Today, we mailed a few more states, in particular,
TN, KY, MD, AK, IL, WV.
What are we still missing: IN, MI, NC, OH, SC, VA, and WI.
If your MPO area includes multiple states, we are waiting to have all the
TIGER files for your area before we mail your package. I apologize for the
inconvenience. Also, I am missing 5 TIGER files for miscellaneous states, so
you may be among this list. We hope to complete all the mailouts by the end
of next week (May 7), that is, if we receive the TIGER CDs from the Census
Bureau.
Other information:
One MPO has already submitted their TAZs for 2000 to the Census Bureau. Most
of it looked excellent. The MPO renumbered their entire area, and created many
new TAZs. However, there were 2 small polygons which did not get correctly
assigned NEW TAZ numbers. They still had their OLD TAZ number, and since the
numbers were not overlapping, the contiguity check did not identify them as a
problem. We recommend that before sending in your file, to please visually
scan the entire county area to see if there are any small polygons (they would
be a different color), that have not been reassigned as you may want.
Finally, we will be sending out shortly a list of each COUNTY in each STATE.
We will be asking which MPO or STATE, or combination will be the responsible
party for each county. You should be talking to neighboring MPOs and with
your State DOT to coordinate any counties for which there may be some question
about who will take responsibility. There can only be ONE TAZ definition for a
particular piece of geography.
Funding for the CTPP itself. As you know, AASHTO passed a resolution to
support CTPP as a pooled fund project last fall.
AASHTO will be sending the letters of commitment to the State DOTs very
shortly. I understand they will want them signed in about one month's time.
The cost is approximately 1.1 cents per person, for year 2000 population.
Thus, for example, an area with a population of 4 million, the cost is about
$44,000. The total package cost for the entire U.S. is a little over $3
million. FHWA has agreed to waive the local match (for SPR and PL funds) for
this project--this means for the CTPP package itself, as produced by the
Census Bureau, and NOT for all MPO and State Workplan tasks for such related
work such as the TAZ-UP and LREF. In addition, we expect that FTA will also
waive the local match for SPR funds.
------
I'll be out of the office from May 4-6, 1999.
Yours truly,
Elaine Murakami
phone 202-366-6971
I got this note and thought it was very relavant to the statewide taz
discussion.
Simons, Chris wrote:
> Please add me to your mailing list for future CTPP information &
> correspondence. Thanks!
>
> I am Georgia DOT's primary travel demand modeler. Can I ask who you are and
> how you are involved with upcoming CTPP data structures and tools? I've got a
> few enhancements I'd like to convey if you will be having input into the 2000
> CTPP.
>
> I got a copy of your recent message about polling states regarding the need
> for statewide TAZs. I would comment that having some sort of TAZ structure
> for states could prove useful. In Georgia, census tracts or counties could
> serve as TAZs for statewide level models & analysis. I would note that some
> states have very large counties, so counties may not sufficiently subdivide
> the state for some types of analysis. If there were more levels of analysis
> available in the CTPP (i.e., census tracts or block groups) and the polygons
> could be exported in a GIS format, users could create their own TAZs. For
> metropolitan areas, I would like to be able to query and export information,
> preferable as ArcView shape files, at the county, census tract, and census
> block group levels in addition to those available in the 1990 CTPP.
mike, i hope you don't mind but i felt this was an issue for the list
serve. the concept of a dual tabulation (tracts and tazs) within MPO
areas has come up at several different times. i believe that the current
thinking is that inside mpos the tabulation for worker data and the flow
data will be at the taz geography only. however, the table content is
still being developed so i believe this is an open issue. to me you
make a very valid point. at a tract or coarse level the flow data
becomes higher in quality but at a smaller area, like a taz, the
household demographics are needed and quite reliable.
it has also been my understanding that we are working towards a ctpp
product that has TAZs within mpo areas and whatever the states' want
outside the mpo areas. the only parameter would be that either zonal
system nest within counties and thus states. if the states can work
with their mpos to have their tazs nest within tracts, as chuck purvis
of san francisco's mpo is doing, then an equivalency file would take
care of the tract issue. the only difficulty there might be with
variables using medians but a good approximation could certainly be
calculated.
we have also talked about the possibility of developing a (for cost)
special tabulation for those who wish to get something above and beyond
the ctpp. right now i can't think of any examples except maybe some type
of 8-way cross tabulation no one else ever thought of.
mike i may not have given you any definitive answers but i would suggest
that you keep in mind the concept of having TAZs nest within other zonal
systems and the use of equivalency files for aggregation. maybe we can
get some creative folks to develop an "on the fly" buffering ability
that can, with the push of button, produce data summaries for whatever
zones we desire. coming out of 1990 we at cats in chicago were dabbling
with this idea but then conformity issues and a long range took us
over.
