Local Sources of Air Pollution and Environmental Health
in New York City
Andrew Maroko1, 2, Juliana Maantay*,1, 2, and Rosa Perez1
1Lehman
College, CUNY; 2The Graduate Center, CUNY; *Advisor
GOAL: To analyze
and quantify the potential association between (or the contribution of) air
pollution from local sources and
selected health outcomes in New York City.
HYPOTHESIS: There is a positive correlation between exposure to air
pollution from local sources and
increased risk of hospitalization for selected cardiovascular diseases and
asthma in New York City.
Heterogeneity: Left: The orthophoto
and cadastral map of a city block in Manhattan show uneven distribution of
land use and residential units at the tax-lot level. Right: Heterogeneity
is shown at the block group level in the Bronx. (a) aerial photo; (b) land use map; (c)
Population estimate and limited access highway (Bruckner Expressway) 150m
pollution buffer. Note that CEDS estimates most of the block group’s residents
live outside of the impact zone.
AERMOD inputs: Left: Only buildings within a distance of 5x
their own height need to be included in
the model.
Right: Emission release points(stacks), populated tax lots, and irregular receptor grid in NYC.
Right: Emission release points(stacks), populated tax lots, and irregular receptor grid in NYC.
CO Dispersion in the Bronx: Average annual
CO concentrations (classified by quantile) from NEI stationary point sources
in the Bronx, NY. Note the difference between dispersion modeling
and proximity analysis.
Spatial
distribution of hospitalization rates: All
the data, except for conductive
disorders, appear be spatially clustered.
Proximity
analysis results: As can be seen by the
bar chart (below), only proximity to toxic release inventory facilities (TRI) seem to have a consistently harmful effect on
hospitalization rates when compared with the city-wide rates.
Seasonality
of hospitalizations: Only asthma shows a clear seasonal pattern.
ABSTRACT:
Connections between air quality and
environmental health have been well established. However there have been relatively few studies exploring
the contribution of criteria air pollutants (CAP) to respiratory disease and cardiovascular disease in an
ecological framework while using high resolution pollution concentration surfaces derived from dispersion
modeling as the explanatory variables. The analysis will attempt to quantify the effect of CAP
concentrations on hospitalization rates for asthma and cardiovascular diseases (ischemic heart disease,
congestive heart failure, and arrhythmias) in New York City. The hospitalization data will be provided at the
patient level (residential location of each patient). This, in combination with the utilization if the
Cadastral-based Expert Dasymetric System (a technique which disaggregates census data to the tax-lot level)
and the modeled pollution surfaces, provides the opportunity to conduct a fine-grained analysis utilizing
datasets with extremely high spatial resolution.
CEDS:
The Cadastral-based Expert Dasymetric
System (CEDS) estimates total population and specific sub-population distribution for urban areas or any
geographies with sufficient cadastral (tax lot) data in order to develop an improved “denominator,” allowing for
more correct rates and superior estimations of exposure due to residential location in GIS
analyses. Rather than using data
aggregated by arbitrary administrative
boundaries such as census tracts, dasymetric mapping, a disaggregation method
using ancillary information
to delineate areas of homogeneous values, can be used.
CEDS is particularly applicable to hyper-heterogeneous urban
environments, such as that of New York City. This heterogeneity can be particularly troubling
when quantification of an exposed population (e.g. people living near major highways) is heavily biased
within the geographic unit (e.g. the people are not evenly distributed in the census block group). Additionally,
CEDS uses modeling by expert system routines
and validation against various census enumeration units and other data in
order to further refine the
estimates.
PROXIMITY
ANALYSIS: This proximity analysis consisted of delineating fixed
distance buffers around the
known major local sources of air pollution in NYC. These include toxic release
inventory facilities (TRI, ½ mile
buffer), National Emissions Inventory (NEI) stationary point sources (SPS, ¼
mile buffer), limited access
highways (LAH, 150m buffer), and major truck routes (MTR, 150m buffer). CEDS-derived population data was used to estimate “exposed”
and “unexposed” populations. Patient residences
were used to calculate the number of hospitalizations. These rates
(hospitalizations / population)
were then age-adjusted using NYC (2000) as the reference population
DISPERSION
MODELING: AERMOD (American Meteorological Society/Environmental
Protection Agency Regulatory
Model) was used to model the dispersion of pollutants from the National
Emissions Inventory (NEI) point
sources. This model operates by taking environmental variables into account
such as meteorological data
(surface variables: ceiling height, wind direction, wind speed, pressure, temperature, relative humidity, cloud cover - upper air
variables: wind direction, wind speed and temperature), topography, building downwash (structures
close enough to the stack to influence the dispersion), and emission parameters such as emission
rate, gas temperature, emission velocity, stack diameter, and stack height. In this study, a continuous
and cumulative pollution surface will be derived for each CAP of interest for all of New York City and
analyzed vis-à-vis cardiovascular disease and asthma. A receptor grid, which indentifies the points where
AERMOD calculates pollution concentrations, must be defined or created. A regular point grid is
inefficient since there is no reason to calculate concentrations in areas where there is no population.
This issue was resolved by first creating an irregular grid over the NYC based on populated
tax-lots.
