1 / 47

Tim Trainor Chief, Geography Division U.S. Census Bureau

The Census Bureau's Geographic Support System Initiative – An Update Council of Professional Associations on Federal Statistics September 21, 2012. Tim Trainor Chief, Geography Division U.S. Census Bureau. For the 2020 Census – The GSS Initiative.

darena
Télécharger la présentation

Tim Trainor Chief, Geography Division U.S. Census Bureau

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. The Census Bureau's Geographic Support System Initiative – An UpdateCouncil of Professional Associations on Federal StatisticsSeptember 21, 2012 Tim Trainor Chief, Geography Division U.S. Census Bureau

  2. For the 2020 Census – The GSS Initiative For the 2010 Census – Realigned the street network through the MAF/TIGER Enhancement Program For the 2000 Census – Introduced the Master Address File For the 1990 Census – Introduced TIGER Census Geographic Support – Major Initiatives Over Time

  3. A Change in MethodologyIn Taking a Census

  4. Improving Data Quality 1: Establish quantitative measures of address and spatial data quality Existing MAF/TIGER Data New incoming data 3: Monitor and Improve the quality of the: 2: Assign Quality Indicators to MAF/TIGER data IT processes for updating the MAF/TIGER System Geographic products output from the MAF/TIGER System

  5. Improved Partnerships New Tools Partners TIGERweb Community TIGER Crowd Sourcing Web-based Address Tools Volunteered Geographic Information (VGI) Enhanced collaboration Expand Existing Partnerships Engage New Partners New and Enhanced Programs Enhanced Feedback Address Feedback adhering to Title 13 confidentiality laws Build on and Expand Feedback for Spatial Features Utilize new tools and programs to acquire address and spatial data in the most efficient and least intrusive ways

  6. Research Activities • GSS-I Working Groups • Address Summit • Address Pilots • External Expert Reports • Research Project Examples • iSimple • GSS Lab Data Viewer • Quality Indicators • Census 2010 Road Update Operations Evaluation • Targeted Address Canvassing Continuum

  7. Project/Contract Management Quality Assessments Feature Coverage and Sources MAF/TIGER Integration/ Linkage Policy FY2011 10 GSS-I Working Groups Geocoding Research and Development Address Coverage and Sources Partnerships Global Positioning Systems (GPS) To date, 11 IPTs formed Problem Capture Tool Quality Indicators Improving Group Quarters Data iSIMPLE Metadata Improvements Features Source Evaluation Highway Median “Flag” Improvements Parcel Data and Centroid Use MTAG The CATT Better Meeting MAF “Facility” Data User Needs

  8. Census Address Summit Goals • Educate our partners about the Geographic Support System Initiative (GSS-I) and the benefits of targeted address canvassing • Gain a common understanding regarding the definition of an address • Learn how our partners are collecting, using, and maintaining address data

  9. Address Summit Participants

  10. Participants - All Levels of Government

  11. Observations • Continuous partnerships are needed and welcome • Public safety is a driving factor for local governments • Urban and rural areas will pose different challenges • Address coverage varies and is sometimes not known or quantifiable • Communication and engagement are key

  12. Results of the Address Summit • Five Pilot Projects • Address Authority Outreach and Support for Data Sharing Efforts • FGDC Address Standard and Implementation • Federal/State/Tribal/Local Address Management Coordination • Data Sharing – Local/State/USPS/Census • Hidden/Hard to Capture Addresses

  13. 2012 Address Pilot Schedule

  14. Moving Forward These pilots will provide: • The Census Bureau with a testing ground for future geographic partnership programs • The Census Bureau with an opportunity to identify the best methods for the continual update of the MAF/TIGER System • www.census.gov/geo/www/gss/address_summit/

  15. Benefits of Establishing an Census Address Ontology • Establishing an Ontology allows for • Effective communication • Common language • Ease the burden of data sharing • Explicit terminology, concepts, and relationships

  16. Expert Research at Census • Five reports created by outside experts: • The State and Anticipated Future of Addresses and Addressing • Identifying the Current State and Anticipated Future Direction of Potentially Useful Developing Technologies • Measuring Data Quality • Use of Handheld Computers and the Display/Capture of Geospatial Data • Researching Address and Spatial Data Digital Exchange • http://www.census.gov/geo/www/gss/reports.html • Summer at Census: • Steve Guptill; USGS Chief Scientist (Retired) • Quantifying the Quality of the MAF/TIGER Database • David Cowen; Distinguished Professor Emeritus • Use of Parcel Data to Update and Enhance Census Bureau Geospatial Data • http://www.census.gov/geo/www/gss/qaewg.html • 2 In-Progress Reports • Change Detection • Master Address File (MAF) Evaluation

  17. Analysis of the MAF/TIGER System • iSIMPLE • Evaluation of road features in TIGER • Is TIGER consistent with imagery? • 852,090 grid cells reviewed • 94% had NO missing features • 5% had 4 or less missing features • 70% had NO misaligned features • 26% had 4 or less misaligned features • First web service based review • Research will assist with targeting efforts

