World timezones: the shapefile

The insane hairball that is the world's timezones has been rendered as a free/open shapefile for all your cartographic/temporal needs. The project, maintained by Eric Muller, has been going since 2008, and is a work of data-heroism, given how totally screwed up and complex the world's timezone divisions are.
The tz_world shapefile captures the boundaries of the TZ timezones across the world, as of TZ 2011b. The geometries are all POLYGONs, and a TZ timezone will sometimes have multiple polygons. There are about 28,000 rows.

The tz_world_mp shapefile captures the same boundaries. The geometries are either POLYGONs or MULTIPOLYGONs, and there is a single geometry for each TZ timezone. There are 394 rows.

There is a companion map for the TZ timezones used in Antarctica stations.

The geometries are primarily derived from the fip10s data (itself derived from the VMAP0 data), augmented with data presented in the pages for the maps of the United States, Canada, Russia and China.

tz_world, an map (via O'Reilly Radar)



    1. I’m in India right now, and India also has only one time zone. I guess it’s easier as a developing country to avoid the complications of timezones.

  1. Yay for freely available complex GIS data! But based on the image being shown, there are some strange patterns. 1) Newfoundland and Labrador are the same province with the same timezone, but have different colours. The lower peninsula and most of the upper peninsula of Michigan are eastern time zone, but marked differently than the rest of the eastern time zone. Why are US states joined but not Canadian provinces?

  2. Cool project, I’m sure. But serious question–what’s screwed up about time zones? Don’t they actually work quite well for what they are?

    Sure, the borders are irregular, but that’s because populations aren’t evenly distributed and it’s worth it to most countries not to split themselves into two or more zones, if they can help it. There’s nothing particularly holy about solar noon being 12:00 anymore. And that map greatly (visually) exaggerates the reality on the ground, as a glance at a “normal” time zone map will tell you. You could cross a lot of different colors on the map before you had to change your watch.

    I can think of a few changes that have been made–some shenanigans with islands wanting to be first into the millennium, and tinkering with the specific days that daylight savings took effect. But so what? How much simpler does it need to be?

    1. Yeah, I’m sure this is useful and accurate (and seems to include historical information) but the image doesn’t just show today’s “world timezones.” For example, about two dozen countries in Western and Central Europe are in the same time zone, yet they show up as lots of individual colorful patches.

  3. The TZ file reflects the actual practices, not necessarily the official ones. Newfoundland sets time by their official zone, but Labrador doesn’t.

    1. What is being captured by multiple timezones in China (and some places in Canada that I’m fairly certain have the same timezone) is historical differences. Beijing and Harbin have the same time now, but before 1949 they were different:

      I am in awe of the anal-retentiveness that led to this map (and the underlying zoneinfo database).

      I am also somewhat surprised that the map is this *simple* given that it captures not only current but historical differences.

  4. Some of the areas that are seperate that you’d think would be connected are divided not because of a different time zone but because of different rules regarding daylight saving time (or the local equivalent).

    On a semi-related note, Russia just did away with changing between summer time and winter time due to all of the trouble the change causes twice a year. They’re now permanently on summer time. Kind of wish we’d do the same here in the States but then there wouldn’t be any dates for pols to change every couple of years to make it look like they’re doing something to combat rising energy prices.

  5. Alas, that map is riddled with errors. Looks like someone is conflating regions with timezones. Not only is China wrong, but so is Russia. It’s not even right if you used Russia’s timezone map from 2009 (See: ). Also, whoever made the map, can’t color it correctly. US Eastern and Central have the same color.

  6. This map seems to be inaccurate in at least one manner. Looking at the US, it still shows the outline of Indiana below Lake Michigan because it used to be a standout of daylight savings time and belonged to neither Central or Eastern timezones. Notice the past tense? 3-4 Years ago It was changed such that most of Indiana is now Eastern. The map or page linked to does not reflect this.

  7. There are about 4 counties in the upper peninsula of Michigan that border Wisconsin that are in the central time zone. All the rest of the UP and LP are in the eastern time zone. It makes no sense for *all* of Michigan to be shown as some oddball sitting on the northern tier.

    1. If you look at the US map closely, you’ll see that those 10 Michigan counties are in a separate zone (“America/Menominee”) than the rest of Michigan (“America/Detroit”). The reason Michigan is shown as “some oddball” is that the zones reflect historical changes, not just the current time zones. According to the TZ database authors, “Each location in the database represents a national region where all clocks keeping local time have agreed since 1970.” Michigan is set apart to reflect its transition from Central to Eastern; its clocks have been in continuous agreement with neither the clocks of Chicago nor the clocks of New York since 1970. (This also explains why Indiana has so many zones: various parts of Indiana have switched between Central and Eastern at various times.)

  8. Most of Indiana is now permanently part of the E.S.T., which means we get to change clocks now too. Most of us enjoyed not doing that.

  9. as noted above, there are many errors.
    In Canada, Newfoundland and Labrador have the same time.
    Manitoba and Saskatchewan have the same time.
    And why doesn’t the map show areas that have a time difference of just one-half hour?

  10. Maybe this has already been said, but Newfoundland and Labrador do actually have different timezones.

    1. Never mind, I looked it up.

      For anybody else wondering: ‘multipolygon’ is a term that refers to multiple disjoint polygons, considered as a single entity.

Comments are closed.