MIKE JAFFE wrote:
> To: Ed Christopher
> I have a question that I originally asked Tom Mank
> about last December (but unfortunately didn't get an
> answer), and I think its important for other MPO's to
> know what the census is planning.
>
> We're participating in the TAZ-Up program because
> we want certain census data (including CTPP data)
> tabulated at our TAZs. This is particularly helpful in the
> development of our household sociodemographic
> cross tabular data (e.g number of workers by
> household size) that are inputs to our travel model.
> On the other hand, CTPP at the Census Tract level is
> useful for more aggregrate analysis, such as journey
> to work patterns and mode choice, where the
> statistical reliability at the census tract level is
> preferred. Therefore, ideally we need the CTPP data
> summarized for both TAZs and Census Tracts within
> MPOs; outside MPOs, as Ed Arabas notes in his
> email, census tracts are more appropriate.
>
> Therefore, I need to know whether MPOs will get
> Census data for both TAZs and Census Tracts within
> their MPO boundaries. The 1990 CTPP came in two
> geographic levels: the urban level CTPP and state
> level CTPP. Will this be the format for 2000, and if so,
> can the urban level have the MPO TAZs and the State
> Level have census tracts, including census tracts
> inside the MPO?
>
> I look forward to your response.
>
> Mike Jaffe
> Senior Transporation Planner/Modeling
> Salem Keizer Area Transportation Study (MPO in
> Salem, Oregon)
>
> Phone: 503-588-6177
> email mjaffe(a)open.org
First, census tracts (and block groups) are not designed with traffic
forecasting in mind. Tract boundaries do not reflect the access to the
transportation network that you are trying to model. Depending on the level
of detail that you need, they may "straddle" several transportation
facilities. If you want to save a lot of front-end time by using a unit of
geography already defined by others for other purposes, that is your choice,
but don't be surprised later on if you get bad answers at the other end of
the process.
Second, I'm curious to know on what basis you consider QRSII a "simplistic
model." If a simplistic approach to forecasting is all that your project
requires, that's fine, but that's a self-imposed user limitation rather than
a software limitation. (I do recognize that this model does allow for some
very elementary setups that others won't, making it the model of choice of
many agencies who put forth minimal efforts in forecasting, but it is not
limited to only these setups.)
> -----Original Message-----
> From: Ed Christopher [SMTP:berwyned(a)mcs.com]
> Sent: Friday, April 30, 1999 7:53 AM
> To: Jack Pascoli; ctpp maillist
> Subject: Re: [CTPP] statewide TAZs
>
> Jack, first as you can see i posted your note to the listserve. this is
> exactly the type of information we need to hear and i think other would
> benefit. as to the specific of the zone sizes if it were a perfect world
> and
> we were doing typical travel demand forcasting we would want to look at
> trips procuded and trips attracted and use those numbers to ease into
> zones
> that are all realatively ballanced on a travel basis. i suppose this is
> one
> reason why folks gravitate to census tracts. in theory they are balanced
> atleast on the residents side. other issues play into the zone creation as
> well like geography and network configuations.
>
> it sounds like counties are certainly a geography that is too rough for
> you
> but something smaller would work. given the context of what the ctpp is
> (resident location data, work location data, and data on the flows between
> zones) and where you might speculate the state is heading, could tracts be
> a
> suitable geography or would you prefer something smaller? in one hand
> sounds like you have the population base and travel demands to develop you
> own TAZs in the non-MPO counties. on the other hand tracts are more off
> the
> shelf and a decent geography to use as a starting point. ultimately, you
> will be making the call.
>
>
> Jack Pascoli wrote:
>
> > Ed,
> > In West Virginia, we have several county studies that are not in MPO
> > areas. We use essentially the same procedures that an MPO would use.
