Geography 120 #1: Mapping the European Immigration Crisis

So I was excited this semester to be asked at the last minute to teach Geography 120, Geography of World Affairs, for the first time in a few years. We are doing webmaps for homework (what else would I do?)!

Our first project is to map the European Immigration Crisis, like this:


View Full Size

It is amazing how quickly this is growing and changing. Over the years, Ceuta and Melilla, the Spanish exclaves on the Moroccan coast drew all the attention. Last year, the dangerous crossing of the Mediterranean from Libya to Lampedusa and Italy by Africans reached crisis proportions. This summer, the story has been the Balkan route from Syria through Turkey, the Balkans, and Hungary has eclipsed all previous waves of immigration. What is next?

Maps + Lacrosse = Bliss

So I am involved neck-deep in Lacrosse (go Warriors!). Utah is currently in the midst of reorganizing its high school lacrosse league, and the prospect of state sanctioning is always a background issue. As I wondered about what is happening in other states, it suddenly occurred to me, “you’re a cartographer, for heaven’s sake: make a map!”

This map tries to portray three themes:

  1. Whether each state officially sanctions/sponsors boys lacrosse or not, gathered from each state high school activities association (I found a handy list at New Jersey’s site).
  2. How many teams each state had in Spring 2014, gathered from laxpower.com (I tried to only include school teams in sanctioned states, not club teams)
  3. How “prevalent” lacrosse is; that is, the ratio of the number of teams to the state population. If I could calculate the percentage of all high schools that play lacrosse, that would probably be a better measure, but I couldn’t find a table of total number of high schools by state without a lot of leg work.

Besides the obvious pattern of lacrosse being strongest in the Northeast, there are a number of interesting patterns here:

  • The much-hyped rapid growth in the South Atlantic (ACC country) is apparent.
  • Although lacrosse is generally stronger in Colorado than anywhere else in the West, it is actually more prevalent in Utah!
  • Several non-sanctioned states are “pregnant,” and should consider becoming official in the near future, especially Ohio, Utah, Oregon, and Washington (I know full well that the politics in each state are different).
  • What has happened in Minnesota? Did sanctioning lead to their great growth, or vice versa? Are they a model for other developing states to follow?

I welcome any comments, suggestions, corrections, and better data!

An Esri Basemap Tile Cache with No GIS Server?

So, I tasked my Geog 521R Web Mapping class with building a custom interactive web map, including a custom basemap tile cache. In previous semesters, I had a virtual machine running ArcGIS Server, but it was always a beast to keep running reliably (without me spending all my time as a sysadmin). I had thought of using ArcGIS Online, but we were having trouble making it work for students. I had the students investigate several other GIS tile servers, including open source tools, but they weren’t impressed with any.

It occurred to me that since the standard tile cache structure, as developed by Google, mirrors a set of folders and files, why do I need any cache server at all? Yes, the servers are good for regenerating the tiles, compressing the tiles to take up less space, and delivering the tiles with high performance; but they create another layer of complexity that is too cumbersome for simple applications like what my students were doing (i.e., only a few hundred tiles). In other applications, I have had success using Tilemill to design a basemap (which saves it as a .mbtiles file) then using a custom tool I found on the Internet to explode it to a file cache. In this case, I wanted to let the students use ArcGIS to design the basemap without having to learn another program. So here’s what we did:

  1. Design a basemap in ArcMap, using the best practices suggested by Esri, such as scale-dependent data, symbology and labeling
  2. Share it as a tile package, saving it as a file, rather than uploading to ArcGIS Online
  3. Use the “Export Tile Cache” tool to an exploded cache. Remember to edit the area of interest, and save it to a place on a web server (yes, you need a web server with enough space for the cache)
  4. For some mysterious reason, Esri’s exploded cache structure has all the right folders and files with filenames representing the zoom, row, and column numbers, but it uses hexadecimal numbers like R00002fcd (WHY???)
  5. The Leaflet web mapping API can load any basemap that uses the standard URL template, but doesn’t handle the hex numbers. However, there is a plug-in called FunctionalTileLayer that lets you define a function to generate the URL for each tile. With javascript base conversion functions, and some leading 0 padding, we can use the Esri exploded hex structure.

Here is an example of the result for Zion National Park, using the landscape expression technique I describe below. Feel free to look at the source to see how the Leaflet function works.

A Portrait Map

I have been working on techniques for designing maps that more effectively express the beauty and complexity of landscapes. Typically, maps and GIS are great at representing the facts on the ground (visible and invisible), but not so great at portraying the qualitative aspects, much like the difference between a photograph and a portrait. So far, I have been tinkering with some previously developed techniques, such as the sky model relief shading of Patrick Kennelly at Long Island University (his most recent paper is Kennelly & Stewart 2013) and the vegetation bump mapping originally developed by Jeff Nighbert at BLM-Oregon (a short background and some ArcGIS tools are at the Esri Mapping Center). I first implemented these in a “portrait map” of my favorite place, Zion National Park.

