[CTPP] Development of TAZs for the 2000 CTPP (From Thabet Zakaria)

The following message is from Thabet Zakaria <DVRPC>, not from Margaret Shaw. Please direct all correspondence regarding this mailing to Mr. Zakaria, NOT to Ms. Shaw. Mr. Zakaria's e-mail address is: tzakaria(a)dvrpc.org Thank you. Development of TAZs for the 2000 CTPP This is a brief description of the Traffic Analysis Zones (TAZ) in the Delaware Valley Region, which includes an area of 3,833 square miles and a population of about 5.2 million. I have prepared this summary in the hope that other areas in the country will find it useful for preparing a zonal system for the 2000 CTPP. In 1960, the Penn Jersey Transportation Study developed a set of TAZs for data collection and travel simulation within the cordon area (urbanized area) which did not consider county boundaries. The basic unit was a modified grid square, 0.10 mile on a side in city centers and 0.50-mile elsewhere. A total of 162 districts (TAZs) were aggregated to 51 Superdistricts for evaluation and presentation. In the late 1970s, DVRPC realized that it could not continue to collect massive data for transportation planning, and must depend on secondary sources, including the Bureau of the Census. It was decided to adopt census geography and convert the old TAZs to census tracts and block groups. Census tracts were considered adequate for regional travel simulation, except for the Philadelphia CBD, where block groups were used to define TAZs. The cordoned area was expanded to the region's county boundary and divided into 1,330 TAZs, which were aggregated to 71 county planning areas to simplify analysis and presentation. This system of TAZs was expanded in the 1980s, because the Bureau of the Census in cooperation with the counties divided some census tracts for the 1990 Census. This system proved to be easier to work with and facilitated the preparation of a correspondence table for the 1990 CTPP between TAZs and blocks. It should be noted that the new TAZ system is not free of problems. Some tracts are too large or too small for network coding and modeling, and they must be reviewed each decade because of tract splitting. Also, some tracts are too narrow or curvilinear, and are thus not suitable for travel simulation purposes. It is essential that each tract in the area be examined regarding its layout, size, type and magnitude of development. Several criteria could be used to develop TAZs form census geography, such as population, employment, trip production, trip attraction, zone centroid location, and trip loading (network grain). DVRPC used an average of about 4,000 persons and 3,000 employees per zone in the development of its TAZ system. Obviously, the size of TAZs should be smaller for corridor and sub-area studies and small MPOs.