and provide your data in context. Also, I might add that I actually used a paper map of the gardens the two days I visited. Paper!chihuli
  • The other day I was very surprised to get a box from Dave Imus ( I think you’ll remember him as the creator of The Greatest Paper Map of The United States You’ll Ever See) and completely touched to find in it a framed print of his beautiful Chesapeake Bay Watershed map. It’s so good that I’m replacing the Maroon Bells aspen-tree hiking photo in my office with the Imus map. And that’s saying a lot. Thanks Dave, it will provide inspiration for many years to come!imus

The Books at the Esri User Conference

August 1st, 2014

Cartographer’s Toolkit was in the top 50 books sold at the Esri UC out of 236 titles on display there. Cartography is usually considered a niche even in the world of GIS. So this means that we are getting a lot more people on board with the “make better maps” mantra. I’m thrilled with this result.

Also, the funniest thing happened. One of the books that Esri ordered for the conference, and only 1 out of an entire box, had the cover of a different book on it! What do you think? It’s a beautiful cover and now I’m wishing–if only it were a little more map oriented–that it really was the cover for Cartographer’s Toolkit. :) The inside is the real inside to the book, only the cover is wrong.

Wrong cover, correct insides

Friday Thoughts on Cartography

July 18th, 2014

Today a few of us decided to create a new organization called Emotional Cartographers Anonymous. While I am decidedly not an emotional cartographer–in fact there was another faction arguing for a Rational Cartographer’s Guild or some sort that I would be a reasonable fit for–I think it would be rather fun to attend an ECA meeting. I imagine there would be lots of shouting concerning the acceptability of halos on labels, old-school vs. new-school color schemes (think light vs. dark), and of course an open source vs. proprietary discussion in which someone inevitably brings up child labor. Or something.

And when we’re done with the meeting we can all gather around and have some map cake, which won’t make us upset at all.

USA Cake

All kidding aside, I also managed to publicly declare today that: You really have to understand the infrastructure underlying your map to be an effective cartographer. This statement received a fair bit of attention on twitter. The reason I’m pointing it out is that it may seem like a departure from my oft-repeated stance that to be a good cartographer you have to learn the fundamentals of cartographic design independent of the software tools.

If you think about it though, these are not two opposing ideas. It is true that excellent cartography is based on a sound foundation of basic principles as well as application of both analytical and creative forces to the data and design. None of that has to do with software and shouldn’t be constrained by software.

If you want to create a certain visualization that is perfect for your purpose and you don’t have the ability to do that within your normal day-to-day software then you should seek software that does offer that functionality (or, if you’re a dev type, build it). In other words, I don’t have to teach software in order to teach cartographic principles.

Figure-ground differentiation, label-font hierarchy, balance of margin elements, choropleth color maximums, and many other principles remain the same regardless.

However, what I perhaps have not emphasized enough is that there’s also a need for the cartographer to understand exactly how features and maps are created, stored, styled, modified, updated, and published in order to have a maximum of control over stylistic changes. In short, if you can’t operate your map stack technology then you’ll be relying on others to do it for you, which will represent a significant delay in getting edits published. It also leaves a gap in your knowledge of what the possibilities/capabilities of your stack are.

So, learn cartographic principles. That’s key. Also learn data analysis, general design, and creative skills. Finally, learn your software, whether that’s traditional desktop GIS or an entire map stack comprised of a map server, a statistics package, a web mapping client, a tile cacher and so on.

This is a lot and you probably can’t be an expert in all of it. In fact, team effort is fully warranted, particularly if you are serving up complicated maps. For sure a lot of speed is garnered from a team that has a server expert, a styling expert, a design expert, a front-end developer, and so on. But a cartographer should be familiar enough with all these pieces to be able to at least ask the right questions and coordinate the experts to provide a substantial, informative, and timely map product.