In my opinion, it’s a good first start. The Kennelly method (a very simplified form I implemented as a ArcGIS model) creates a relief image that portrays the incredible cliffs and canyons of Zion better than any other method I have seen. The Nighbert method (with some modification to fit the situation) shows the variety and complexity of Zion’s ecology in a simplified, understandable way. Combined with subtle representations of roads, buildings, water, and geology, and compositing in Photoshop and Illustrator (which have transparency tools far superior to ArcMap), we begin to get an appreciation for the physical and human landscape of one of the Earth’s greatest places. Here is the final Zion portrait map.

kolobcanyon

The primary inputs were the 1/3 arc second elevation data from The National Map, and 30m vegetation cover from the Southwest Regional Gap Analysis Project. Although Kennelly and Nighbert (in the links above) provide full GIS tools, here are the tools I implemented:

I realize that these are completely undocumented and would need to be considerably altered to work in other landscapes, but if you find them useful, great! I also created a custom tiled cache basemap from the image, but that is another story . . .

Optimal States of America?

One of the joys of being a geographer is the variety of things I’m interested in.  Do we have the “right” state boundaries? I recently submitted a paper about why 19th Century politicians drew our western state boundaries where they did; last year Neil Freeman’s map of 50 equal population states went viral; then recently there was some news about the latest proposal to carve up California (about the 300th so far :-) ). This got me thinking about what makes “good” state boundaries: put everyone with their nearest capital (voronoi polygons)? keep cultural/functional/physical regions together (watersheds, metro areas, etc.)? Equal population (and thus equal representation in Washington)? Cultural/historical relationships? So, I tried to do it myself: carve up the United States into “ideal” states, using some combination of these factors. As a policy proposal, this is of course completely pointless. We are not the U.K. or France: we can’t redraw state boundaries on a whim; this would never happen in a federal system, and I’m not convinced it ought to. However, it is a very interesting geographic experiment, getting us thinking about regions, relationships, and the relationship between government and everyday spatial life.

The factors I used in drawing this included:

  • Existing maps of functional regions, like the Neilson Designated Market Areas and Census metropolitan statistical areas
  • A formula I developed for predicting city hierarchies (a la Christaller) based on distances from a town to significantly larger towns
  • Physical features that unify (e.g. valleys, rivers) and divide (e.g. mountains, rivers) regions
  • Past citizen proposals to divide or alter states.
  • I tried not to include political & cultural inertia (the advantages of keeping the status quo, like the fact that people identify with their state, even if it doesn’t “make sense”), but it probably crept in here and there.

So check it out (sorry about the poor resolution–server limitation–see my facebook page for the full-resolution version). Did I get it right? What would you change?

optimalstates_small

What I did this summer . . .

Serpent Canyon, Zion N.P.

Serpent Canyon, Zion N.P.

Contrary to popular student belief, professors don’t spend their summers at home watching TV. Yes, I did take some fun trips to Philadelphia and San Diego (twice); well, I also spent a couple weeks (spread over 4-5 trips) hiking and building cabins in Zion National Park; oh, and I worked really hard on growing out my beard. BUT, I spent most of the summer writing a paper on geographic masses (as opposed to objects) and researching the mapping of textbooks onto the GIS&T Body of Knowledge. So there!

Another geo-historical narrative

If it isn’t obvious from my atlas, I am intrigued with the idea of using maps to tell a historical narrative. I’m also a sucker for atlases (especially historical atlases) focused on the places where I travel. This week I was in Philadelphia for the NCAA Lacrosse Championships (we were rooting for Cornell and Denver, but oh well), and while we were visiting Gettysburg, I picked up the brand new book Gettysburg: The Story of the Battle with Maps. It attempts to tell the very complex spatio-temporal story of the three-day battle with maps as the primary means of communication and the text in a supporting role. While they don’t quite accomplish that goal (I still had to carefully read the text to understand what is going on), it’s about as close to success as I’ve ever seen on paper. I’m not a civil war buff, nor am I very into military history, but I feel like I finally “get” Gettysburg after having worked through this book. Two techniques I really liked were:

  • Using the exact same base map (same area, same scale) and same symbology throughout to make it easy to compare one map with another.
  • Embedding “clips” of the map in the text to clearly tie specific parts of the text to the map

An impressive piece of work!

Mapping Mormonism wins Best Atlas of 2012!

Award presentation at Esri User Conference

Award presentation at Esri User Conference

I recently found out that Mapping Mormonism won Best Atlas in the 2012 Map Design Competition of the Cartography and Geographic Information Society (CaGIS), the most prestigious cartography awards in the U.S.! It turns out we were a tiebreaker away from Best of Show, but I’m happy to hear that someone likes the atlas. I only hope that CaGIS eventually updates their website so it shows something official. . .