This is the fourth iteration of this site. I'm not exactly sure, but I think I'm coming up on roughly six years of having this thing online. When I was writing the first version, hand-coding everything in HTML and CSS, Game of Thrones was still on TV. While I don't think I'd design it the same today, I was pretty pleased with it at the time.

But man, it was hard to update. Every new post required a new HTML page, which is kind of insane to think about today. A couple years ago I switched to Wordpress, the idea being that if the site was easier to update, I might do it more. Partially this plan worked out, and yet I never really felt at home with the design. The themes were too limiting, the options to edit specifics too restricted. Most problems seemed to require a plugin to solve. Each plugin added to the list of ways the entire site could theoretically be compromised. Tutorials constantly tried to shill paid classes, paid themes, and paid plugins. Wordpress never felt like a place I wanted to spend time in, never felt like a place where I could make what I wanted.

Which brings us to version four. It's now a static site, made with Eleventy, which will hopefully draw the right balance between customizability and ease of use. Right now it's pretty bare bones; the design isn't finalized and I have a long list of features to implement and posts to write. But I'm optimistic. It feels right leaving behind the databases and pre-made templates. I get to code all the features I want and can write posts easily in markdown.

More than that though, it feels right having a place to post things away from all the platform-acquisitions and moderation-decisions.

Welcome to the new site!

Lots of the kids in my class have trouble learning phonics, and for good reason! It’s not an easy skill to learn, and can be difficult to practice without someone on hand to correct pronunciation. Meanwhile learning phonics is (as far as I know) the best way to learn how to read which is (as far as I know) one of the most fulfilling things a person can do with their day.

To help my students practice at home, I made a website where they can click on a phoneme and listen to the pronunciation.

Additionally, added a quiz mode that will play a sound, and allow them to click on the corresponding card.

After they run through the quiz, they get a little certificate that they can show their parents (or teachers) and the site will let them practice any sounds they might have missed.

Coding the site was a lot of fun. It was the second programming project that I undertook in conjunction with ChatGPT, which had some useful advice for how to handle some of the audio issues I ran into.

For anyone interested in looking at the code, feel free to check out the github repo. I recorded the audio myself using audacity, and the phonics cards themselves were made by my colleague Andrew.

A bit over a year after completing my world map, I’ve finished a map of the United States. In the same style as the world map, it features important cities, rivers, lakes, and ecoregions. Decorated with over 75 handmade illustrations depicting significant plants, animals, cultural and historical events, this map showcases some of what makes the US a unique and interesting place.

I’d been wanting to make a US map for some time. The longer I’ve lived outside the US, the more I find it intriguing, interesting, and bizarre in the best sense of the word. The research process for this map turned up lots of cool places I’d like to one day visit, and provided a chance to illustrate some of my existing favorites.

While making the map, I also wanted to emphasize the US’s international connections. While there wasn’t space to include them in full, neighboring countries are depicted at the same level of detail, as are the network of undersea cables that provide the US with internet connectivity to the rest of the world.

I also wanted to include as many illustrations as possible of the diverse natural beauty found in the US. While the trails themselves are not included on the map, approximate start and end points for both the Pacific Crest Trail and Appalachian Trail are marked with tents. Useful for explaining the rough outlines of a through-hike to younger relatives.

I think it turned out well, and I’m really proud of it! Enjoy the images posted here, and stay tuned — hopefully I’ll be able to offer prints sometime soon for those who are interested.

As always, if you have any feedback or ideas for things that would be cool to include in the future, let me know.

I hope you like the map!

My neighborhood is a great place to live. A great place, that is, until a concert or baseball game finishes at Kyocera Dome and the streets, restaurants, and grocery stores are flooded by rivers of people.

In an attempt to combat this issue, I wrote a Google Apps Script that will scrape the Kyocera Dome monthly schedule and create google calendar events for each event. At least this way I can try to avoid going to the supermarket when I know it’s going to be a problem.

I’ve definitely done more useful things with programming, but this was deeply, deeply satisfying. If, for some reason, you also have a problem with Kyocera Dome and want to subscribe to the calendar, you can find it here.

If you happen to live near an arena and want to write your own version, you can find the code on my github.

I made this world map for kids (or adults!) who want a playful, friendly, world map.

