13 maps in 13 days: Joachim Ungar

Sending off the year 2015, we present to our readers the mapmakers who contributed their work to the 2015 GeoHipster calendar.

***

Joachim Ungar

Q: Tell us about yourself.

A: I’m a cartographer working at an Austrian IT company, named EOX IT Services, based in Vienna. We are mainly involved in the Earth Observation domain, most of the time being contracted by the European Space Agency (ESA). The first time I got in touch with GIS was in high school where we first heard about vector and raster data, projections, and so on. I liked it from the first day, so it was a rather easy decision to study Cartography and Geoinformation at the University. Back then I was rather drawn into data processing and learning more about the Open Source tools (GDAL, QGIS) than analyzing data or creating thematic maps with commercial tools.

Q: Tell us the story behind your map (what inspired you to make it, what did you learn while making it, or any other aspects of the map or its creation you would like people to know).

A: There were a couple of things coming together. I want to mention that we, at EOX, build an Open Source, back-end system for managing and serving EO data to portals for viewing and download, called EOxServer. Soon thereafter, we started to build clients as well and needed background maps. As we are committed to Open Source and Open Data, it was clear that we wouldn’t use one of the big commercial solutions (e.g. Google Maps). First we used a couple of voluntarily maintained OSM servers where we got our maps from, but we needed to have guaranteed uptime.

Mapbox was just starting off getting huge, but at this point I decided to get back to my cartography roots and do it myself. Moreover, ESA needed to have the maps in the WGS84 projection and not in Spherical Mercator, but almost all of the aforementioned solutions just supported the latter. But there was TileMill, which in my eyes revolutionized the process of styling maps for the web. I probably would have failed, mainly in keeping the motivation up, if TileMill (or something comparable) weren’t there already.

Downloading some SRTM data, creating a hillshade, and combining it with some OSM data shouldn’t be that hard, I thought. Well, let’s say I pretty soon found out this couldn’t be done in two weekends.

In the end it took us over a year and several iterations until we were able to publish Terrain and later Terrain Light. I was pretty obsessed for quite a while about all kinds of details. But it was worth it, I think. If you look close enough you’ll discover many neat details, like that the contour lines are light blue over ice, blue on the ocean floor and brown over land. We recently published a blog post about creating smooth centerlines from polygons if you are interested: https://eox.at/2015/12/curved-labels/

Of course we use EOX::Maps on a daily basis, and even ESA is now one of our customers, paying for the maps as a service. We also use it for our new product, called mapalupa (www.mapalupa.com), a tool where you can easily publish your global (or large area) data on your website, overlaid on an interactive globe using Cesium. Terrain Light is the default background layer there, and I think it works quite nicely on a globe as well.

However, there are always things to improve. So I guess what I have learned is that it can be very satisfying to map the world once, but you are never done doing so.

Q: Tell us about the tools, data, etc., you used to make the map.

A: For preprocessing I wrote some tools in Python which make use of GDAL, OGR, rasterio, and so on. For styling we used TileMill and some Mapnik hacks, so we could render the maps in a WGS84 projection. The tile cache is created and hosted by MapCache, a MapServer project.

For the hillshading we first used SRTM, but later switched to ASTER GDEM and filled some messy parts (bigger ice shields e.g. in Greenland) with GTOPO30. For the rest we used, of course, OpenStreetMap and some datasets from Natural Earth.

It’s great that there is so much data and software out there which can be used. Especially, I would like to mention the person from USGS who was extremely friendly and supportive when we asked to get the global ASTER data. We are not even US taxpayers, but this commitment to Open Data and customer support was very impressive. Last but not least I like to thank all people who spend their time and effort for the Open Source / Open Data idea and therefore provide so many powerful tools which help to compose and realize new things.