Adjusting U.S. Census TIGER/Line Blocks

The Census employs a crude level of spatial accuracy and detail in the creation of TIGER/Line data. This level is sufficient for the collection of the census and to allow demographic GIS analysis of the data. Consequently TIGER/Line data cannot be overlayed with more accurate data. This paper presents County of Contra Costa Community Development Department to recreation of TIGER blocks using our highly accurate local street, hydrological, and orthorectified photo data. Our process and the lessons learned from it will be the focus of the presentation. The session will be of interest to other local governments attempting a similar process. History: Contra Costa County is one of the nine counties in the San Francisco Bay Area and is one of the fastest growing. Contra Costa County has begun implementing an enterprise GIS program. As a planning tool, the ideal will be to have have all land development activities (approvals, permits, subdivisions, etc.) reflected in the GIS. As a government entity, the County has a large number of projects that require the use of census geography in our GIS system. Grants, entitlement programs, travel/traffic modeling, government data all require the use of census geography in one way or another. The County, in the past, has made use of TIGER files for census geography. TIGER files were originally created to support Census Bureau operations. These operations do not require high levels of positional accuracy in their GIS files and their policy is not to develop the data beyond what is needed internally. The result of this situation is that census geography is crude and positionally inaccurate. In contrast, Contra Costa County currently owns, develops or licenses high quality, extremely accurate GIS files including parcel data, street centerline, city boundaries, elevation contours etc. One common result of these two contrasting datasets is that work must be done completely with either TIGER data or Contra Costa data; the two datasets will not register and are not used together for county projects. This is inefficient. The utility of the County’s high quality data is compromised by the fact that any project, which requires the use of census geography, is typically done using all TIGER data. The posistional quality of TIGER data is poor and results in a lesser quality product. Given the situation above the County decided to implement a program to adjust the census block files to register with Contra Costa data, creating a GIS system that is fully integrated with census geography. Contra Costa County is nearing completion on its Census geography adjustment project. With this, the County hopes to have more accurate population forecasts and with that, improved associated/derived statistics, economic development, traffic, housing, schools, etc Introduction: This paper describes the effort by Contra Costa County, California to conflate United States Census 2000 TIGER Blocks boundaries to our more accurate local geography. It discusses spatial and attribute conflation of the Blocks as well as the TIGER polyline features composing the Block boundaries. TIGER conflation is not something to be attempted if your municipality is new to GIS. It takes a substantial investment of time, no small amount of GIS skills, a well thought out systematic approach, and a comprehensive digital geodataset inventory of local features. The smaller the demographic unit you wish to adjust the more difficult the project becomes. Tracts are fairly painless, block groups are numerous and more difficult, and blocks are a major investment of effort. Efforts to affect TIGER conflation generally follow two approaches: a manual method and a hybrid of automatic feature matching and user error correction. This project utilized a manual approach to conflation. A hybrid approach was considered but rejected because of the technical difficultly of automation, the high chance of errors and the lack of some the spatial data needed to do automation. For a good example of a hybrid approach see Delaware County’s (Ohio) participation in the Census’s Local Update of Census Addresses program [1].