Well, I've received a few responses to my plea for direction, and I thank those who responded thus far. Jonathan Lupton is correct, if anyone seems to have a reasonable handle on this issue, please "chime in". We don't want to leave him holding the bag with his "pedestrian knowledge" as the epitome of the state of travel demand! ;-) What I have developed to date is a TAZ system that should reasonable provide us with enough detail to get a better handle on forecasted volumes on the region's "minor arterial" network. In the past the focus had been primarily on the "principal arterials". In the next phase of this TAZ development, I plan to aggregate this system into a nested system that could be used as a "fall-back" in case the more detailed one resulted in large numbers of empty cells when sent to the Census Bureau. This is where I find myself at the moment. How many of the smaller TAZs do I need to combine? Frank, your notes indicated 240 households or about 600 people. I looked at that and found 39 percent of my TAZs fall below that number when I "populate" the cells with year-2000 Census data. Jonathan suggested 1000 population or employment. Again, looking at my TAZs, I find 56 percent of them fall below THAT number. Fortunately (I think), the problem TAZs are found overwhelmingly in the rural or developing areas of our region OR they are TAZs in which I have large employment concentrations. Given that scenario, I am attempting to aggregate these TAZs into something larger. The question is, what should my minimum be? Patricia, I agree with your thoughts on the use of ACS data. However, in this instance, my concern is with the upcoming 2010 census and the information we hope to obtain from a CTPP-like product that has been urged by AASHTO and its Standing Committee on Planning. Having dealt with CTPP (or UTPP) tabulations for several decades, this new realm of ACS and LEHD apparently will be the way to go, but I am still hoping that something will come out of the 2010 census that will reasonably approximate that older product. Thanks so far for the comments. I hope they continue as we head into the decennial census year. ________________________________ From: ctpp-news-bounces(a)chrispy.net [mailto:ctpp-news-bounces@chrispy.net] On Behalf Of Jonathan Lupton Sent: Tuesday, August 04, 2009 4:32 PM To: ctpp-news(a)chrispy.net Subject: RE: [CTPP] TAZ Development Mr. Paddock: I bounced this same question off Nanda Srinivasan years ago (ca. 2006) back when he was involved in the CTPP. He seemed to have as firm a grip on this kind of issue as anybody. I remember being told that the minimum TAZ size should be about 1,000 population or 1,000 employment, to avoid disclosure problems. I do not know what would happen in a TAZ that met one of these thresholds but not the other. I am not offering the information above as anything definitive; as best I remember there seemed to be a lot of deliberate vagueness about appropriate TAZ size. I know our TAZ's in my region are presently quite a bit smaller than these thresholds in most cases. I would dare to predict that LED may end up replacing CTPP in practical terms. At its "On the Map" web site, LED already provides TAZ-level employment by detailed industry for much of the country, including my region. LED doesn't seem to have a disclosure problem, although I think it doesn't yet provide the depth of journey to work data as in pt. 3 of the CTPP. If anybody knows more, please, please chime in. I'd hate to think that the pedestrian knowledge of a peon like me represents the best information among the Great Minds of transportation modeling. Jonathan Lupton AICP Research Planner Metroplan Little Rock AR From: ctpp-news-bounces(a)chrispy.net [mailto:ctpp-news-bounces@chrispy.net] On Behalf Of Paddock, Bob Sent: Tuesday, August 04, 2009 3:33 PM To: 'ctpp-news(a)chrispy.net' Subject: [CTPP] TAZ Development I must admit that, of late, I have gotten lost in the woods while trying to see all the trees. During these past months I have been engaged in developing a new TAZ system for the Minneapolis - St. Paul MPO area that will both serve the needs of travel demand modeling here at the Metropolitan Council AND provide us with the information we want from a 2010 CTPP-like product. This new TAZ system is substantially more refined than those of the past and should result in reasonable travel loadings on our arterial roadway network. However, certain issues have been raising their knobby heads regarding zone size that probably will clash with Census Disclosure Board rules. Does anyone have a reasonable idea as to what minimum population or household levels a TAZ will need to attain in order to circumvent large numbers of "blank" cells? I suspect that there is not an easy, uncomplicated answer but would like to hear one nonetheless. Bob Paddock Transportation Planning Metropolitan Council Bob.paddock(a)metc.state.mn.us 651 / 602-1340