The map includes hand-placed labels for countries, overseas territories, as well as oceans and seas. Additionally, the map features nearly a hundred hand-drawn illustrations showcasing some of the most amazing plants, animals, and cultural accomplishments from around the world.

The base-map features climate zones, providing a clear and easy-to-understand visual representation of our planet’s diverse climates. Furthermore, the map includes informative panels on earth’s population, atmosphere, and core, making it a great educational tool as well.

I poured a lot of time and effort into creating this map, and I’m thrilled with how it turned out. If you’re interested in getting a copy for yourself, please let me know!

This is my contribution to Daniel Huffman’s projection cards project. My projection, Eckert V, isn’t recommended for use anymore, and was introduced in 1906. I thought it would be interesting to try and portray the world as it was when Eckert V was first released, with much of the world divided up between global empires. That data came from the University of Minnesota’s Historical National Boundaries dataset, which the authors very kindly granted me permission to use.

It was so much fun being part of a larger project, and thrilling to get a copy of my final card in the mail!

This is the first time I made in Blender using non-USGS data. The vintage map as well as the digital elevation model was readily available from the Geospatial Information Authority of Japan, but getting the data into a usable format was a bit of an adventure.

I haven’t done a full-size render of this one yet, so it might be a tiny bit fuzzy, but the key elements are all there!

Good apartments can be hard to come by in Osaka, and one of my friends had his eye on a specific unit that he knew was about to be put on the market. The only catch was that applications are handled on a first-come first-served basis, and his job wasn’t compatible with constantly refreshing the housing company’s homepage.

I wrote this web scraper in python to check available units against a list of desirable units, and send an email when a new listing appeared. Time was of the essence so the code is probably less elegant than it could be, but it worked!

Anyone looking to create something similar can have a look at the code here.

View full-resolution map here

While I would deeply, deeply love to report otherwise, a 2011 Macbook air is not the computer to use if you want to layer modern hillshades with vintage maps in Blender. After many fruitless hours trying, this is my best shot using QGIS alone.

I’m not entirely satisfied with the results. They’re a far cry from the soft light and gentle shadows I had been hoping for. Still, I think the results are decent given that everything was done in QGIS. You can read the full write-up for this project here.

John Nelson makes some amazing tutorials and how-to videos for ArcGIS. While I love ArcGIS, I wanted to see how close I could get using the free and open-source QGIS. For anyone looking to learn more about QGIS, I’d highly recommend this approach. The steps are sometimes fairly different, so it’s a great way to get to know the software.

I see this process as honing my aesthetic sense for what works on the page, as well as expanding my technical knowledge. It’s also good in that it gives you a project, and a rough set of ideas for how to proceed, but you can’t fall back on following a written list of instructions. You really have to do some research and figure out how to get QGIS to do what you want.

Occasionally I’ve had to start in QGIS before moving to Illustrator, and some of the tutorials have been flat out impossible given the differences in the programs, but it’s a great way to practice making visually-appealing maps, learn more about cartography, and push the limits of what QGIS can do.

I haven’t made step-by-step instructions for these maps the way John Nelson has, but if you have any questions, feel free to reach out and I’ll see what I can do to help.

Here are a few of my favorites. Based on this map of the Rift Valley.

Based on this map of Indonesia.

Based on this hurricane map tutorial.

Based on this map, which is in turn based off of Tolkein’s map of Middle Earth.

To continue practicing and improving my R and ggplot programming skills, I occasionally contribute to the Tidy Tuesday programming challenge. If you’d like to see the graphs, or check out the accompanying code, feel free to [check it out on GitHub]https://github.com/benleamon/tidy-tuesday).

Several years ago I (and a lot of the internet) was very taken with Daniel Huffman’s techniques for creating hillshade in Blender, and the subsequent explosion of modern hillsahde imagery layered with vintage maps. After getting a new computer and basking in the increased processing power, I was able to make some pretty significant improvements on my original attempt in QGIS.

I stared with a USGS topo map, o which I added the hill-shade.

Which I then colored with the original map. I’m really pleased with the level of detail, even close up. Here is Point Reyes station and the surrounding hills.

Compared to what I was able to do on my old computer in QGIS, the difference is night and day. The shadows are softer, the light gentler, and the details better. Last May I was impressed with how much QGIS could do. Side by side, however, there is no comparison.

