5 Thirty-eight urban regions
Selecting cities, determining boundaries, mapping regions
What would you do if you wanted to understand urban regions? So many
cities, so variable in size and geography -- the task seems daunting. You might
enjoy traveling to and studying a good batch of them, but these are large complex
objects and the enterprise would take years. Or you might devour books and arti-
cles on the subject, also a protracted process, which would provide a skewed pic-
ture dominated by a limited number of much-studied cities. Or simply talk to the
experts (who wrote those books and articles). Here is the story of how I learned.
Selecting cities worldwide
Togetthe big picture at the outset I pulled out maps, all sorts, and
sketched the shapes and sizes of cities and especially how they are arranged rel-
ative to water bodies and mountains. Since I have lived in parts of North America,
Europe, Latin America and Australia, initially the focus was on other areas, then
gradually becoming worldwide. Quickly I was able to group the sketches into
three big categories characterized by: (1) continent or geographic area; (2) loca-
tion relative to rivers, bays, seacoasts, etc.; and (3) city size, as indicated by area.
Idecided that my mix of urban regions to be studied should include the typical
range ofvariation within each of these groups, and that probably other useful
categories or subcategories exist which should at least be represented. Also I
estimated that 25 urban regions would be the minimum needed (which might
or might not provide clear results), and that 50 regions would be the maximum
possible based on time and resources available.
When I am stuck indoors, the opportunity to study maps, read books and
articles, and talk with experts is a pleasure, so these continued intensively for