Ed is correct, a concept of "Super TAZs" may very well happen with the 2010 Census, in addition of course to "Base TAZs". A Super TAZ would be an aggregate of Base TAZs (or census tracts or block groups), with a threshold of probably 20,000 population. "Base TAZs" may feature a threshold of 600 persons (residents or workers), but like Ed said, if NCHRP 8-79 proves to be successful, then there may be no reasons for a minimum TAZ threshold per say. The idea behind all this is that one could perform socio-economic analysis and forecasts at the "Super TAZ" level (with proper disaggregation to the "Base TAZ" level), while detailed traffic assignment could take place directly at the "Base TAZ" level. More on TAZ disaggregation methods and techniques can be found at: http://www.ampo.org/assets/604_arctazdisaggregator.pdf The shape and size of zones, the way the zones are delineated, and how well the zones integrate with the highway network all influence the representation of highway and transit access in a model. One way to assess the reasonableness of zone systems (with respect to highway and transit access) is to determine the "optimal" zone size for which the highway and transit access procedures are created. Also important is how the zone system size corresponds to activity patterns. More concentrated activity necessitates smaller zones in those areas, as the increased activity requires the reduced geographic unit to minimize the amount of intra-zonal movements and maintain consistency among trip distribution characteristics throughout a model area. Activity levels vary over time, so both the base and future year activity levels should be compared with their zone sizes. I hope this helps for now, more guidance to follow in the fairly near future. Thanks, Guy Guy Rousseau Traffic Modeling Manager Atlanta Regional Commission -----Original Message----- From: ctpp-news-bounces(a)chrispy.net [mailto:ctpp-news-bounces@chrispy.net] On Behalf Of Ed Christopher Sent: Thursday, August 06, 2009 10:45 AM To: ctpp-news(a)chrispy.net Subject: Re: [CTPP] TAZ Development Watching this discussion on TAZs has been interesting and the size of TAZs has been shown to be more of an art than a rule based science. With that said I want to toss a few more items on the table when folks are considering TAZs. 1. TAZs to be used in the production of a CTPP 5-year ACS product will not have to be developed until the Summer of 2011. The process used will be an equivalence method where folks will be asked to say which Census Blocks (in their entirety) will be placed in which TAZ. 2. There will likely be at least two TAZ sizes. The regular TAZs that we have all come to know and the subject of Bob's question and maybe a larger one somewhere in the neighborhood of 20,000 residents. Just this week the AASHTO CTPP Oversight Board established a subcommittee to work out the TAZ details. Guy Rousseau of the Atlanta MPO is chairing the group. 3. The TAZs will be developed to support a tabulation of the 5-year ACS data using (2006, 07, 08, 09 and 2010) data that is planned to be released in the Summer of 2012. For areas that do not submit TAZs Tracks (and maybe Block Groups) will be the default. 4. For the CTPP 5-year tabulation there is a NCHRP project (NCHRP 8-79) getting underway that, if successful, will take any disclosure issues off the table. The RFP for the project, "Producing Transportation Data Products from the American Community Survey that Comply with Disclosure Rules" was just finalized and put out on the street two days ago. See http://www.trb.org/TRBNet/ProjectDisplay.asp?ProjectID=2708 to view the RFP and details. 5. Assuming NCHRP 8-79 is successful, which there is no reason why it should not be, TAZ sizes will not be affected by the artificial disclosure rules and instead can go back to being determined based on good planning practice. Which doesn't answer Bob's original question but does help put the focus where it should be. 6. Regarding disclosure rules we know from our attempt to get some CTPP tables produced from the 3-year ACS that if the same rules for the 3-year were extended to the 5-year data the losses would be so devastating that there would be no usable 5-year data, which is why we fought for the NCHRP project. 7. In terms of a 3-year CTPP ACS product we hope to have one released in the Spring/Summer of 2010 using 2006, 07 and 08 data. This release will be subject to a variety of data suppression and disclosure rules and will only be for areas of 20,000 residents or more. To see the tables that were designed for this go to http://trbcensus.com/products/ 8. In terms of TAZ sizes for planning the Florida DOT through its modeling group has published a report on the subject. I do not have the link to the report handy but maybe someone else the list can find it and post it. I hope this helps clarify a few things Paddock, Bob wrote: trying that from past roadway not ------------------------------------------------------------------------ -- Ed Christopher 708-283-3534 (V) 708-574-8131 (cell) FHWA Resource Center Planning Team 4749 Lincoln Mall Drive, Suite 600 Matteson, IL 60443 _______________________________________________ ctpp-news mailing list ctpp-news(a)chrispy.net http://www.chrispy.net/mailman/listinfo/ctpp-news