Skip to content U.S. Department of Transportation/Federal Highway AdministrationU.S. Department of Transportation/Federal Highway Administration

Office of Planning, Environment, & Realty (HEP)
Planning · Environment · Real Estate

HEP Events Guidance Publications Glossary Awards Contacts

CTPP Data Products

Transportation Analysis Zones (TAZ) FAQs

To view PDF files, you need the Acrobat® Reader®.

Delineation criteria and guidelines

Q. Is there a population threshold for TAZs?

A. No, however, it is strongly suggested that TAZs should be delineated with a resident or worker population of 1,200 or greater.

This minimum population suggestion is the result of the relatively small sample size in the ACS which even after 5 years of data accumulation remains small with some large Margins of Error (MOE). For example, in an average census tract with a population of 4,000 residents the Census Bureau estimates that there are 120 unweighted households, and 150 unweighted workers. Assuming that there are approximately 5 TAZs in the census tract, there would only be about 30 unweighted workers in a TAZ from which to tabulate the Census Transportation Planning Product (CTPP). Thirty unweighted records are not going to give very good results in the CTPP.

The TAZ delineation software, TAZ MAF/TIGER Partnership Software (MTPS) has a verification check that identifies TAZs if the population (resident populations based on 2010 Census), and/or worker estimate (based on CTPP 2000 and distributed to 2010 blocks by land area) is below 600.

Q. Can a Transportation Analysis District (TAD) include multiple counties?

A. Yes, a TAD can include multiple counties or parts of multiple counties. This would be more likely to occur in more rural areas where the population in a county falls below 20,000, and by combining areas across county lines the TAD population threshold of 20,000 can be reached.

Q. What, if any, are the nesting requirements for TAZs?

A. TAZs must nest within counties (and also cover the entire area of a county). Additionally, TAZs must nest within TADs.

Q. What, if any, are the nesting requirements for TADs?

A. TADs are not required to nest within a county. TADs must only nest within the delineation coverage assigned to the MPO/DOT (which are composed of complete counties). Additionally, TADs must be composed of whole TAZs.

Q. If TAZs are not delineated for a county, how can TADs be defined?

A. If no TAZs are present, TADs can be built from 2010 block groups or 2010 census tracts.

Q. Do TAZs have to nest within Counties (and defacto, States) besides nesting within TADs?

A. TAZs must nest within counties (and by default, States). An easier way to think about it (instead of TAZs nesting within TADs) is that TAZs are aggregated to create TADs (i.e., TAZs are the geographic building blocks for TADs). If you are delineating TADs, then delineate TAZs first, and then build TADs by aggregating TAZs.

Q. Will the TAZ MTPS contain criteria and guidelines verification checks for TAZs and TADs?

A. Yes. The TAZ MTPS will check for code uniqueness (TAZ codes are unique with the county, TAD codes are unique within the delineation coverage), TAZ/TAD nesting (TADs are composed of complete TAZs), completeness (TAZs cover the entire county, TADs cover the entire county/counties within the delineation coverage), minimum populations (600 residents or workers for TAZs -; suggested, and 20,000 residents or workers for TADs - required), and possible TAZ/TAD delineation errors/anomalies (contiguity and shape index).

Q. Can multiple MPOs with adjoining territory define different TAZs for the same county?

A. No. In the case of a county for which two MPOs want to define TAZs, the agencies need to agree on the TAZ delineation that meets both of their needs and submit one file for the county. We expect good cooperation between the MPOs to define TAZs that are agreeable to both agencies.

Q. What is the purpose of a TAD?

A. Right now, ACS 3-year tabulations are limited to geographic areas with populations of 20,000 or more. This results in a "swiss cheese" of data availability (see Ed Christopher's slides) when examining counties and places. If there were a nationwide coverage of geographic units that met this population threshold, it could potentially be used as an ACS 3-year tabulation geography, either for ACS standard products, or for future CTPP.