That said, there is still a lot to do. I still want to improve the lighting and contrast, and figure out how to make maps with irregular outlines, like the original Point Reyes one on the right. I have a feeling that two or three iterations later, the process I use for these maps will be at least somewhat different, but for the moment, I’m quite happy with the results.

While working on a different project, I came across these maps of forested land in the US, Australia, and Europe, from [Earth Art Australia]https://www.etsy.com/shop/EarthArtAustralia). I was at a bit of a roadblock in my other project, and making a map of forested land in Japan seemed like an interesting — and most importantly easy— diversion. The project, however, quickly spiraled into interminable raster-processing on my aging macbook air.

Highlights:

Compared to the maps of The US and Australia, the forests mapped closely to hills and mountains. North-Eastern Hokkaido, around Nakashibetsu, was interesting. Grids of agricultural fields are bordered by stands of trees.

Unsurprisingly dense urban areas like Tokyo are largely free of dense vegetation.

Osaka, Kyoto, and Nagoya are also pretty sparse, and in the image below you can just see the ghost of an outline around Lake Biwa. Most of the countryside is pretty green!

It should be noted that this data shows vegetation over five meters, and there are bound to be some errors. Overall though, it felt good being able to find something online, and reproduce it without being given any guidance. Here’s how I made the map:

Process

I used QGIS, as well as the data linked to in the original EarthArtAustralia post.

The plan for the project went as follows:

1. Download the digital elevation model, or DEM, a data layer that contains elevation data for every point in our area of interest.

2. Turn the DEM into a hillshade layer, and reproject it using a projection suitable for Japan.

3. Download the forest data, and reproject it using the same projection as the hillshade.

4. Write a script to delete any part of the hillshade layer that doesn’t contain any forest.

5. Use this layer as our base map, then use the forest data from step two to color it green.

Ultimately, either my computer or QGIS (probably my computer) simply wasn’t up to running a lot of the geoprocessing operations the way I wanted them to. Batch processing failed to even start on a lot of the raster layers, and performing multiple operations (for example reprojecting and saving in one step instead of reprojecting and then saving the resulting output layer separately) triggered hours-long wait times. Ultimately I completed each step on every raster tile individually, then joined the finished tiles together.

If anyone wants to make a map of their own and is running into problems, feel free to contact me, I’m happy to help!

One of the podcasts I listen to, The Modern Mann, features honorary “Mannbassadors”, patrons of the show who represent their various locales. I made an interactive 3D map of all the Mannbassadors over the years, written in D3.JS.

As I was working on the project, the host, Ollie Mann, mentioned that he was looking for a map. My map was one of several submissions, and was ultimately chosen to be featured on his site.

You can see the full map on the podcast's website or here

Since 2017 I’ve been keeping track (roughly) of how much time I spend working on personal projects. I got the idea from Cal Newport’s book, Deep Work, and it’s been really useful. Keeping track of how long I work helps keep me motivated as well as honest about how hard I’m working. It’s also good to know how my projects actually take.

I made this chart using D3.JS.

You can see the time tracker here.

In 2018, I wanted to create a long-form infographic similar to those found online and in magazines. At the same time, I wanted to refresh my knowledge of HTML, CSS, and Javascript, which I hadn’t used since college, and learn how to create interactive web maps and graphs using R.

I think I bit off more than I could chew, and the project ended up taking much longer than I anticipated. Still, I’m proud that I could teach myself these skills during my breaks at work.

Throughout the project, I used a variety of tools, including HTML, CSS, Javascript, Leaflet.js, ggplot, and Illustrator.

This project is no longer live -- it requires PHP which isn't supported by my current static-site setup. However if you're interested in having a look at it, you can check out the github repository for the projetand run it on a local development server.

This is a simple map of Japan that I made in 2017 to get into the swing of things with QGIS and practice styling a simple QGIS map in Illustrator. If you’d like a copy, feel free to download and print it!

Download here!

Orange blossom water is a versatile ingredient used in Middle Eastern cuisine. It has a unique, fragrant flavor which is truly delightful. I made this infographic to practice using Illustrator and as a project for a Japanese class I was taking. As a brief caveat, my 2016 Japanese was not what it is today, and that lack of experience probably shows.

Link to PDF