Hello: We have several TAZ's that have zero population and zero employment because we are modeling areas that are entirely agricultural. One area is about 2 square miles. We also have some TAZ's in central portion of the modeling areas (industrial) that are distinct in terms of the surrounding land use (residential). For us, the issue has always been getting TAZ's are similar in landuse and are defined by access to the transportation network rather than population. Last Census we split everal TAZ's that were developed before we did a revision to our model and it was a real complex problem to split TAZ's in terms of creating matrixes and balancing the network. I would advise against that procedure unless absolutely necessary. Is the Census Bureau going to enforce this rule of 1200 (employment + population) for Taz's as part of the 2010 census? Mike LeBlanc, AICP Planning Manager, Special Projects Division Rosa Parks Transportation Center Lafayette Consolidated Government Department of Traffic and Transportation P.O. Box 4017-C 101 East Cypress Street Lafayette, LA 70502 Ph: 337-291-7179 Fx: 337-291-8458 Planning for the 21st Century -----Original Message----- From: ctpp-news-bounces(a)chrispy.net [mailto:ctpp-news-bounces(a)chrispy.net] On Behalf Of Patricia Tice Sent: Wednesday, August 05, 2009 7:02 AM To: ctpp-news(a)chrispy.net Subject: [CTPP] RE: ctpp-news Digest, Vol 66, Issue 1 The seminar I went to two years ago seemed to indicate minimum TAZ size in the range of 1200 people (again, pop or employment or combination). The issue is that in most areas transit accounts for 0.5-1.5% of the mode split and if you want to keep from having that data suppressed you will need to have that many people in it. I was told that suppression usually will occur when the real value is 3 or less, but true integer value reporting doesn't start until the real value is more than 10 (I think...it's been 2 years). The problem then becomes that area totals over several TAZ's are way off because you are adding "rounded" numbers rather than real numbers. I suspect that the real question will have to do with which data set you are going to use. If you are going to try to use the ACS 1 year aggregate you really need large TAZ's. They say they are not going to suppress as much in the 3 year aggregate and may not suppress anything in the 5 year (at least for category totals). As a modeler, this drives me completely crazy. To get the network to load well, you need to have the TAZ size and connections that are appropriate relative to the geography and roadway network. If you're dealing with a subdivision with 500 houses and one or two driveways the 1200 people limit is fine. If you're dealing with a city block with single family units, you've got a real problem. One way to get around this (thinking out loud) is to take an area with a relatively uniform density and report that area as a whole as a census TAZ. Then when you get the data back, you can split that TAZ proportionately for your model's sake. Most travel demand models are only accurate enough to get you lane calls anyway so making some internal estimates from the Census data will keep you away from suppression issues and not adversely impact your accuracy. It will take a little data massaging on the back end but it may be worth it, especially in areas with dense roadway network connections. Patricia C. Tice, PE, AICP, LEED AP Glatting Jackson Kercher Anglin, Inc. 120 N. Orange Ave Orlando, FL 34787 407-284-4753 _______________________________________________ ctpp-news mailing list ctpp-news(a)chrispy.net http://www.chrispy.net/mailman/listinfo/ctpp-news