Q. Is the 20,000 population threshold for TADs a requirement or a suggestion?

A. Because we want to use it for 3-year ACS tabulation, it should be treated as a requirement, not a guideline.

Q. Are TADs optional?

A. Yes, delineating TAZs and TADs are both optional. However, we are encouraging TAD delineation, so that there will be a nationwide coverage of TADs, similar to nationwide coverage of PUMAs, that could be used for ACS 3-year tabulations.

Q. Should I build TADs from completed TAZs?

A. Yes. If you don't delineate TAZs, then you can make TADs from 2010 Census Tracts or 2010 Census Block Groups, or entire counties (in rural areas).

Q. If no Census TADs are defined for an area, will the Census define them?

A. If no TADs are delineated by the participant, AASHTO is in discussion (as of March 9) with a consultant to delineate TADs for those areas for which no TADS are defined. The Census Bureau will NOT perform this task.

Q. Can I delineate a large TAZ (US Forest Service land) with no population?

A. Yes.

Q. Can I create one dummy TAZ for all the water bodies, e.g. "H20", in the study area?

A. Yes, but if you have multiple water bodies, they will be discontinuous, you should report that this is what you are doing when you submit the files back to the CB.

Q. Should the TADs use resident population and not worker population?

A. Yes, the TADs should use 20,000 resident population as a requirement, as the goal is to have a geographic unit that will meet the CB's Disclosure Review Board requirement for ACS 3-year tabulation.

Q. Are there any plans for the CB to use these TADs for "standard" ACS 3-year tabulations?

A. No, at the present, the CB does not have any plans to use these in their standard ACS products, but a future CTPP 3-year tabulation could use them and reduce the "swiss cheese" problem that is readily apparent in the CTPP 2006-2008 product.

Q. Should the agency defining TAZs define the TADs?

A. Yes, if you are defining TAZs, then you should define the TADs using the TAZs as a building block.

Q. Will TADs replace place data in CTPP tabulation?

A. No, "Place" is a different tabulation summary level than TAD.

Delineation software (TAZ MTPS) and partnership shapefile data download

MTPS TAZ Software - The TAZ module is part of the MAF/TIGER Partnership Software suite being developed by Caliper Corporation for the Census Bureau, specifically for the CTPP. This is a GIS-based program that will create block equivalency files to be submitted to the Census Bureau, and subsequently added into TIGER for the CTPP tabulation to be completed.

Q. When will we get the final CTPP TAZ delineation software?

A. The Census Bureau plans to have the software finalized by March 16, and available from their website, to which you will need a password.

Q. Why can't I see my data when I first open the software?

A. When the TAZ MTPS first launches, it will not be zoomed into your specific county coverage. You must use the "zoom to area" tool in the TAZ toolbox to see your county coverage.

TAZ screen shot as described in the text.

Q. Do I need to have ARCGIS on my PC to do this?

A. No, the software is a standalone software module built by Caliper Corp for the Census Bureau for the CTPP program.

Q. How much space is the software going to take?

A. The MTPS software takes roughly 150 MB. The data can be roughly 15 MB to over 850 MB per county (depending on the county size and feature density).

Q. Can the software be installed on multiple PCs within one agency?

A. Yes, if you would like to divide your work among multiple staff members you can install the MTPS software on multiple machines. You would then administratively divide the work, and ask the staff members to work the assigned area. Once the staff members create output files of their TAZ/TAD delineation work you can pull the multiple files together onto one machine (using the multiple BEF import functionality) and run them through the verification checks before returning the complete submission to the CB. The participant guidelines will have additional information on either administratively dividing the work or creating multiple setup files to divide the work amongst multiple staff members.

Q. If I want to use the CTPP2000 TAZ as a starting point, do I need to download a file?

A. If TAZs were defined for CTPP2000 for your area, the software automatically includes this choice as a starting option. (Note: TAZs were not universally defined in 2000).

Q. Can I use a shapefile from my travel demand model to import TAZs into the MTPS TAZ software?

A. Yes, you can use a shapefile as a reference layer in the MTPS TAZ software, or you can have your GIS system generate a Block Equivalency File (assuming your GIS has imported the TIGER files with the 2010 Census block geography).

