From the Department of Urban Studies and Planning at MIT
Boston, Massachusetts, Develops an Industrial Archaeology Mapping Project With GIS
By Kris Kolodziej, Project Manager, Massachusetts Institute of Technology
Many cities in America have a long and rich history of varied industries and land uses. Despite the improvement in decision making processes, recent decisions on types of land use have often been made without any relation to a site's past. A parcel of land used for residential or commercial building might once have been a leather tannery, lead smelter, or foundry. Add the current mix of residential, commercial, and industrial zoning and land uses to this historic patchwork of successive industries and you are left with a complex tapestry of in situ conditions with potential environmental health concerns.
The Department of Urban Studies and Planning at the Massachusetts Institute of Technology (MIT) is developing a procedure to link 120-year-old fire insurance maps from Sanborn Maps with current land use to use as a planning tool. A GIS mapping database and a customized GIS application have been developed to map the determinant factors of potential environmental hazard exposures.
This project's aim is to produce mapping tools for professionals in government and private industries such as public health officials, legislative branches of government, city and environmental designers and planners, and private developers. The investigators of this project are exploring the various uses of such a tool to conduct research and design health intervention programs in a way that enables them to target the zones of greatest concern. Moreover, communities can utilize this mapping tool as a community information system in an attempt to mobilize around issues of health and potential risk. Developers and communities, in partnership, can use the mapping to make better investment decisions and to allocate funds more aggressively for cleanup.
While the initial pilot project concentrates on a single neighborhood, the final maps will show snapshots of an entire city over a 100-year period, each slice of time showing zones of any significant industrial activity and areas of potential contamination (along with the types of land uses and chemicals of concern that potentially could have been released or emitted). A composite overlay of this historic data with present data layers (e.g., U.S. Census demographics) can then be prepared to allow the user to show the combined exposure risk potential accumulated over time and to delineate zones according to predefined degrees of potential historic contamination. Raul Lejano, professor of urban studies and planning, says, "The vision of this effort is one of being able to combine the spatial with temporal and, in so doing, better understand how some communities must coexist in the present with a history of neglect."
This project involves three steps: (1) heads-up digitizing and data entry of historical information (e.g., facilities found on Sanborn fire insurance maps), (2) identifying and assigning the pollutants associated with the historical use/facility, and (3) GIS data analysis. ArcView 3.x was chosen for this project because of its presence in the department's computer labs and the project team's familiarity with the software.
First, Sanborn maps, which are the primary source of data for this study, were chosen because they are comprehensive. They provide valuable, detailed, historical information on building typology, building footprint, and land use as well as physiographical information on the landscape. This project considers three time periods: 1888, 1900, and 1962.
The heads-up digitizing process begins with a careful surveillance of Sanborn sheet maps to identify any industrial facility that is suspected to have relevant effects on public health. Once the facilities are identified on Sanborn sheets, they are then transferred individually onto a present-day digital map of the study area in ArcView comprising three themes: parcels, Topologically Integrated Geographic Encoding and Referencing (TIGER) system roads, and orthophotographs. Each industrial facility is entered as a point of approximate geographical orientation to matched parcels. The attributes of each point include a unique identifier, the name of the facility, and the type of facility according to the Occupational Safety and Health Administration (OSHA) Standard Industrial Classification (SIC) or, as soon as cities and other agencies are able to convert, to the new North American Industry Classification System (NAICS). However, there is a limitation associated with the Sanborn maps and entry of the SIC/NAICS codes. Sushila Maharjan, a research assistant, asserts that "the Sanborn maps do not tell you much about the type of industrial facilities. So we often have to use our best discretion to assign SIC/NAICS codes to these facilities."
Second, in order to gauge the severity, persistency, and overall risk to public health from each of the mapped facilities, Professor Lejano conceived a table of pollutants associated with four-digit SIC codes found at Superfund sites and the Accidental Release Information Program (ARIP) database. The lookup tables are used to link pollutants to facilities that have SIC/NAICS indentifiers that are either listed on the Superfund list or the ARIP list.
Third, ensuring work on GIS analysis involves finding out the "hot spots" and neighborhood statistics for data correlation purposes of health data and location and/or hazard ranking of these hot spots. The process involves assigning hazard rankings for each facility based on the persistency and severity of the pollutants associated with it. Then, a customized ArcView function is used to perform such statistics as sums and averages of the hazard rankings for all facility points within an area of interest such as a census tract. Avenue script was used to perform needed tasks such as point-in-polygon analysis. For example, a user of the final product may want to explore which areas have the most hot spots for the user to investigate in a particular workweek. Moreover, the user's criteria for investigating a particular hot spot area might be census tracts that have the highest percentage of children under the age of five with a blood lead threshold of about 10 micrograms/deciliter. In such a way, epidemiological analyses can be carried out using this tool.
Chika Sassa, a research assistant, says, "I'm really excited about this project because I believe that digital cartography, and the spatial database that goes with it, would greatly aid public health officials in visualizing risk, identifying the hot spots, and devising policy interventions accordingly." She hopes that this project will also serve as an intimate survey into the historical making of cities' landscape, something of great interest to her as a landscape planner. In addition, she says, "Through my work of digitizing the facilities data layer, I came to understand how something so intangible and invisible such as 'risk' could be broken down into its spatial and temporal components and be represented graphically on a map."
Professor Lejano says, "Our project is an effort to move mapping beyond representations of concrete, immediate spatiality and onto other complementary logics, whereupon we realize upon reflection that mapping has always been about more than just representing the physical." Researchers at MIT are hoping that the maps may enhance the understanding of how communities find themselves having to bear a history of neglect and, more important, improve the possibility of creating a better and healthier future for the public.