Map Curmudgeon

July 2nd, 2014

It’s been 4 years since I started writing this blog. So that means I’m now old and curmudgeonly. This was startling apparent today when these three things occurred to me:

  • If you edit a publication with a circulation around say 1 million, and you include a picture of a person squatting on a table in what is ostensibly supposed to be a business-setting, then shame on you for sloppy work. As my highschool yearbook teacher would say, “nothing is illegal about printing that picture, but ethically, you will want to think hard about including a picture in a publication that might make the person in the picture embarassed.” Or maybe everyone thinks squatting on tables is perfectly accceptable business practice these days. Beats me.
  • In the same publication an article by a pre-eminent cartographer shouldn’t be composed of 5/6 ridiculous non-meaningful chatter about why maps are great and 1/6 description of why the author is great. How about giving us some real ideas about how to be better cartographers? At the least provide us with a modicum of value. Please.
  • Finding a dataset on something as simple and as ever un-changing as the Oregon Trail shouldn’t be difficult. Instead, you try to find, say a shapefile of the Oregon Trail (a single line, or perhaps a single line with a loop at the western most end) and you’ll wind up in a maze of government web pages where it as if each page is a government employee passing off the request to another (i.e., a link), who then immediately passes you off to another employee as if it isn’t their problem. That’s how hard it is to find a dataset with a single line of the Oregon Trail. And before you tell me that it’s available in ArcMap as a dataset you can get through their online service, I’ll tell you that indeed you can but it is utterly useless because you can’t make a local copy of it and you can’t even trace it with the tracing tool.

And because I’m not ever going to be eternally sour, let’s leave this rant with a triple set of tips on how to make your maps better:

  • Learn some digital cartography, I don’t care if it’s big-name online or open-source online, your clients/customers/constituents need you to know this. I know you’ve got your expertise in your niche. Maybe you are an expert in parcels for the county, an expert in salmon (raising hand), or a geologist who can use GIS to find the best archaeological sites, you still need to know how to make compelling (or at least usable) digital, zoomable maps. It’s actually quite difficult to learn this skill if you’re not a dev. But it’s doable (raising hand again) and pretty necessary.
  • Give us something other than sensational maps. Pop maps have had their day and we now crave intellectual, even sophisticated, if you will, maps that teach us what’s important today around the world.
  • A two-hour sequester of 2-4 people brainstorming how to take your county’s parcel map from passable to extremely useful is not to be underestimated.

There are lots of tips to impart but I also want to be wary of pedantry. There are plenty of absolutely amazing maps out there changing the world today. Let’s leave with the London Tube Map in 3d:


Making a basemap with OSM data, some notes

June 22nd, 2014

Creating a zoomable map with openstreetmap data that covers the entire world is doable even if you’re a single-person entity without your own server infrastructure. You’ll have to use a provider like AWS to scale up to the level of performance that you’ll need, of course. As with any cartography project, minding the data becomes about 50% of the work and multi-zoom basemaps of the world are certainly no exception to that rule.


A tool like imposm or osm2pgsql is usually required to parse data from a source like geofabrik (pbf). At Boundless we’ve had some great success with imposm3, though it is still in the experimental phase, but so far proves to be much faster than imposm2. The benefit of imposm is that you don’t dump everything into 3 tables like osm2pgsql; instead you dump it into any number of tables based on data type, usually around 25 tables (you customize the download via a json file with whatever parameters you want to specify). This makes querying faster. You also get “diff support,” which means that you can easily incorporate updates from OSM diff files, thus enabling easy updates however often you want to update. (We haven’t tested that in imposm3 yet.)

Going along with the “data is 50% of the project” maxim, you need to become familiar with PostGIS in order to deal with this data. You need to be able to use a viewer like pgAdmin and/or get familiar with the command-line tools both for viewing the tables and their contents so that you can actually use and style the data but also to manipulate the data if needed.

For example, performance is often enhanced if you create attribute indexes on any attributes that you are using for styling. Let’s say I’m showing parks at a high zoom level. My GeoServer SLD that contains the styling rules for my map might specify that type=parks in table landusages needs to be green with a dark green outline. The table landusages may benefit from an attribute index being created on the type field.


Personally, I believe the very best way to get familiar with all the tools that I’ve mentioned above is to play with them yourself in a real-world environment. For this my favorite two options are 1) learn on your own (I love Boundless’s new online training courses) and 2) attending a maptime near you. And on a final note, don’t try to download and parse the entire world’s worth of OSM data your first time. ;) Start with a city, county, or small state of interest and scale up from there.

Spending even a few nights beginning to fool around with OpenStreetMap, imposm2 or imposm3, PostGIS, and/or GeoServer will make you more marketable. Investing in this knowledge acquisition will certainly pay dividends.

Cartographer's Toolkit

Map Making Tips, Tricks, and Inspiration