Using GIS to Examine Environmental Injustice
in the South Bronx

The Case of Waste Transfer Stations¹

By Zvia Segal Naphtali, Carlos Restrepo & Rae Zimmerman

Introduction

The U.S. Environmental Protection Agency defines environmental justice as "...the fair treatment and meaningful involvement of all people regardless of race, color, national origin, or income with respect to the development, implementation, and enforcement of environmental laws, regulations, and policies."² Environmental injustice has been defined as the disproportionate exposure of communities of color and poor people, or other vulnerable groups, such as children and the elderly, to environmental risks.³

In the analyses described in this article, Geographic Information Systems (GIS)⁴ techniques and models were used extensively to facilitate and streamline the analysis of demographic and socioeconomic data about people living in close proximity to waste transfer stations and major highways, and to determine whether a disproportionate number of people in communities of color and poor people live in proximity to these sites. The area of application for this analysis was a portion of the South Bronx, New York, as shown in Figure 1.

The analysis of the racial, ethnic, and other socioeconomic characteristics of the population around existing waste transfer stations in the South Bronx was conducted primarily for those residing at various distances within one mile of these stations. A set of four buffers including 1/4-, 1/2-, 3/4- and one-mile radii from the waste transfer stations were evaluated for the relative stability of these characteristics in the immediate vicinity of these sites.

Information on racial and ethnic characteristics was obtained from the 2000 U.S. Census at the Census Block level (which is the smallest geographic unit for which these data from the Census are available). Information on other socioeconomic characteristics such as income, poverty, and housing values was obtained from the 2000 Census data aggregated at the Census Block Group level (which is, again, the smallest geographic unit for which these data from the Census are available). Figure 2 shows the locations of waste transfer stations in Bronx County, with buffers of different radii around them, and Census 2000 Blocks.

The data for the locations of the waste transfer stations were obtained from the New York City Department of Sanitation. Waste transfer stations are facilities where solid waste, such as putrescible trash, recyclables, and construction debris, is consolidated from multiple collection vehicles. The waste products are then compacted and loaded onto larger vehicles to be transported to final disposal sites.⁶

According to the NYC Solid Waste Management Plan, about 13,000 tons of residential trash are discarded every day in the city.⁷ New York City's waste stream also includes an additional 20,000 tons of commercial waste per day. In order to discard this waste, New York City utilizes approximately 60 private waste transfer stations. These facilities are generally located in neighborhoods that are zoned for manufacturing land uses. The South Bronx has many areas zoned for such land uses, which contributes to the fact that the area has a disproportionate share of these facilities. Approximately 15 waste transfer stations, or about 24% of the city's total number of these stations, are estimated to be located there. These stations handle over 31% of New York City's solid waste. Meanwhile, the South Bronx houses about 6.5% of the City's population.⁸

Traffic patterns are also of concern in the South Bronx. More than 3,000 trucks drive through the Hunts Point peninsula of the South Bronx every day. This includes traffic related to waste transfer and other commercial activities. Residents have been complaining for years about the air and noise pollution associated with the traffic patterns in the area. Many residents believe that the diesel fumes associated with traffic generated by these activities are associated with the asthma rates observed in the South Bronx.⁹ These rates are among the highest in the United States and have been steadily increasing since 1980.¹⁰

As mentioned earlier, GIS technology was used in our analysis to examine the socioeconomic characteristics of the people living around waste transfer stations. The general findings with respect to socioeconomic characteristics—including race, ethnicity, household income, poverty, and housing value—of the populations that live within a mile or less of the waste transfer stations' locations are summarized as follows:

1. Populations living in close proximity (within one mile) to the South Bronx waste transfer stations tend to be more Black and Hispanic than in the Bronx as a whole.
2. The socioeconomic characteristics of the population in this area—median household income or value of owner-occupied housing units—are generally lower within one mile distance from the waste transfer stations than in the study area as a whole, the Bronx, and the other NYC boroughs. Supporting this finding is the fact that poverty rates are higher within the one-mile radius.

Modeling & Sensitivity Analysis Using ArcGIS

This study used GIS to create buffers around five putrescible waste transfer stations. Census data on the population, race, ethnicity, and median household income in the buffers and the study area were analyzed and compared to the Bronx and NYC as a whole, as reference areas.

