I’m taking next week off so that I can respond to emails from various places in Ontario and Quebec instead of from my desk. The out of office messages really fly at this time of year, so it’s usually a pretty good time to try for a recharge.
Because of that, this post feels appropriate.
Sahil Chinoy of the Washington Post recently looked at anonymous cell phone and vehicle data (from Here Technologies) to see how far you could drive in one hour if you were trying to escape the downtown of various U.S. cities on a Friday afternoon in the summer.
This exercise was done for 3 departure times on July 28, 2017: 4pm, 7pm and 10pm. The mappings all leverage 3 years of historical speed data.
Here is a first set of maps showing a few cities in the northeast and in the mid-atlantic. Every city is shown at the same scale so that they can be easily compared.
And here is a second set of maps showing a few, more car-oriented, cities.
Not surprisingly, older transit-oriented cities like New York don’t do well in this contest. No matter what time you leave, it’s hard to make it past 30 miles. Whereas in the case of Vegas, it doesn’t really matter what time you leave. You should be able to clear 50 miles.
That’s the other interesting thing to note about these maps – the spread between distances at the various times.
I’m sharing these because I’m a sucker for diagrams, but I don’t think they tell the whole story. The modal splits and the population and employment densities are all very different across these cities. New York’s core competency is in moving lots of people in trains, not in cars.
Although, perhaps the ironic thing about these diagrams is that a tighter drive radius might actually say something about how efficiently land is being used.
PLOS One recently published a paper and a set of maps that looks at commuter flows across the United States (over 4 million data points). The objective was to identify all of the country’s “megaregions.”
Here is one of those maps. I think it says a lot.
We often think of cities as having discrete boundaries and population counts, but the reality is that studies and maps such as these provide a much better sense of the overall economic geography of a place.
It’s worth noting that the commuter dataset used for this study is from 2006-2010. So things may look a bit different today. The full report can be found here.
The following diagrams were taken from LSE’s Urban Age website. I’ve sorted them from lowest to highest peak residential population density. In each case I’ve also included the year of the dataset.
It’s amazing how much these simple extrusion diagrams can tell you about the city. It also shows you that high population densities don’t necessarily need to equate to tall buildings. Barcelona, in particular, stands out for me.
Berlin (Peak residential density: 21,700 people/km2, 2009)
Stockholm (Peak residential density: 24,900 people/km2, 2012)
London (Peak residential density: 27,100 people/km2, 2013)
São Paulo (Peak residential density: 29,380 people/km2, 2009)
Mexico City (Peak residential density: 48,300 people/km2, 2009)
Barcelona (Peak residential density: 56,800 people/km2, 2013)
New York (Peak residential density: 59,150 people/km2, 2012)
Shanghai (Peak residential density: 74,370 people/km2, 2011)
Istanbul (Peak residential density: 77,300 people/km2, 2013)
Hong Kong (Peak residential density: 111,100 people/km2, 2013)
Mumbai (Peak residential density: 121,300 people/km2, 2013)