Q. If I start with 2010 block groups, can I do multiple counties at once or does it only work one county at a time?

A. It can be used for multiple counties.

Q. If I use 2000 TAZs as the starting point, will there be any problem with a 2000 boundary no longer being valid?

A. If you use the 2000 TAZs, there may be some areas where they do not match up with the 2010 blocks. You will have to correct those areas using the software (they will be easily identifiable using the unassigned areas check in the verification toolbox)

Q. The MTPS TAZ software includes an estimated count of workers. Is this the same as employment?

A. No, the count of workers is an estimate based on Census 2000 long form results distributed proportionately by land area to 2010 census blocks. This count will be less than total employment (or jobs), because the Census 2000 long form allowed only 1 work location, and a worker might have 2 jobs.

Q. Does the TAZs name have to be unique only within state or nationwide?

A. The full TAZ code includes State and County FIPS, so each TAZ code is unique within each county

Q. After we complete TAZs and return to the CB, can we export the TAZ results to a shapefile?

A. YES. When you report changes after completing your TAZ/TAD delineation county-based shapefiles will be output into the RETURN ZIP file. If you delineated TADs across county boundaries, the shapefiles will be clipped at the county boundary. The pop and worker counts for these TADs will contain the counts for the entire TAD, not the portion of the TAD in one county or another.

Q. How do I create a Block Equivalency File?

A. Please see the Census Bureaus' BEF documentation file: http://www.fhwa.dot.gov/planning/census_issues/ctpp/data_products/tazmtps.cfm that they shared during the webinar. The TAZ download page is not yet active. Once active, the site will be: http://www.census.gov/geo/www/cob/tz2000.html. The BEF documentation will be posted in the appendix of the participant guidelines available through the 'Download Traffic Analysis Zone (TAZ) Documentation' link.

Q. Can the software be installed across a network?

A. No, this is a desktop application, but it can be installed on multiple machines.

Q. If my MPO crossed multiple states, will there be one set up file?

A. Yes, there will be a single set-up file. The user guide will include directions for creating multiple setup files if you want to split up the work.

Q. What is my username and password to download the TAZ MTPS software?

A. Call the GEO TAZ team at 301.763.1099 for this information.

Q. What is the password to launch the TAZ MTPS software (after installation)?

A. The password to launch the TAZ MTPS software (after installation) is "TAZ" (case sensitive).

Q. Can I download partnership shapefile data for counties outside my delineation coverage?

A. Yes, however you should only download counties within your delineation coverage since the Census Bureau will only accept TAZ/TAD delineations for counties within your assigned delineation coverage.

Q. What do I need to start my TAZ/TAD delineation project? How will I obtain these materials?

A. To start your TAZ delineation project you'll need the TAZ MTPS, the county-based partnership shapefiles for the counties within your delineation coverage, participant guidelines, and a setup file.

The TAZ MTPS, county-based partnership shapefile data, and participant guidelines that will be used in the TAZ MTPS will be available for download from the Census Bureau's website. The custom setup file (necessary for each project) will be emailed to program participants. More specifically, when the county-based partnership shapefiles for your delineation coverage are available for download, the Census Bureau will email the TAZ program participant with a customized setup file, asking that the participant call the GEO TAZ team for the username and password to download the TAZ MTPS from the Census Bureau's website. The emailed setup file contains a list of the county(ies) within your delineation coverage. The participant will download the TAZ MTPS, the partnership shapefiles that correspond to the county(ies) within the setup file, and the participant guidelines from the Census Bureau's website. At this time, the participant will follow the instructions for installation and starting the TAZ MTPS within the participant guidelines.

Delineation program (general)

Q. How firm is the 90 day window and why?

A: Firm, otherwise, it will likely delay the delivery of the CTPP 5-year tabulation (2006-2010 ACS records), which we now hope will be at the end of 2012 (or early 2013). Otherwise, it could set back the CTPP production by as much as 12 months.