"Sensitivity Analysis" was used to portray how applying each of the different methods of buffering the waste transfer stations using GIS (described below) might have altered the results. Earlier studies that did not employ GIS used different methods for aggregating population characteristics, for example, by using circles, aggregating Census Blocks, or the larger Block Groups. These studies noted how results changed dramatically depending on how areas were defined with distance.¹¹

Buffers around Waste Transfer Stations: Comparing Three GIS-Based Techniques for Estimating Demographic Characteristics¹²

Three methods of buffering were compared as part of the GIS analyses: the polygon intersection method, the areal interpolation method, and the polygon containment method. These can be applied to buffering that involves Census Blocks or Block Groups. Each of these GIS-based methods involves a different approach to identifying the Census Blocks or Block Groups encompassed by the buffer.

The areal interpolation method is the most accurate, but requires calculation of the area of the Blocks or Block Groups intersected by the buffer, and various assumptions associated with such a calculation. The polygon intersection method tends to overestimate the population, whereas the polygon containment technique tends to underestimate population. The three methods are illustrated in Figure 3.¹³

Census Blocks offer the most detailed Census geography and yield the most precise estimates, but only limited demographic information is reported at the Block level. Figure 3 shows the Census Block and Census Block Group geographies used to derive population estimates for a one-mile study area surrounding a proposed bridge location. The shaded Census units show the area from which the one-mile area estimates are determined. Note that the same one-mile buffer is shown in each diagram. The population estimates obtained from the various methods and Census units vary by as much as 12,092 persons, or 4.7%.

The Polygon Intersection Method

Figure 4 shows the application of the polygon intersection method to buffering waste transfer stations in the South Bronx. The polygon intersection method adds the population of each Census Block or Census Block Group within the selected buffer distance, or that intersect the buffer.

Figure 4 shows what happens when the edges of the buffer intersect Census Blocks or Census Block Groups that are relatively large compared to those in the interior of the buffer. Looking at these results, we may ask, "Is a large error created by selecting a substantial area which is really further away than implied by the buffer distance?" The reality is that some of the Blocks or Block Groups are larger because they are sparsely populated or unpopulated.¹⁴ Clearly, using this method of buffering overestimates the population.

The Areal Interpolation Method

The areal interpolation method is considered the most accurate of the three GIS methods. The most common way of doing areal interpolation is the area-weighted interpolation. The use of the method is based on an assumption that populations within the Census units (Blocks or Block Groups) are uniformly distributed. Figure 5 shows how this method is applied to waste transfer stations in the South Bronx.

In order to compute the characteristics of the population around the waste transfer stations using the areal interpolation method, we constructed models (a graphical tool for computer programming) that carry out the necessary steps in an automatic manner. For each Census Block or Block Group, we were able to calculate the ratio of the area included in the buffer to the total area of the Block or Block Group. This enabled us to calculate the portion of the population in a Block or Block Group that was within the buffer, as well as the number of Blacks and Hispanics, and additional socioeconomic variables of interest.

However, the assumption of uniform distribution of population is not valid in the area that is the focus of this study.¹⁵ Even a cursory examination of the Census data for the Blocks and Block Groups that make up the buffers we created support this fact. As Figure 6 shows, a number of Blocks or Block Groups near the waste transfer stations are zoned for commercial or industrial use, or contain parks and have population only at the edges of the Blocks or Block Groups.¹⁶ (There were, however, only a few Blocks or Block Groups that contained only parks with no population.)

The Polygon Containment Method

The polygon containment method underestimates population. This method only includes those Census units that are entirely within the selected buffer distance (in this case 1/4 mile) in estimating population. As Figure 7 shows, it excludes the Census units that are not contained by the buffer. Note that when the polygon containment method was applied to Census Block Groups, buffering five putrescible waste transfer stations, the 1/4-mile radius buffer did not contain any Census Block Groups.

A variation on the polygon containment method is a method provided by ArcGIS in which the centroid¹⁷ of each polygon is used to determine if it is included within the buffer, instead of requiring the entire polygon to be within the buffer. This method is not illustrated in this article.