> If
> > we had the information from the Census that we use as our independent
> > variables on a TAZ basis it would save us a lot of work in trying to
> > manipulate the block group data into TAZs. In addition to
> > that, we have many projects that will have a profound effect on our
> > smaller communities that we would like to use simplistic models (QRS
> II)
> > to determine the traffic impacts of project alternatives.
>
> > While we do not have a statewide model at this time, I suspect that in
> the
> > next ten years we will. Depending on the requirements for the work to
> > develop TAZs for CTPP, as a minimum we would expect to coordinate the
> TAZs
> > in the five (5) separate active county studies we currently have. As a
> > maximum we would do the whole state outside of the
> > MPO areas.
>
> > I would appreciate your comments.
> > Jack Pascoli P.E.
> >
Jack, first as you can see i posted your note to the listserve. this is
exactly the type of information we need to hear and i think other would
benefit. as to the specific of the zone sizes if it were a perfect world and
we were doing typical travel demand forcasting we would want to look at
trips procuded and trips attracted and use those numbers to ease into zones
that are all realatively ballanced on a travel basis. i suppose this is one
reason why folks gravitate to census tracts. in theory they are balanced
atleast on the residents side. other issues play into the zone creation as
well like geography and network configuations.
it sounds like counties are certainly a geography that is too rough for you
but something smaller would work. given the context of what the ctpp is
(resident location data, work location data, and data on the flows between
zones) and where you might speculate the state is heading, could tracts be a
suitable geography or would you prefer something smaller? in one hand
sounds like you have the population base and travel demands to develop you
own TAZs in the non-MPO counties. on the other hand tracts are more off the
shelf and a decent geography to use as a starting point. ultimately, you
will be making the call.
Jack Pascoli wrote:
> Ed,
> In West Virginia, we have several county studies that are not in MPO
> areas. We use essentially the same procedures that an MPO would use. If
> we had the information from the Census that we use as our independent
> variables on a TAZ basis it would save us a lot of work in trying to
> manipulate the block group data into TAZs. In addition to
> that, we have many projects that will have a profound effect on our
> smaller communities that we would like to use simplistic models (QRS II)
> to determine the traffic impacts of project alternatives.
> While we do not have a statewide model at this time, I suspect that in the
> next ten years we will. Depending on the requirements for the work to
> develop TAZs for CTPP, as a minimum we would expect to coordinate the TAZs
> in the five (5) separate active county studies we currently have. As a
> maximum we would do the whole state outside of the
> MPO areas.
> I would appreciate your comments.
> Jack Pascoli P.E.
>
thanks ed. i am reposting your comment to list. as to my county comment i made
it for a coupla reasons. first, to get something for folks to react to but also
because this time through, all TAZs even inside MPO areas will be nesting inside
county boundaries. knowing that tract sized TAZs will work for you is important
to know--thanks agian.
Edward.P.ARABAS(a)odot.state.or.us wrote:
> Ed,
> Since I am not on the CTPP maillist, your message was forwarded to
> me by Bill Upton. Hence, I do not think that this response will be
> broadcast to the ctpp maillist. If you are so inclined, you are welcome to
> post it to the maillist.
> The Oregon Department of Transportation is currently building a
> second-generation statewide model integrating land use, transportation, and
> the economy. Descriptions of our modeling efforts can be found at
> http://www.odot.state.or.us/tdb/planning/modeling/modeling.html. As far as
> the construction of our statewide TAZs was concerned, we basically use
> aggregations of census tracts. In the MPO areas, this works out well
> because they have created TAZ districts that respect census geography. We
> did not create any statewide TAZs that were larger than a county, and in
> only two cases did we create TAZs that cross county lines (due to the
> effects of physical features). As long as CTPP continues to report
> transportation-related responses at the census tract level, we are
> relatively happy. In a fairly rural state like Oregon with its strict urban
> growth limits, many trips that might be considered of an intra-urban
> distance are actually inter-urban, and thus the CTPP data are quite useful
> for comparison against our modeling efforts.
> Your suggestion, "... that in non urban areas they should be
> counties. In urban areas they would, of course, represent the TAZs developed
> by the MPOs...", might be difficult to operationalize for MPOs that do not
> use census geography as a basis for their TAZs. And some non-urban areas
> that are experiencing rapid growth may find data reported only at the county
> level to be too gross for their planning objectives.
