Here is a copy of a post I made back in March about the relationships to
Tracts, Zip codes, Census Bock Groups and Census Blocks, the full post is
here with views to larger map images:
http://www.dslreports.com/forum/r22157982-National-Broadband-availability-a-
simple-solution-to-mapping
National Broadband availability a simple solution to mapping
[del]
Zip Code Tabulation Areas (ZCTA Polygons)
[del]
Census Tracts
[del]
Census Block Groups
[del]
Census Blocks
(smaller thumbnails)
The 350 million dollars allocated for a national broadband mapping is way
more than necessary. Read through this message to get an idea of the issue
and examine the attached maps to see what we are dealing with using any
particular level of mapping detail. This is obviously just my opinion but
one worth consideration.
I have attached map images of Tom Green County, Texas with the different
polygons the Census Bureau uses in their demographic tabulations. I chose
this county because it seems to be a decent cross section of rural America
but also has a high population density area.
»en.wikipedia.org/wiki/Tom_Green_···y,_Texas
Here are the raw numbers, but you need to look at the attached images to see
how the totals can be deceiving when compared to the map:
Zip Code Tabulation areas = 13 Polygons (These polygons are made up by the
Census Bureau, the post office does not create zip code polygons, zip codes
are linear routing for them)
The FCC already has this data collected using the Form 477.
Census Tracts = 23 Polygons (look in the rural areas outside San Angelo to
see that they are actually much bigger than the zip code areas)
This is the level of reporting required on the new Form 477.
Census Block Groups = 101 Polygons
Census Blocks = 5241 Polygons (even in the rural areas these are much
smaller than Tracts or Zip Codes). Blocks are the most granular level
studied by the Census.
The problem with the FCC data in the current state is, if there is just one
single customer reported as served in a polygon, they show the whole area as
being served by broadband. We know the number of households in each of the
polygons (Census 2000 Figures). If the FCC totaled the number of subscribers
for all form 477 respondents (by zip code) and then divided that by the
total households, we could have a percentage of the households served within
each polygon. This would be much better than an all or nothing reporting
method. This would also not cost anywhere near 350 million dollars to report
broadband availability to the public. If the total subscribers was
aggregated by all carriers (removing the data for Satellite Internet), you
would not know the specific totals for each provider, thus preserving
private information.
Just thought I would post this for all to see and become familiar with the
issue.
Thank You,
Brian Webster
-----Original Message-----
From: wireless-boun...@wispa.org [mailto:wireless-boun...@wispa.org]on
Behalf Of Martha Huizenga
Sent: Wednesday, May 06, 2009 12:59 PM
To: WISPA General List
Cc: w...@part-15.org; Motorola Canopy User Group
Subject: Re: [WISPA] Response to the FCC Regarding Form 477
Hi Matt,
This was a great note to the FCC! Well done. This doesn't help me, since
I am an urban WISP, but my guess is that as you stated a lot of rural
WISPs either had this problem and knew it and just decided to file
inaccurate data or didn't know the data was inaccurate.
It certainly should help the FCC to know that they chose a parameter
that wasn't as easy as they stated (here is a list of vendors for
Geocoding info : ). I know I had to do mine by hand and contemplated
just picking a few tracts to enter for all my customers, which would
have been very inaccurate.
Martha
Martha Huizenga
DC Access, LLC
202-546-5898
*/Friendly, Local, Affordable, Internet!/**/
Connecting the Capitol Hill Community
/*
Matt Larsen - Lists wrote:
> I thought I would share this email that I just sent to the FCC regarding
> the Form 477 report. I am late filing this report because we don't have
> accurate data and thought that my reasons why were worth sharing with my
> colleagues. I support what the FCC is trying to do with Form477, but was
> not able to in good conscience turn in our data by the report deadline.
>
> I hope that this is valuable to some of you out there.
>
> Matt Larsen
> vistabeam.com
>
> ---
>
> Hi Suzanne,
>
> I am not really in a position where I can give a projected date to have
> this information completed for you. However, I do feel it would be
> valuable to explain why and provide you and your management some more
> information as to why I am unable to give you a better date on when we
> intend to have it completed.