Examples of Sensitivity Analysis: Comparing Three Methods for Estimating Population Characteristics

Sensitivity Analysis for Estimating Population & Percent Black & Hispanic

The first example, summarized in Figure 8, with details in the Appendix tables referenced at the end of this article, compares estimates obtained using the three GIS methods described above for total population and for estimates of the percentage of Hispanics and Blacks living within buffers around five putrescible waste transfer stations. The results show important differences among the three methods. For example, for the total population estimates within a one-mile buffer around the waste transfer stations, the values range from 160,081 (polygon containment method) to 199,721 (polygon intersection method). For the estimates of the Hispanic and Black population, the differences in the estimates are less pronounced, but slight differences appear. The polygon containment method estimates that 67.3% of the population living within one-mile buffers around the waste transfer stations is Hispanic whereas the estimate using the polygon intersection method is 66.0%. The areal intersection method estimate for Hispanics is somewhere between the other two estimates: 66.6%.

Sensitivity Analysis for Estimating Median Household Income

The second example uses the three GIS methods to estimate median household income for the population within different sized buffers around waste transfer stations in the South Bronx. The results are summarized in Figure 9 with details shown in the Appendix tables referenced at the end of this article. In this case, the differences are relevant for some buffers but not for others. The polygon containment method is not applicable for a buffer of 1/4 of a mile since no Block Groups are included within those buffers. For the 3/4-mile buffers, the polygon intersection and areal interpolation methods produce the same result ($19,500), but the polygon containment method produces a slightly higher value ($20,000). For the one-mile buffers, all three methods produce the same estimate.

Conclusion

GIS is a critically important approach to evaluating environmental justice conditions around waste transfer stations and other facilities that are associated with pollution emissions. There are a number of different ways of applying GIS, and thus, it is important to conduct a sensitivity analysis to determine the most appropriate method and to choose a meaningful size for a buffer zone.

Acknowledgement

The work presented in this article was supported by the South Bronx Environmental Health and Policy Study, a project funded with a Congressional Appropriation sponsored by Congressman José E. Serrano and administered through the U.S. Environmental Protection Agency under grant number 9821520-03. The views expressed in this article are those of the authors, and do not necessarily reflect the views of the EPA.