>
> Ed Arabas
>
> Associate Transportation Planner
> Transportation Planning Analysis Unit
> Oregon Department of Transportation
> 555 13th Street NE, Suite 2
> Salem, OR 97301-4178
> * Ph: 503-986-4398
> * Fax: 503-986-4174
> * Email: edward.p.arabas(a)state.or.us
>
>
>
> > -----Original Message-----
> > From: Ed Christopher [SMTP:berwyned(a)mcs.com]
> > Sent: Thursday, April 29, 1999 11:08 AM
> > To: ctpp maillist
> > Subject: [CTPP] statewide TAZs
> >
> > This post has two purposes. First to welcome 20 folks to the list serve
> > from states where we have ctpp contacts who have email addresses. There
> > are about 20 states we still need to track down. To the new folks let
> > me formally say" welcome to the CTPP listserve". We created this
> > listserve as a mechanism to keep the flow of information relative to the
> > CTPP moving. If you ever want to get off the listserve or change your
> > address the best thing is to drop me an email. You can also do this
> > through an automatic procedure which I will not bore you with now.
> >
> > The second and CTPP related reason for this post is to inquire about
> > statewide TAZs. Although the voice of one or two is not a
> > representative sample, the folks working on the CTPP are discussing the
> > issue of statewide TAZs and their construction, and could use some
> > input. One of the central questions that keeps coming up is, do we need
> > statewide TAZs? and, will they be useful for states?
> >
> > one could argue that the CTPP is journey to work data and it represents
> > short trips, mostly intraurban travel, when one considers statewide
> > travel in some of the larger (spatially big) states. On the other hand,
> > there are smaller states where the opposite is true. The bottom line of
> > all this would be the type of planning underway and the zonal structure
> > behind it. Ultimately it may take a state-to-state poll to identify who
> > needs what. But for now we could use some thoughts on this. Anyone
> > care to comment on the issue of statewide TAZs and more importantly what
> > they should be?
> >
> > To get things moving let me suggest that in non urban areas they should
> > be counties. in urban areas they would, of course, represent the TAZs
> > developed by the MPOs.
> >
> > (although i welcome your replies, i would hope that we keep this
> > discussion on the listserve by making sure you reply to listserve
> > address)
> >
> > thanks
> > ed christopher
This post has two purposes. First to welcome 20 folks to the list serve
from states where we have ctpp contacts who have email addresses. There
are about 20 states we still need to track down. To the new folks let
me formally say" welcome to the CTPP listserve". We created this
listserve as a mechanism to keep the flow of information relative to the
CTPP moving. If you ever want to get off the listserve or change your
address the best thing is to drop me an email. You can also do this
through an automatic procedure which I will not bore you with now.
The second and CTPP related reason for this post is to inquire about
statewide TAZs. Although the voice of one or two is not a
representative sample, the folks working on the CTPP are discussing the
issue of statewide TAZs and their construction, and could use some
input. One of the central questions that keeps coming up is, do we need
statewide TAZs? and, will they be useful for states?
one could argue that the CTPP is journey to work data and it represents
short trips, mostly intraurban travel, when one considers statewide
travel in some of the larger (spatially big) states. On the other hand,
there are smaller states where the opposite is true. The bottom line of
all this would be the type of planning underway and the zonal structure
behind it. Ultimately it may take a state-to-state poll to identify who
needs what. But for now we could use some thoughts on this. Anyone
care to comment on the issue of statewide TAZs and more importantly what
they should be?
To get things moving let me suggest that in non urban areas they should
be counties. in urban areas they would, of course, represent the TAZs
developed by the MPOs.
(although i welcome your replies, i would hope that we keep this
discussion on the listserve by making sure you reply to listserve
address)
thanks
ed christopher
In reference to Mr. Purvis' recent posting, the Census Bureau would like to
respond as follows:
With regards to the amount (2,445) of invalid TAZ boundaries that are
contained within Mr. Purvis' area.
Large numbers of invalid boundaries may in fact represent much smaller numbers
of actual boundaries since the TAZ-UP software tallies each line segment as a
separate failure. Thus, a single invalid line could be responsible for many
separate errors.