DISEASES:
Disease information was acquired from
the Statewide Planning and Research Cooperative System (SPARCS). This data contains all hospital
admissions (i.e. hospitalizations) in the year 2000 where the primary, secondary, or tertiary diagnoses were
asthma, ischemic heart disease, congestive heart failure, or arrhythmias. As can be seen in the figures
below, only the asthma hospitalization
rates show clear seasonality, but
all of the diseases, with the possible exception of conductive disorders, display at least some modest spatial clustering.
CONCLUSIONS:
Proximity analysis is not sufficiently
sensitive to detect the relationship(s) between air quality and health outcomes in New York City. Other
than toxic release inventory facilities which were consistently associated with higher hospitalization
rates of acute myocardial infarction, arrhythmias, heart failure, and asthma, the fixed-distance buffers
were not able to identify any negative health impacts related to other major local sources of
pollution (stationary point sources, limited access highways, and major truck routes) in the city. Important
limitations with the proximity method include the assumption that all facilities/land uses of the same
type will produce equal amounts of pollutants. This is clearly not the case, as certain sources are far
less productive than others. Also, proximity analysis ignores physical properties of the emission sites and
the pollutants themselves, as well as meteorological variables. It is possible that through dispersion
modeling a more realistic estimate of exposure will reveal these expected associations. At the printing of this
poster, this modeling is being processed. Other sources of air quality data, such as ground-based
sensors and satellite remote sensing, are being explored as well.
Proximity
buffers: Local sources of buffers in
pollution in NYC (a),
fixed-distance arts of Manhattan and the Bronx (b), CEDS-derived population (c) and an orthophoto (d)
showing heterogeneity and
affected population in a small area in Upper Manhattan.
(a)
(b)
(c)
(d)
DISPERSION
MODELING CASE STUDY: At present,
modeling has been done for carbon monoxide
(CO) in the Bronx as a pilot area to test the model and its output. Stacks
were defined and plotted for
the NEI sources and the emission parameters assigned to their geographic
locations and imported to
the AERMOD software. Hourly meteorological data was acquired from the National
Climatic Data Center (NCDC) of
the National Oceanic and Atmospheric Administration (NOAA). The highest levels
of modeled CO
concentration emanating from NEI stationary point sources in the Bronx can be
found near the Bronx Zoo.
Other areas of comparatively high CO concentrations include Montefiore Medical
Center and Albert
Einstein College of Medicine.
SELECTED REFERENCES:
1.Brook, R., Franklin, B., Cascio, W., & al., e.
(2004). Air pollution and cardiovascular disease: a statement for healthcare
professionals from the Expert Panel on Population and Prevention Science of the American Heart Association. Circulation, 109, 2655-2675.
2.Chakraborty, J., & Armstrong, M. P. (2004). Thinking
outside the circle: Using geographical knowledge to focus environmental risk
assessment investigations. In W. B. Janelle D, Hansen K. Dordrecht (Ed.), WorldMinds: Geographical Perspectives on 100
Problems. (pp. 435-442). The
Netherlands: Kluwer Academic Publications.
3.Cimorelli, A., Perry, S., Venkatram, A., Weil, J.,
Paine, R., Wilson, R., et al. (2005). AERMOD: A Dispersion Model for
Industrial Source Applications. Part I: General Model Formulation and Boundary Layer Characterization. Journal of Applied Meteorology, 44, 682-693.
4.Cora, M. G., & Hung, Y.-T. (2003). Air Dispersion
Modeling: A Tool for Environmental Evaluation and Improvement. Environmental Quality Management, Spring 2003, 75-86.
5.Dockery, D., Pope, C. I., Xu, X., & al., e. (1993).
An association between air pollution and mortality in six U.S. cities. New England Journal of Medicine 329, 1753-1759.
6.Eicher, C., & Brewer, C. (2001). Dasymetric mapping
and aerial interpolation: implementation and evaluation. Cartography and Geographic Information Science, 28(2), 125-138.
7.Maantay, J., Maroko, A., & Herrmann, C. (2007).
Mapping Population Distribution in the Urban Environment: The Cadastral-based Expert Dasymetric System
(CEDS) Cartography and Geographic Information Science, special issue:
Cartography 2007: Reflections, Status, and Prediction, 34(2), 77-102. .
8.Maantay, J., Maroko, A., & Porter-Morgan, H. (2008).
A New Method for Population Mapping and Understanding the Spatial Dynamics of
Disease in Urban Areas: Asthma in the Bronx, New York. Urban
Geography, 29(6), 1-15.
9.Maantay, J., Tu, J., & Maroko, A. (2008).
Loose-coupling an Air Dispersion Model and a Geographic Information System
(GIS) for Studying Air Pollution and Asthma in the Bronx, New York City. International
Journal of Environmental Health Research, 19(1), 59-71.
ACKNOWLEDGMENTS:
This research is funded by NOAA-CREST
grant # NA06OAR4810162. This research was also partially supported by grant # 2 R25 ES01185-05 from the National Institute of
Environmental Health Sciences of the NIH. Special thanks to NOAA collaborators
Ralph Ferraro and Bruce Ramsay,
NOAA/NESDIS/STAR/Satellite Climate Studies Branch.
