Datawrapper has added population cartograms to its map collections, and in its blog post discusses the advantages and disadvantages of cartograms vs. geographical maps, as well as the advantages and disadvantages of some of the different types of cartograms. Turns out that cartograms are kind of like map projections: each has its pros and cons; each is better suited to some uses than to others. [Caitlin Dempsey]
Maarten Lambrechts discusses how to make a dot density population map, in technical detail. He uses QGIS to process data for his home country of Belgium. The map above is one result (he also zoomed in on Brussels).
NASA’s Socioeconomic Data and Applications Center (SEDAC) has produced a population estimation service “for estimating population totals and related statistics within a user-defined region.” Basically, it provides a population estimate for an area drawn on a map. Available as data via map and GIS clients, it’s also accessible via a web app. I’ve noodled about with it; its population estimates are generally not insane. [Kottke]
Statistics Canada released population and dwelling data from the 2016 Census yesterday. MountainMath’s CensusMapper project already has interactive maps based on that data: population change since 2011 (absolute and percentage), population density, and unoccupied dwellings—with presumably more to come, since the interface allows you to make your own census-derived maps.
Reddit user joostjakob posted the above map to MapPorn with the following caption: “Why population density maps are hard to read: exactly the same amount of people live in the black and in the blue areas.” [Maps on the Web]
Every Person in Scotland Mapped, a dot density map by Heikki Vesanto with a bit of a methodological twist. Rather than randomize the location of population dots within a given postcode, the map “creates a random point within a building shell inside of a postcode area, which is repeated for every person in a postcode. This is in contrast to a simpler process, which does not take into account buildings at all, working simply with postcode areas.” Zoom in and see. [via]
There’s a certain kind of map found all over the Internet that drives me nuts. It’s the map that compares two geographic regions by labelling one with the other: show that this U.S. state has the same GDP as that country by labelling with that country (or better yet, its flag). But the comparisons can get awfully recondite: labelling the Brazilian state of Rio Grande do Sul with Zimbabwe’s flag because they have similar populations is cute but ultimately useless, unless you have some familiarity with both Rio Grande do Sul and Zimbabwe. They’re bad maps because they’re not really informative—they’re just showing off.
But the problem isn’t necessarily the format. For an exception to the above, see TimeOut London’s maps of London. The first map (above) shows London’s population size by illustrating how many other cities’ populations could be crammed inside London’s boundaries; the same is done with greater metropolitan areas, U.S. cities, Scotland and Wales, and other countries. These maps work because a British-based reader will have some sense of what’s being compared to London: they’re not, in other words, esoteric comparisons. [via]
The population of the world from 1 CE/AD through the end of the 21st century (projected) is mapped in this video and interactive map from Population Connection, a group concerned with the carrying capacity of the planet and the environmental impact of overpopulation (they used to be Zero Population Growth back in the day). In each, one dot represents one million people. [via]