Q. Why have people been talking about using two different TAZ structures: a TAZ for the regional model and a TAZ for CTPP?

A. In the past, it made a lot of sense for the TAZs used for the regional models to be the same as for aggregating census data. However, with the advent of GIS and advanced statistical tools, it is possible to fit zone sizes to the quality of the underlying data and have zones sensitive to travel forecasting and modeling needs. For the ACS data, bigger is better, but for modeling, smaller is preferred as agencies move toward microsimulation. One recommendation is for the regional model TAZs to nest within the Census TAZs, especially if regional model TAZs are very small.

Q. Can TAZ and TADs be submitted to the Census Bureau incrementally, that is one county at a time, or do I have submit "all" the counties for which I am responsible at one time?

A. The Census Bureau requests that all the counties for which you are responsible be submitted at the same time.

Q. Do you have any advice on how to define my regional model TAZs?

A. Please see the links to recommendations from the Florida State Model Users' Group, and Guy Rousseau from Atlanta Regional Commission that may help you.

http://www.fsutmsonline.net/images/uploads/reports/FR1_FDOT_Taz_White_Paper_Final.pdf

http://ctpp.transportation.org/Documents/ARC%20TAZ%20DELINEATION%20STRATEGIC%20PLAN.pdf

Q. We will be creating a whole new TAZ structure county-wide for our travel demand model. Will the document from FL DOT give us the best guidance?

A. In addition to the FL DOT document, you may also want to look at materials from the Atlanta Regional Commission: http://ctpp.transportation.org/Documents/ARC%20TAZ%20DELINEATION%20STRATEGIC%20PLAN.pdf

Q. Does the TAZs name have to be unique only within state or nationwide?

A. The full TAZ code includes State and County FIPS, so each TAZ code is unique within each county

Q. What is the starting point for defining TAZs?

A. The software has several different options for the starting point. These include: starting with CTPP 2000 TAZs, 2010 Census tracts, 2010 Census Block Groups, or bringing in your own Block Equivalency Files (BEF).

Q. Where do the residence counts and worker counts in the file come from?

A. The residence counts are the 2010 Census block total populations and will match the PL-94-171 block totals. The worker counts use the results of the Census 2000 and CTPP2000 and are an estimate of workers in 2000 at the place of work.

Q. How can I use my own local resources with more current block level worker counts (or estimates)?

A. The simplest method is to bring in your shapefile (or other file) as a reference. Keep in mind that the worker counts displayed in the TAZ toolbox would still reflect the 2000 worker counts interpolated to 2010 data.

Another method would be to bring in the more current block level worker counts into the tabblock2010 shapefiles. The more current worker counts must be loaded into the WORKERS00 field in the tabblock2010 shapefiles for the MTPS to recognize the worker counts in the TAZ toolbox. This means a couple things: (1) your current worker counts must be based on 2010 tabulation blocks. Worker counts based on block groups or tracts will not load into the MTPS. (2) No extraneous information (fields, etc) should be added to the the tabblock2010 shapefiles. Also, the data types, lengths, etc. of the existing tabblock2010 shapefiles should not be changed. This could cause the MTPS to not function correctly. (3) The updated worker counts need to be loaded into each county's tabblock2010 file in your coverage. This means that if your MPO has 50 counties, you will need to load your updated worker counts 50 separate times, one time for each county.

Q. Is it required to use the Census Bureau's MTPS TAZ module, or can I write my own code and use the 2010 TIGER files to make my own block equivalency file?

A. The Census Bureau Geography Division requests that all 2010 TAZ and TAD delineation files submitted to the Bureau be output from the TAZ delineation software (TAZ MTPS). This is to ensure that all TAZ and TAD criteria are met, and that the files submitted to the Bureau contain a standardized format and content. However, participants may delineate TAZs and TADs using other types of software by creating block equivalency files (BEFs) to import into the TAZ MTPS.

To clarify; If you've created block equivalency files (BEFs) from the 2010 TIGER/Line shapefiles, then this process should be relatively quick and simple:

If you've delineated your TAZs and/or TADs in a GIS and have the delineations in shapefiles or another GIS file format (but have not created BEFs), you can:

Q What if we ask for all our States' rural counties with the intention of Delineating TAD, but decide in process not to do it.

