Traffic analysis zones (TAZs) are universally used in travel demand modelling to represent the

spatial distribution of trip1

origins and destinations, as well as the population, employment and

other spatial attributes that generate or otherwise influence travel demand. The urban area is

divided into a set of mutually exclusive and collectively exhaustive zones. While travel actually

occurs from one point in the urban region to another, all trip origins and destinations in a travel

demand model are represented at the spatially aggregate level of the movement from an origin

zone to a destination zone. These movements are further aggregated within network assignment

models as originating and ending at single points within the origin and destination zones – the

zone centroids.

The road and transit networks coded into the computer network model reflect this zonal

aggregation of space in that they are designed to carry the flows between zones, not within

zones. Zone centroids (which are the “sources” and “sinks” of trips) are connected to the road

and transit networks via artificial links known as centroid connectors which are a highly

abstracted representation of the actual fine-grained local street network, thus representing

another element in the spatial aggregation associated with the use of traffic zones.2

The choice of a TAZ system is thus a first, very critical step in travel demand modelling, since it

represents the fundamental level of spatial representation, precision and accuracy in the model.

Despite this importance, TAZ definition has received very little attention in either the academic

or professional literature. A review of the recent academic literature, frankly, provides little

practical guidance for operational traffic zone design; nevertheless, several of the citations found

are included in the list of references. Similarly, a scan of the Transportation Research Board

(TRB), Travel Model Improvement Program (TMIP) and Travel Forecasting Resource

(TFResource) websites turns up little in the way of useful guidance. The one excellent exception

found to this statement is Cambridge Systematics (2007), which provides excellent, detailed

guidance concerning TAZ design (even if it is US-oriented, as well as sometimes Floridaspecific).3

This report is recommended as “required reading” for any agency considering

reassessment of its traffic zone system, and it represents the starting point for the discussion

presented below. In particular, this report does not attempt to reproduce/summarize the

Cambridge Systematics report in any way, but rather augment it with what is hoped to be key

points concerning TAZ design from an explicit travel demand modelling perspective.