Author Archive

A New Take on Time Zones

January 11, 2019
Tim Montenyohl

It began with a single line.

A while back I wrote a blog post where I applied our Sovereign Limits database to the International Date Line, redrawing the line to respect states’ sovereign footprints, something the “conventional” IDL doesn’t do. As I wrote that post I realized that I could take the concept significantly further, and apply Sovereign Limits to time zones for a new perspective on the time zone map. It took much longer than I’d ever expected, so I have a fair amount to say about the resulting map. Aside from a little bit of boundary lingo in the next section, a vast majority of this post is about the design process behind the map. If you’re not familiar with our work and research on international boundaries, check out Sovereign Limits, and don’t hesitate to reach out to us on twitter with any comments or questions.

Very few of these lines were made up, which is no small feat  

Recontextualizing time zones with Sovereign Limits

Time zones maps use a lot of “cartographic license” in maritime space, drawing square, blocky lines to divide zones. We don’t need to fabricate these lines, we can use lines in existence that have real-world implications. Applying key concepts of maritime sovereignty marries perfectly with time zones. Using established maritime boundaries, equidistance boundaries, and Exclusive Economic Zone (EEZ) limits to divide time zones in maritime space would make for a more objective time zone map. You could claim I’m being a bit “radical” by using EEZs, as it isn’t conventional. But I’d like to make the argument that it should be conventional. I don’t think I’m adding any overreaching authority to EEZs, rather, most people are ignorant to their whole existence. If the “cartographic license” I’m taking is extending time zones into EEZs, I still find it more equitable than the current “standard practice” of entirely making up lines.

The problem with your average time zone map

Quite frankly, a lot of time zone maps just aren’t good. Conceptually you’d think time zones are a fairly simple thing to portray, but it gets rather complicated pretty quickly. Therefore a lot of time zone maps are complex enough that they aren’t very clear, or they’re flat out wrong. Countries toy/tweak/experiment with their own time zones with surprising frequency, so the sad truth is that even the map presented here in this post won’t remain accurate for very long. (Six months, tops.)

This old CIA map is informative, just not the easiest thing to look at

For my own sanity, I’ve ignored the whole concept of Daylight Savings Time. Not all countries observe DST. Countries that observe multiple time zones can sometimes have regions that practice or abstain from DST, causing irregularities within a single country. Some countries declare that Daylight time is their “standard” time and they roll their clocks BACK an hour to observe “Winter Time.” Other countries take that a step further and permanently observe DST all year round. Morocco is a great example of these shenanigans. They used to observe UTC-0, and participated in DST. In October of 2018 they declared that UTC-0 DST is their permanent, year-round time. I’m going to simplify that and say they observe UTC+1. If I didn’t commit to standardizing countries with peculiar practices, I wouldn’t have made it very far.

Simplify as much as possible

It’s astoundingly easy to get lost down the rabbit hole, which really reinforced the need to simplify every aspect as much as I could. Time zones had to be represented by color fills, using lines to divide the zones wasn’t going to cut it—as something that applies to the whole world, it should cover the whole world. This also places the focus on the subject of the map (where it should be), but means I need to be very careful about the colors used for these fills.


In terms of color palette for the fills, I wanted to use as few colors as possible—I’m trying to stay simple here. This gets tricky. Say I want to cycle between a set of colors, there are areas where zones mash up and jumble together (once again, Morocco…), so even with a sparse set of 4 or 5 colors it can quickly become unclear which zones certain time exclaves are associated with.

Is that bubble (Crozet Island) UTC+5 or is it UTC+1?

When simple doesn’t cut it

Using a limited color palette not only has problems with clarity, but has a whole other issue: tonality. Whatever palette I picked would really dictate the whole vibe of the map. And it would be completely arbitrary. An arbitrary set of colors is unintuitive.

Randomizing the repetition of colors felt …very random (note: these colors aren’t too far off from the final colors)

By simplifying the number of colors I’m causing more problems than I’m solving. Maybe the simple solution is using intuitive colors. When it comes to the passage of time, that means the shade of the sky. But now I need to step through colors for 12 hours of the day (midnight to midday, then reverse the sequence). Initially that didn’t work, as there’s not enough contrast between adjacent zones to make them distinct from each other. To solve that, I used a gradient for each zone. This helps reinforce the temporal quality of a time zone: even though a single zone is observing the same hour, different longitudes within the zone will have the sun at different heights in the sky.