A There is no "penalty" or consequence for asking for all the counties and then opting to not delineate some of them, however, if you choose to do it this way, CB Geo asks that you either let them know as soon as possible that you have decided not to (within the allotted delineation timeframe of 90 days), so that if an automatic TAD aggregator is available it can be implemented on those counties, or that you, at a minimum, return a firm statement of that you are not delineating, and not just let those counties age past the submittal deadline.

CTPP and ACS Data

Q. Is the CTPP data going to be suppressed at the TAZ level if I make very small TAZs?

A. No, the data will not be suppressed, but it will be of questionable value due to high sample error and small sample size. Right now, regional models are moving toward smaller and smaller TAZs, but there are fewer and fewer ACS records available for CTPP tabulation. Therefore, if the TAZs used in your regional models are small, it is recommend that the TAZs used in the regional models be different than the TAZs used for Census and CTPP data products. The TAZs for CTPP could be an aggregate of 2 or 3 model TAZs.

Q. There was too much data suppression in CTPP 2000 at the tract and TAZ level. Is this going to happen again?

A. No. There is a project underway (NCHRP 08-79) to take all the tables that would be suppressed for disclosure and confidentiality reasons and statistically manipulate them to satisfy the disclosure rules. The company doing the work is Westat, and their method will perturb the ACS microdata records sufficiently to protect individual confidentiality, but still allow for TAZ level tabulation without destroying the underlying statistical reliability of the data.

Q. I don't like the idea of synthetic data.

A. Can you rephrase that in the form of a question? The Census Bureau is already using data synthesis in the ACS, you just didn't know about it. This is just another layer of data synthesis for the custom CTPP. Some of the tables in the CTPP do not need any synthesis applied, so those will be "real." Since the data synthesis will be done at the Census Bureau, the results can be compared against the "true" ACS results, so that the method chosen can be validated before it is implemented.

Imputation is one kind of data synthesis, and imputation has long been used even with decennial census "long form" tabulations for CTPP. For example, household income has a high imputation rate, and place of work has consistently required about 25% imputation for block level geography.

Q. Are they any other planned uses of Census TAZs and TADs after the CTPP 2006-2010?

A. The current CTPP is funded to include the 3-year (2006-2008) tabulation and the 5-year (2006-2010) tabulation, but depending on how things go, future 3-year and/or 5-year CTPP tabulations are possible if someone wants to pay for them. The Census Bureau did not create a geographic tabulation unit for the 20,000 population threshold (for the 3-year ACS), to provide complete coverage, therefore, 3-year ACS tabulations have a "Swiss cheese" coverage of counties and places. This is the transportation data community's opportunity to do something about incomplete coverage. The PUMAs (100,000 population threshold) provide a consistent geographic tabulation unit for the 1-year ACS, but how they have been defined has not always been useful to transportation planners. Although PUMAs were not designed to be tabulation units, they have become tabulation geographic units in the ACS.

Q. Do you think it is worthwhile to spend the time on delineating TAZs or should I just get CTPP tabulations for census tracts or block groups?

A. It depends on how easy or difficult it will be for you to allocate tract or block group data to your model TAZs. Even if you choose not to define TAZs, we highly recommend that you delineate TADs for potential future 3-year CTPP tabulations as there is currently no other geographic entity of equivalent size.

Q. How good or bad are the CTPP data using the ACS survey results going to be at the TAZ level?

A. That is an excellent question, and that is why a TRB Conference on Census data for Transportation Planning is scheduled for October 27-29, 2011. We hope that you will have time to review and use the 2006-2008 CTPP, and also the 2005-2009 ACS, so that you can share your experiences with small area ACS results. This kind of evaluation will be critical to determine the future of the CTPP program, especially to consider alternate data sources for travel flows for small geography.

Updated: 07/29/2013
HEP Home Planning Environment Real Estate
Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000