The Census Bureau requires that participants in all of the 2000 statistical
geographic programs avoid the use of features that we deem to be unsuitable
for use as tabulation block boundaries. Many of these invalid TAZ boundaries
are the result of 1990 census block features that have been either deleted or
realigned. Most changes to the affected TAZ boundaries will involve a simple
adjustment to an adjacent feature that corresponds to the original 1990 TAZ
boundary feature. In a minority of situations, the TAZ boundary is on a
feature that is not properly identified. In those situations the participants
will maintain their TAZ boundaries on those features, and submit a sketch maps
to the RCC that provide the proper identification of these features. (Please
review the TAZ Program Instructions contained in the Reference Section of the
TAZ-UP manual for the complete instructions for how to resolve invalid TAZ
boundaries.)
With regards to the identification of invalid TAZ boundaries as F84 (2000
statistical area boundary).
Originally, all lines with an F84 label were labeled F72 (1990 block
boundary). During the processing to create the TIGER/Line 98 files these
lines were inappropriately relabeled. These lines are not 2000 statistical
area boundaries (census tracts, block groups, census designated places, and
census county divisions), but are the invalid boundaries of the 1990
statistical areas. Participants in the Participant Statistical Areas Program
(PSAP), just like in the TAZ program, are required to move the boundaries off
of these invalid lines.
In addition, none of the current boundaries of legal entities (incorporated
places and minor civil divisions) are identified as invalid boundaries. TAZ
boundaries may follow the boundaries of these entities. However, the
boundaries of these legal entities are subject to change. Changes to these
legal entities may require changes to the TAZ boundaries.
With regards to Mr. Purvis' need to know the 2000 census tract and block group
boundaries.
The Census Bureau recommends the following to all TAZ participants that need
to know the 2000 census tract and block group boundaries.
Contact the PSAP participant and request that they furnish you with a copy of
the PSAP proposal that they submitted to the Census Bureau, or a copy of the
approved plan if that is available. The regional census center (RCC) staff
can furnish you with the PSAP contacts for your area. Please DO NOT request
the RCC's to provide 2000 census tract and block group boundary information.
(The RCCs will supply an addition set of PSAP annotation maps to a PSAP
participant that requests the maps for the purpose of furnishing 2000 census
tract and block group information to a TAZ participant).
Adjust your TAZs where necessary to the proposed or approved census tracts and
block groups, and also ensure that your TAZ plan conforms to all requirements
including the elimination of the use of invalid boundaries.
During the verification phase of the TAZ program, the Census Bureau will
furnish participants with TIGER/Line files that contain the approved 2000
census tracts and block groups. Participants can make the necessary changes
to the TAZs in order to realign them to the approved boundaries, and then
resubmit their TAZ plan to the Census Bureau. Again, TAZ plans must conform
to all requirements. (Concurrently, the PSAP participants also will be going
through a verification phase of their program in which they are allowed to
make adjustments to their census tracts and block groups. Contact them and
inquire if are making any changes at this time.)
If you have any questions or comments regarding this information, please
contact Dave Aultman. His telephone number is (301) 457-1099, and his e-mail
address is: daultman(a)geo.census.gov
I just received this message from the Census Bureau:
We just recently uncovered, using new types of edits, some topologic errors in
three TIGER/Line 1998 files that are part of the TAZ program.
We have corrected the errors and are recreating these files.
The affected files and agencies are:
06013 Contra Costa, CA (Chuck Purvis)
30063 Missoula, MT (Missoula Office of Community Development)
42077 Lehigh, PA (Lehigh-Northampton Cos. Joint Planning Commission)
----------------------
The Census Bureau will send FHWA replacement CDs, and I will mail them out as
soon as possible.
Also, I am expecting to receive TIGER files for many (but not ALL) of the
remaining states this week. Cross your fingers!
TO: CTPP News mailing list & MPO staffs
FR: Chuck Purvis, MTC
We received our copy of TIGER'98 and the TAZ-UP software on April
7th. I already had ArcView 3.1 installed on my 233 MHz PC, actually
on a 1 gigabyte jaz cartridge. ArcView takes up 83 MB of space;
TAZ-UP an additional 61 MB. So far, so good.
