I appreciate what they're doing. It's pretty cool looking. But This image, too: Perhaps more importantly, why? You don't think the fact that Reno is across state lines has something to do with all the San Francisco traffic there? Also: Right. Like "cleaning up" the connections between Los Angeles and Las Vegas. I've literally driven there for meals. We're surrounded by Vegas billboards down here. Vegas exists for Los Angeles. Yet it's its own region because it's bigger than Reno, I guess. Right - a different one shows Phoenix as stretching north of the Grand Canyon. Data visualization is a useful skill but I'm not convinced that their approach reveals any insight. There's clearly a lot of shaping going on. From the PLOS paper: They took ZCTA data and fudged it until it looked cool. 'K. It looks cool. But if part of your method is "making it look cool" are we really learning anything from the process?In this image it’s easy to see that the main centers of work are in cities, including San Francisco, Oakland, San Jose, and Sacramento, and that they are highly connected.
But where should planners draw the edges of a megaregion encompassing this activity?
The results of the algorithmic analysis took some cleanup and iterations—such as eliminating superlong commutes between places like New York City and Los Angeles and excluding nodes with only very weak connections—to produce a coherent map of plausible megaregions.
A map shows all the commutes of 50 miles or less in the greater San Francisco Bay Area.
A heuristic approach is, put simply, one which involves exploration through trial and error to produce results which are useful, but not necessarily optimal. Simon [25] refers to this as the concept of “heuristic search” in which sub-optimal but effective “satisficing” solutions are arrived at. We take this visual approach with the ACS commuting and workplace data by working iteratively towards representations which identify core labor market areas based on networks of flows. In addition to providing a visual ‘sense check’ on the underlying data, this also provides a useful reference point for the algorithmic approach described later.
Yeah, the thing is that not only has it been done before: ...it's been done with more careful thought and research into the history and culture of the continental US, too: You can read about the reasons for dividing it up here. His book has been on my reading list for a year or two now. Don't know if it's any good but someone who understands the influence the Dutch have had on NYC must know at least some history. PS: the Las Vegas / LA combination was filtered out before their heuristics even came into play: But hey, pretty graphs!
When filtered by distance (d < 160km)...
The Kentucky and Ohio ones hit me. Lexington and Louisville are to vastly different cities with incompatible cultures. Louisville is an old industrial Catholic city, Lexington is the horse capital of the world (so they say) very agrarian and very Protestant. Those living in either city may travel to Frankfort, but they do not go to each other except for events, definitely not for work. Lexington is closer to Cincinnati both in style, culture and geography. Cincinnati itself is the major hub in the area pulling in people from West Virginia, northern Kentucky and most of Southern Ohio. Texas looks about right to me. My gut instinct is to also call Florida and the Northeast about on point. So, they made a pretty map. Wonder what the point of it is? Oh, hell, I did not catch this one: They ran out of time approaching the deadline? Ran out of money for the computer time? Why not run the data and see how many "mega" regions pop up and go from there? Found the link to the high-res versions of the maps hereThey took ZCTA data and fudged it until it looked cool. 'K. It looks cool. But if part of your method is "making it look cool" are we really learning anything from the process?
One of the decisions the researchers made was to limit the algorithm to 50 megaregions, which can be seen in the map above, where every node is colored according to the region it belongs to. This made the map more plausible visually. While 50 may sound like an arbitrary number, it makes sense mathematically because a very high percentage of commutes lie entirely within a megaregion relative to paths that cross boundaries between regions.
One of the things veen and I did when we were citing the birth center is deal with isochrones. See, when you're talking about people and what kind of bullshit they'll put up with, distance matters fuckall. For example, it took me longer to get from Playa Del Rey to Pasadena (30 mi) when living in LA than it took me to get from Snohomish County to Mount Rainier (76 mi) because there's only 6 million people in the way, as opposed to 18 million. If you get an ESRI demographic & income profile they'll break things down by "10 minute drive time radius." I'd already run the numbers on bunch of client data for two birth centers and determined that 80% of all clientele comes from 15 minute drive-time (less traffic). But isochrones are iterative and time-consuming so no wonder they didn't run this analysis on it... because really, it would have told them what we already know. I mean... one of them is a grad student. They probably started working on it and determined that the sun would be a cinder before they ran isochrones on an entire census dataset. So they made some pretty pictures instead.