Appendix Tables

Footnotes

1. This article describes the work of researchers at New York University's (NYU's) Institute for Civil Infrastructure Systems (ICIS), a multi-disciplinary research institute affiliated with the Wagner Graduate School of Public Service that focuses on innovative approaches to infrastructure management, planning, and policy. This work is part of the South Bronx Environmental Health and Policy Study, a collaborative research project that also includes a research team from the NYU School of Medicine, as well as four local community groups, and is funded by the U.S. Environmental Protection Agency (see acknowledgement). The ICIS portion of the project focuses on the relationships between solid waste management, transportation, air quality, and public health in the South Bronx.
2. See: United States Environmental Protection Agency (EPA). Environmental Justice Website. Accessed 23 March, 2007: www.epa.gov/compliance/environmentaljustice/.
3. For a literature review of environmental justice and a discussion of definitions, see Hirschstein, C. 2004. "Waste Transfer and Environmental Justice: A Literature Review". In Restrepo, C. and Zimmerman, R. (eds). South Bronx Environmental Health and Policy Study: Transportation and Traffic Modeling, Air Quality, Waste Transfer Stations, and Environmental Justice Analyses in the South Bronx. Final Report for Phase II & III. Available online at www.icisnyu.org/admin/files/ICISPhaseIIandIIIreport.pdf; see also Maantay, J. (2007). "Asthma and Air Pollution in the Bronx: Methodological and Data Considerations in Using GIS for Environmental Justice and Health Research," Health & Place, 13 (1): 32-56.
4. GIS is a collection of computer hardware, software, and geographic data for capturing, managing, analyzing, and displaying all forms of geographically referenced information. (www.gis.com/whatisgis/)
5. Figure 1 shows the project study area, defined as community boards 1, 2, 3, 4 and 9 in Bronx County, and the location of truck routes and waste transfer facilities.
6. See U.S. Environmental Protection Agency (EPA). 2002. Waste Transfer Stations: A Manual for Decision-Making. www.epa.gov/epaoswer/non-hw/muncpl/pubs/r02002.pdf.
7. The New York City Solid Waste Management Plan is available at www.nyc.gov/html/dsny/html/reports/swmp-4oct.shtml.
8. See Restrepo, C. 2002. "Waste Transfer Stations In The South Bronx." In Zimmerman, R., Restrepo, C., Hirschstein, C., Holguín-Veras, J., Lara, J., Klebenov, D. South Bronx Environmental Studies, Public Health and Environmental Policy Analysis: Final Report for Phase I. Available online at www.icisnyu.org/assets/documents/SouthBronxPhaseIReport.pdf.
9. For a discussion of air quality in the South Bronx see: Restrepo, C., Zimmerman, R., Thurston, G., Clemente, J., Gorczynski, J., Zhong, M., Blaustein, M., Chen, L. C. 2004. "A comparison of ground-level air quality data with New York State Department of Environmental Conservation monitoring stations data in South Bronx, New York." Atmospheric Environment. 38: 5295-5304; and Restrepo, C. 2002. "Characterization of Ambient Air Quality in the South Bronx." In Zimmerman, R., Restrepo, C., Hirschstein, C., Holguín-Veras, J., Lara, J., Klebenov, D. South Bronx Environmental Studies: Public Health and Environmental Policy Analysis: Final Report for Phase I. www.icisnyu.org/admin/files/PhaseIWagner930.pdf.
10. See Restrepo, C. 2002. op. cit.
11. Zimmerman, R., "Issues of Classification in Environmental Equity: How We Manage is How We Measure," Fordham Urban Law Journal, Vol. XXI, No. 3 (Spring 1994), pp. 633-669. See page 654.
12. See "Effective Methods for Environmental Justice Assessment" in the NCHRP (National Cooperative Highway Research Program) REPORT No. 532. Transportation Research Board, Washington, D.C. 2004. www.TRB.org. Page 30.
13. Figure 3 is from the Transportation Research Board. 2004. "Effective Methods for Environmental Justice Assessment," NCHRP REPORT 532. Washington, D.C. National Cooperative Highway Research Program. Page 30. Image used with permission.
14. For a more detailed description see Naphtali, Z. 2004. "Environmental equity issues associated with the location of waste transfer stations in the South Bronx," in Restrepo, C. and Zimmerman, R. (eds), South Bronx Environmental Health and Policy Study: Transportation and Traffic Modeling, Air Quality, Waste Transfer Stations, and Environmental Justice Analyses in the South Bronx. Final Report for Phase II & III. Available online at www.icisnyu.org/admin/files/ICISPhaseIIandIIIreport.pdf.
15. See Naphtali, Z. 2004 (ibid.) for a more detailed discussion of the non-uniform distribution of the population in the areas close to the waste transfer stations in a section on "unpopulated" Census Blocks and Block Groups.
16. The map in Figure 6 was taken from a paper submitted by Nicole Dooskin and Nick Molinari as a final report in an Advanced GIS class, spring 2005.
17. Informally, the centroid is the average of all points that make up an object or area, in this case a polygon.

---

Author Biographies

Zvia Segal Naphtali, PhD, is a research scientist at the Institute for Civil Infrastructure Systems (ICIS), a multi-disciplinary research institute affiliated with NYU's Wagner Graduate School of Public Service. An Adjunct Clinical Assistant Professor of Public Administration, Dr. Naphtali teaches three courses yearly on spatial analysis and the applications of Geographic Information Systems in urban planning, environmental justice, and health. Carlos Restrepo, PhD, is project manager for the ICIS component of the South Bronx Environmental Health and Policy Study and a research scientist at ICIS. His areas of research include infrastructure security, environmental policy, and sustainable development. Rae Zimmerman, PhD, is Professor of Planning and Public Administration at NYU's Wagner School, and Director of ICIS. Her research and teaching areas are environmental protection, infrastructure security, risk analysis and risk management, and natural hazards.

Page last reviewed: April 19, 2007. All content © New York University.
Questions or comments about this site? Send email to: its.connect@nyu.edu.