Installing county TIGER files into TAZ-UP can be *very* slow for
large urban counties. I started installing Alameda County (CA) on
Friday, April 9 at 4:00 PM. It took 60 minutes to install! My largest
county, Santa Clara County, took 90 minutes to install files of 40.2
megabytes in size. My entire region (9 counties) takes 205 MB in
space. The "log" records that are written as part of this
installation process note that I have 104,600 polygons (this number
of polygons is the number of records that should be exported from
TAZ-UP once all of my edits are complete).
RECOMMENDATION: Use the fastest PC you can afford. I believe the
documentation recommends at least a 200 MHz machine, but I would
highly recommend something like a 450 MHz PC. And I think that a
gigabyte of free disk space should be adequate for most regions.
After installing these county files, my first step was to "validate"
the existing boundaries: "invalid boundaries"; "contiguity errors";
and "completeness errors". This is where I initially started to freak
out: we have 2,445 invalid boundaries in our region; 106 contiguity
problems; and 1 completeness problem. The contiguity and completeness
errors are manageable; the invalid boundary errors were highly
unexpected (and may be unmanageable?)
Many of my invalid boundary errors are "F84" errors which are
basically old 1990 block boundaries (be they tract boundaries or
block group boundaries) that are flagged by the Census Bureau as
invalid because they aren't associated with any physical feature
(e.g., roads, streams, power lines). My spot check of my invalid
boundary errors suggest that 90 percent of mine are "F84" errors. I
even got "F84" errors in a county (Napa County) where I was strictly
using 1990 Census block groups as my 1990 TAZes!
My desire is that my Year 2000 TAZ boundaries are nested on Year 2000
Census Tracts. I essentially want to create what I would call
"user-defined block groups." The big issue is "what are my year 2000
Census Tract boundaries?" This is where you (MPOs working on this
project) will need to contact your regional office census geographer
to check up on the status of your Year 2000 Tract Boundaries. I may
be wrong in my interpretation of this, but Year 2000 Tract boundaries
will be valid CTPP TAZ boundaries.
I had numerous phone conversations this past Friday with our USDOT
and Census folk back in DC, and they've expressed concerns about our
plight regarding these invalid boundary issues. I'm still fuzzy on
this, so I'm hoping that the Bureau and USDOT folks can clear up
these issues before we all get too hot under the collar with all of
these invalid boundary issues. (USDOT is also very concerned because
all of the TIGER files for the entire US aren't out yet :-(
The "contiguity errors" are very manageable. Some of these contiguity
errors are due to correspondence errors that are our fault, leading
to some really unintended results. The rest of these errors are due
to a misallocation of "water blocks" that separate real, physical
islands. An example in my region is Alcatraz Island and Treasure
Island in the middle of San Francisco Bay. They're part of the same
TAZ, but were separated by a "water block" that includes much of the
Bay, plus Fishermans' Wharf and Pier 39. So, a simple solution is to
include the Fishermans' Wharf "water block" with Alcatraz Island, as
one TAZ, and have Treasure Island as its own TAZ.
My point is that I believe that all of the contiguity errors in your
TAZ boundary files can (and probably should) be eliminated. Be sure
to have "water blocks" connect your sandbars and physical islands to
your "mainland" blocks.
In terms of the TAZ-UP software, I have no problems. It works as
advertised, though it's all (ArcView/TAZ-UP) very slow with large
urban counties on a 233 MHz PC. I wish there were "multiple un-dos"
since I tend to make mistakes in batches, but I'll learn to make
only one mistake at a time.
Well, those are my first impressions on TAZ-UP and this TAZ
definition program. Just take it easy and don't freak out (too much)
when all your invalid boundaries start to pile up!
Cheers,
Chuck Purvis,
MTC - Oakland/San Francisco/San Jose metro region.
*******************************************************
e-mail: cpurvis(a)mtc.ca.gov
Chuck Purvis, AICP
Senior Transportation Planner/Analyst, Planning Section
Metropolitan Transportation Commission
101 Eighth Street, Oakland, CA 94607-4700
(510) 464-7731 (voice) (510) 464-7848 (fax)
WWW: http://www.mtc.ca.gov/
MTC DataMart & InfoMart:
http://www.mtc.ca.gov/facts_and_figures/datamart.htm
MTC FTP Site: ftp://ftp.abag.ca.gov/pub/mtc/planning/
Personal WWW: http://home.earthlink.net/~clpurvis/
*******************************************************
