Wooden Topography Map | Written by Curtis
In 2016, while living in South Carolina, I stumbled upon a post on the image sharing site, imgur, that caught my eye. A woodworker had created a topographic map of the United States out of wood using a CNC machine. What’s more, he outlined his entire process in a reddit post, and later would create a small business producing even more of the Topographic states.
I saved the original post, and logged it in the back of my mind as something I’d like to do would I ever acquire access to a CNC Router.
Then in the Fall of 2020, in a scramble to finish a bookshelf that I couldn’t complete in my garage, I discovered the Loess Hills Maker Space, across the river from us in Glenwood, IA. And, in addition to every tool I would ever conceivably need, they had a large 5’x8′ CNC router. As I finished my bookshelf, I found excuses to return to the shop for small projects and slowly learned the programs behind the CNC – specifically VCarve.
My first project was to make a puzzle of the United States for my sister’s family as a Christmas present/education tool, as well as some small geography based ornaments and decorations. But really I was learning about the challenges of geometry and geography, namely the problem of meshing a concave and convex curve with a finite bit.
But once I figured out that I could definitely make it work, I set out in ernest recreating that topographic map. I’d love to say that I found my own way of converting raw USGS elevation data into usable raster files, but that simply wasn’t the case. After much trialing and erroring on the computer, I more or less mimicked the process that the original designer had devised. Using QGIS, I chose a map projection (US National Equal Area is so much more pleasing than Mercator), clipped the state shapes from the raw elevation data, and converted that data into a *.asc file. The most difficult portion was converting the *.asc files (or any type of geographic data file) into a *.dem raster file. Most programs required a rectangular field and would either choke on the data or create a null value that would break VCarve later on. But I eventually found a way around that didn’t involve manually editing megabytes of numbers.
Once I had *.dem files of each of the states, I could complete all the rest of the compression in VCarve. I knew that I wanted the final product to be around 4 feet tall and 6 feet wide and scaled the XY coordinates to approximate that (I think the final scale was close to 1:1750000). But, in terms of linear distance, the earth is very smooth relative to its size, and the highest point in the lower 48 at Mt. Whitney is less than 3 miles tall. Somewhat arbitrarily, and essentially by happenstance, I selected a Z scaling factor of about 10 giving a Z scale of very close to 1:175000, further emphasized by making sea level 10 mm high.
I made several test pieces of different sizes, learned that leaning on the table caused sufficient deflection to be visible, and then spent a long time trying how to get rid of edging effects. During the scaling process, individual pixels have to be averaged out, but pixels near the edge of each map were being averaged with null values which, for some reason, made them negative. As the finishing bit approached the edge of a piece, it would invariably drill straight through the other side.
It took a while to solve this problem, and it would occasionally come back if I wasn’t paying attention, but by March 2021 I had made my first acceptable state – Iowa. From then on, it was more or less just rinse and repeat. I ran into a few unique problems, made a few cosmetic errors, and broke a couple of finishing bits, but I eventually got to the point where I was finishing about one state a week.
Naturally, as the states got further west the topography got more exaggerated, and the states got larger, so I spent even longer at the shop on those days, working on small scrap wood projects.
The final state to be finished was California, not out of any sort of honorific, but because California required all the data to be modified on account of Bad Water Basin and the Salton Sea being below sea level.
Unlike the original map that inspired mine, I wanted my map to be portable (like almost all of my furniture) so I put together the map into two pieces divided East-West to the West of the MN/IA/MO/AR/LA line of states. And truthfully, gluing the states together was probably the second most challenging aspect of the project. The convex-concave problem became extremely exacerbated as more states came together – small discrepancies up in New England resulted in SC and GA not meshing fully together.
But this has easily been one of my most fulfilling projects. I’ve learned a lot about geographic data manipulation, 3D modeling, and the topography of this country. My favorite thing to do has been to learn not only about the physical landforms, but also the human history and habits that they indirectly caused. I absolutely love talking and comparing the Lewis and Clark expedition to the Astor Expedition/Oregon Trail and how incredible it is that the Corps of Discovery made it across such a dense mountain wilderness.
So now what? I estimate I’ve put in close to 500 hours into this project, and my love of topography is still not satiated. Part of me would really love to include Alaska and Canada to the same scale, and a small part of that part would absolutely love to see a 15 foot tall North America from Colombia to the Arctic Circle – I’ll need a bigger wall. But our time here (might) be limited, so I’m mostly focused on finishing some repeat states as gifts, and then maybe some high detail maps of some of my favorite topography in the Colorado Basin.