  18. iSIMPLE Missing Road Features

  19. GSS Lab Data Viewer • An on-line, interactive mapping tool to facilitate visualization of data and information • Examples include: • 2010 Census Data • Address Canvassing adds • Type A adds • Undeliverable as Addressed • Delivery Sequence File Statistics • Natural Disaster Information

  20. Quality Indicators • Evaluating the current quality of the MTDB • Addresses • Features • Geographic areas • Geocodes • And only evaluate MTDB • Unit of work is the (current) census tract

  21. Address Indicators • Overall Address QIs • Address consistency • Mailability • Deliverability • Locatability • Geocode accuracy • Tests for ‘other’

  22. Feature Indicators • Overall Feature QIs • Spatial accuracy • Feature naming • Address ranges • Feature classification

  23. Geographic Area Indicators • For each Geographic Area, four major tests or sub-indicators • Local review/approval of areas • Regional review/approval of areas • Program review/approval of areas • Independent subject matter review/approval of areas

  24. Geographic Area Indicators • Additional tests for statistical criteria, attributes, type of submission, contiguity, etc… • Also tests for geographic interaction (slivers), and block size and shape

  25. Geocode Indicators • Combines specific sub-indicators from each other category • Locatability and geocode accuracy (Address) • Spatial accuracy & address ranges (Feature) • Block size & shape (Geography)

  26. Overall indicators & weighting • Addresses, Features, Geographic Areas, and Geocodes QIs are then aggregated according to subject matter formulas • Each census tract will receive a single overall score, and category scores where relevant • History and tendency will be tracked

  27. External sources • Quality Indicators are MTDB only • In the future, external sources may also help determine MTDB quality, such as: • Population estimates • Building permits (new development) • Comparison to Imagery • Additional tests to check for completeness of MTDB (omission/commission)

  28. Tract profiles • Additional ability to adjust Quality Indicators based upon profile elements of the tract, such as: • Natural disaster • Unique address types • Rapidly changing development • Special land use areas

  29. Rapid Landscape Change: Picher, OK Census 2000: 708 housing units 621 occupied 87 vacant 2010 Census: 30 housing units 10 occupied 20 vacant

  30. The Result • All census tracts will be tested and ranked • Work and updates can then be targeted to specific areas most in need of update • Prioritization of internal work • Prioritization of partner contact and file ingestion • Improved resource allocation

  31. 2010 Road Update Operations Evaluation Project Scope: The project evaluated the spatial accuracy of new road edges added to the MAF/TIGER database (MTDB) by 2010 Decennial Update Operations. The Decennial Operations in the Study: • Address Canvassing • Update Leave • Update Enumerate • Enumeration at Transitory Locations • Group Quarters Enumeration • Group Quarters Validation

  32. 2010 Road Update Operations Evaluation Hypothesis (1):By using imagery to systematically assess the spatial accuracy of road edges added by different operations, we can choose update methods that consistently produce higher quality linear features.

  33. 2010 Road Update Operations Evaluation Hypothesis (2):Road updates made with GPS were more spatially accurate than paper-based road updates.

  34. 2010 Road Update Operations Evaluation Project Phases: SQL Metrics– Queried MTDB for counts of new road edges added during 2010 operations, by county. Sample Design– Worked with DSSD, to design a sample of counties, as random as possible, that would include all Operations and all Regions. Spatial Evaluation– Assessed selected edges, overlaid on imagery. Tested spatial accuracy of the imagery to a CE95 of 5 meters or less. Data Analysis– With DSSD, obtained metrics from observations. Conclusions

  35. 2010 Road Update Operations Evaluation We looked at over 42,000 edges… in 72 counties…….

  36. Conclusions • Road updates made with GPS were more spatially accurate than paper-based road updates. • An estimated 90% of road edges added with GPS were spatially accurate. • An estimated 67% of road edges digitized from paper-based operations were spatially accurate. • By using imagery to systematically assess the spatial accuracy of road edges added by different operations, we can choose update methods that consistently produce higher quality linear features.

  37. Suggestions for Further Study • Find other ways to glean what contributes to spatial quality using the data obtained in this review. • Are edges with SMIDs (Spatial Metadata IDs) more likely to be spatially accurate than edges without SMIDs? • Are roads that were named more likely to be spatially accurate than those not named? • Why was the incidence of roads with no name information 38%? Were road names not collected? • Is there a correlation between the use of NAIP imagery and the number of edges not visible in the imagery because NAIP is collected leaf-on? • Is it possible to operationalize or automate this review so that it may be applied at a larger scale? • What other operations that add linear features would benefit from the use of imagery for quality control?

  38. Targeted Address Canvassing Continuum

  39. Targeted Address Canvassing Continuum

  40. Targeted Address Canvassing Continuum

  41. Targeted Address Canvassing Continuum Scores, Census Tract 6069.04, Howard County, Maryland2010 BaseOverall Score = 93.7

  42. Targeted Address Canvassing Continuum Scores, Census Tract 6069.04, Howard County, MarylandCurrent StateOverall Score = 96.9

  43. High Stability Census Tracts Tract 3406, Harris County, TX Tract 4302.03, Fairfax County, VA

  44. Questions?

More Related