Final gradient vs. flat (and opaque) fills. Flat just wasn’t cutting it, you can see I struggled with how to handle UTC+13 and +14

So now I have a color scheme that looks pleasing and is relatively intuitive. I’ll admit that it falls a bit short in Eastern Russia, but honestly the time zones in that area are messy enough that it’s very possible that I’d never find a solution that works there.

Infographic-style labels

Once I had colored the time zones, I needed to slap some labels on this map. I started to label this the conventional way, but quickly realized I wasn’t doing the map any justice. It was detracting from the zones, even though the labels I was adding were small, light and as unobtrusive as possible. They were so unobtrusive, they would have caught a lot of flak for being illegible. Up against this lose-lose scenario, I decided a different approach. Stack the labels within the zones along the bottom. This gives some quantitative sense of how many states are in each time zone, instead of the obvious geographic sense.

When curiosity drives you to make yet another map

I learned a lot of things as I put this map together. Some countries just don’t seem to properly “get” time zones (Russia, China, Chile), and in Russia’s case they seem to be constantly adjusting time zones every year (hint: try not leap-frogging over zones). I was a little surprised to see how uncommon it is for countries to observe multiple times, a vast majority of countries stick to a single time. The half-hour timezones initially seemed pretty odd, but it makes a fair amount of sense when most countries that use them are using them correctly (Australia excluded). All this leads to an odd observation, which is that a lot of countries observe the “wrong” time. I assume it’s mostly for geopolitical reasons, some much more obvious than others. For example, a bulk of the European Union is UTC+1, even if it means Spain, France, Belgium and the Netherlands observe the wrong time. This compelled me to make a second map—something that puts into perspective how many places observe the wrong time.

I really didn’t expect this much red

Luckily it didn’t involve much, after all the work that went into the time zone map. Everything in red observes a time zone other than the natural time zone it lies within. Even though I knew what I was trying to convey with this map, it wasn’t until I was done that I was really able to gauge how much of the world isn’t observing the correct time. The end result is pretty informative, yet still brings up a lot of questions. I never expected the subject of Time Zones to be as complex an interesting as it is, it only took me making an entire map (or two) to figure it out.

Download this map

Since I made these maps on a whim, we figured we could use them for promotional purposes. If you sign up for the Sovereign Limits newsletter (bottom banner of the page) you get an instant download of both the time zone map and the “wrong time” map (at print-quality resolution, I should add). It’s as simple as that. We plan on sending out a newsletter every other month, so as far as newsletters go, its rather unobtrusive.

PS: I couldn’t naturally fit this random tidbit of information into this lengthy blogpost. But it’s just too good to not share so I’ll tack it on here: I’d like to point out that UTC-12, while only half as wide as your typical time zone, only contains the US minor outlying islands of Baker and Howland. Both those islands are uninhabited. So zero people live in this time zone.

An In-depth Look at 3D Buildings

December 3, 2018
Tim Montenyohl

Starting Off with “Simple” 3D Buildings…

My very first project at International Mapping was for an interactive kiosk for Keweenaw National Historic Park, in Calumet, a mining town in Michigan’s Upper Peninsula. Historic buildings were modeled in 3D, while other buildings had the privilege of remaining as footprints.

Calumet circa 1925

All the modeling was done in SketchUp. For architectural modeling, SketchUp can’t be beat. It’s intuitive enough that it only takes a day or two to get the hang of, at which point you can work lightning fast. (At one time SketchUp was owned by Google, and there used to be a free version of it, but sadly that’s no longer the case.)

I really cut my teeth on Calumet. Learning how to simplify details so buildings are distinguishable on a touch screen of a finite resolution. But interactive kiosks aren’t all that common, the National Park Service is far more likely to have a brochure map, or a map on a large panel somewhere. I’ve been fortunate enough to work on a few 3D bird’s-eye view maps, each with a different level of detail. I thought if I showed off the most detailed as an example, you’d get a pretty good understanding of my process.

