Civil Engineering Reference
In-Depth Information
the year. This would be the best approach were it not the most expensive,
and if people would indeed have the patience to provide so many answers.
In practice, this kind of direct data has been sparsely if at all available.
As our second option, we can consult data that researchers have
collected for various parts of the country on average sizes of households,
numbers of cars owned, and numbers of trips per day for job, shopping,
and social visits. What we get is propensity to travel by various household
characteristics. We learn that two-person households owning one car typi-
cally take 2.4 trips per day. We then find how many two-person single-car
households there are in our zone of interest, and then multiply that num-
ber by 2.4; and then repeat the procedure for each other combination of
household size and car ownership.
As our third option, we can consult trip-generation rates by land use,
collected by underpaid students standing on streets and observing numbers
of trips in and out of buildings. Such rates have been assembled for decades
and may be obtained in printed and electronic formats, often called
trip
generation manuals.
For a shopping mall with a given number of retail estab-
lishments or given square footage, the rates tell us the average numbers of
trips in and out. To be sure, it is wise to verify national data through local
trip-generation studies.
Now comes the more conceptually complicated part of the trip-gen-
eration step: reconciling trips produced with trips attracted (excluding trips
that start and end inside a single zone). The number of trips produced by
households in all zones during, say, a one-hour slot in the morning must
equal the number of trips attracted by jobs, schools, shopping centers, and
other households (some people may just be visiting friends) in all zones dur-
ing that hour. Among the difficulties are trips that cut across the one-hour
slot and chains of trips by one traveler within the hour. The need to rec-
oncile the production and attraction numbers adds accuracy to this method.
For Square City, we have kept our model exceedingly simple: we
assume that zones have identical proportions of residential, retail, and other
land uses and that trip production and attraction are completely proportional
to population.
Trip Distribution
Imagine we are going through our four-step model for a hypothetical metro
region. In the morning peak rush hour, of all the trips Zone A produces
(they leave Zone A for other zones), 50 percent go to work, 30 percent
to school, 10 percent to shopping or professional services, and the rest go
elsewhere for other reasons. If we had perfect data on the zone, we would
also know the zones that are attracting its residents.