>
> For background, Vistabeam (Inventive Wireless of Nebraska) is a wireless
> ISP that covers about 40,000 square miles in Nebraska and Wyoming. We
> have around 2000 customers spread out across this very thinly populated
> area. Even though we are quite small in customer number compared to
> other ISPs, we have a very good billing and provisioning system and
> quite a bit of detail on our customers. However, we did not have census
> tract information for our customers as there had never been a need for
> it until the latest Form477 notice came out earlier this year.
>
> Once we received the Form477 notice, we made plans to modify our billing
> system to add the census tract information, which we were successful in
> doing. We also studied how to obtain geocoding information from multiple
> sources and how to integrate this into our database so that we could
> complete the report. Our initial integration seemed to be successful
> until we started to look at the geocoding data that we received and
> realized that over 50% of the census tract information was invalid.
>
> After going through this data, we found that many of the addresses we
> have for customers are simply not being processed and located correctly.
> The majority of our customers are in rural areas with references to “CR”
> and “Road xxx” and other rural address forms that the geocoding engines
> simply cannot process. Many of these rural counties do not have GIS
> departments with the ability to provide the geocoding information for
> these addresses. In the event that the address doesn’t code, the
> geocoding engine returns the census tract information for the nearest
> Post Office, which is not in the correct census tract.
>
> To get the correct information, we basically have two options.
>
> Option #1 is to drive out to every customer with a GPS unit and record
> the information into our system. Since we have approximately 1100
> customers with inaccurate information, this is going to be a time
> consuming process and would cost us several thousand dollars to collect
> – not to mention the lost man hours.
>
> Option #2 is to go through each customer record and use Google Earth and
> the driving directions to each customer location to determine the census
> tract. This takes about five minutes per customer record, so we are
> looking at about 92 man hours to get that data assembled and inserted
> into our customer database.
>
> We have chosen to go with Option #2 to collect the invalid census tract
> data. However, I do not have the manpower to devote dedicated time to
> this data collection so we have distributed this project among several
> employees and are making as much progress as we can when our workflow
> allows for it. After a month, we are about 10% of the way through it. We
> are now entering our slower time of the year, so hopefully we will make
> a little bit better progress on it going forward, but I cannot make any
> guarantees on when we will get the data completed.
>
> This leaves me with a quandary – I can either provide you with timely,
> but inaccurate information that is going to skew your data, or I can
> take the time to get the information right. Unfortunately, 99% of the
> completed Form477 reports that you have received probably have a
> substantial amount of inaccurate data in them.
>
> II can send the inaccurate data that we have, and then you can check us
> off the list. That is probably what we will end up doing. In reality, we
> probably won’t have a truly accurate report until the next one is due.
>
> I would be happy to provide a computer, Internet connection and a quiet
> room for an FCC intern if you would like to send someone out to
> participate in the data collection process. I realize that this is not a
> likely possibility, but I figured it doesn’t hurt to make the offer.
>
> I really do appreciate the thought process behind collecting this
> information. I am one of the founding members and past president of
> WISPA, the Wireless ISP trade association, and we have actively
> encouraged our members to complete this report and comply with FCC
> regulations regarding our industry. I want to comply with the data
> reporting requirements of the Form477 report and will commit as much of
> my available resources as I can to get the data requested and make sure
> that it is as accurate as possible.
>
> Thank you for your time and consideration,
>
> Matt Larsen
>
> Vistabeam.com
>
>
>
>
> --------------------------------------------------------------------------
------
> WISPA Wants You! Join today!
> http://signup.wispa.org/
> --------------------------------------------------------------------------
------
>
> WISPA Wireless List: wireless@wispa.org
>
> Subscribe/Unsubscribe:
> http://lists.wispa.org/mailman/listinfo/wireless
>
> Archives: http://lists.wispa.org/pipermail/wireless/
>
>
----------------------------------------------------------------------------
----
WISPA Wants You! Join today!
http://signup.wispa.org/
----------------------------------------------------------------------------
----
WISPA Wireless List: wireless@wispa.org
Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless
Archives: http://lists.wispa.org/pipermail/wireless/
--------------------------------------------------------------------------------
WISPA Wants You! Join today!
http://signup.wispa.org/
--------------------------------------------------------------------------------
WISPA Wireless List: wireless@wispa.org
Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless
Archives: http://lists.wispa.org/pipermail/wireless/