More Complex Environments, More Complex Buildings

The Forge (L) and Slitting Mill (R)

Saugus Iron Works National Historic Site is just north of Boston. It isn’t very big, which means for our 3D map everything gets to be seen in detail. Lots and lots of detail. You can find the whole map amongst our portfolio on our main page, but here’s a close up of the Forge (with the Slitting Mill to the right). The actual structure is pretty simple, it’s the waterwheels and sluices that make this a bit overwhelming.

When I model a building I work from reference photographs. I can never have too many photos. Usually we visit the site with the purpose of taking reference photos.

Temporarily out of service

During the scheduled site visit to Saugus, the waterwheel on one side of the Forge was in the midst of being restored. This isn’t as devastating as you’d initially think, it gives me a great opportunity to see the side of the building that’s typically obstructed by the waterwheel. Scouring the internet for additional photos usually makes up for any gaps in our site visit. One thing I make sure to do when I model, is to go deeper than ground level. This helps create the illusion that the structure is part of the environment that it’s in, as opposed to a toy being dropped in place. As for all that crazy detail on either side of the building, it’s just a matter of taking it one step at a time. Eventually everything adds up…

The Forge, modeled in SketchUp

Et voila! It’s the Forge! Except I’ve only modeled the Forge. It still needs textures. This is a pretty critical step that a lot of people have never given any thought to. This is entirely an art in and of itself. (Pay attention to the credits next time you watch an animated film.) It’s called Texture Mapping or UV mapping. (In 3D space, coordinates are along the X, Y, and Z axes, but once you’ve modeled 3D geometry, you map textures to their planar surface. In that space you refer to the coordinates of the surface as the U and V axes, completely separate from the XYZ space. If that sounds confusing, don’t worry—it is.)

A Primer on 3D Texturing

For Calumet I could get away without any textures, there wasn’t any point to adding that level of detail. Not so with Saugus. SketchUp has rather crude UV mapping abilities. That’s okay, I can roughly map my textures in SketchUp. From there I bring the building into Vue, where the final rendering will ultimately occur. After bringing the Forge into Vue, I can finesse the placement of the textures and tweak their render settings. I’m getting a little ahead of myself however, I need textures to map onto the Forge.

Wood texture used for the Forge

Luckily there are resources online when it comes to acquiring textures. The challenge was finding this wood texture (above), since it needed to match the actual wood of the Forge itself. It’s not an exact match, but it’s definitely close enough. Keep in mind that I’m not aiming for photorealism. The end result is still only a representation of the actual location, and part of the point is to exclude some details and to stylize the ones you do include.

The Forge with crudely mapped textures

So now I’ve added textures to the Forge. (At some point in the texturing process the executive decision was made to open its doors.) You might think that this textured version in SketchUp doesn’t look nearly as impressive as the final render. That’s correct. What you see in SketchUp are the color textures mapped onto the 3D geometry. There’s more to texturing than just color. Look at that roof texture. See those shingles? scroll back up and look at the roof. Different roof! The actual Forge roof doesn’t have contemporary asphalt shingles, it uses long wooden planks, similar to siding. When I UV mapped the roof in SketchUp, I used a roof texture I had on hand. Once in Vue, I swapped that texture out for a bump map. This applies depth to a texture, with black representing low portions, and white representing high.

The bump map for the roof. About as simple as bump maps get.

Bump mapping also makes the stucco of the chimneys actually look like stucco. You can map any aspect of a texture: reflectivity, opacity, specularity. It goes on. It turns out the roof doesn’t need a texture map for the color, a bump map and a flat gray color is all it took to recreate the roof. Instead I could have modeled the individual planks that make up the roof, but that would have been tedious—a bump map is the simplest solution.

There you have it. I’ve breezed through some key aspects of modeling and texturing, while entirely ignoring lighting and rendering. There’s a lot of work that goes into these 3D maps, and hopefully you have a slightly better understanding of the process.

Makes sense to finish on the end product. We can’t have the last image be that bump map, can we?

Twitter Facebook Linked In YouTube
Subscribe to our newsletter! SUBSCRIBE