1. Introduction
The present urban environment, especially in the U.S. tends to be designed for cars rather than human. The low density and high level of separation of different age group, different income group and different use also the separation from natural environment actually make the city a big prison where residents are trapped inside and have nowhere to escape to. When you live in a city where without a car you cannot anywhere (which is the case for lots of U.S. cities), when you are not old enough or too old to drive a car, then you are totally trapped. You can connect your friends through phones, facebook and other electric based methods, but you can never see them in person. The only thing you can do is to wait until your parents come back and willing to take you somewhere. This may be one of the reasons why people get bad impression on teenage drivers. After years of “jail" they are so desperate to rush their way somewhere else. If you are too old, you are thrown into special elder citizen residential facilities far away from normal social life. From then on you are never going to see the younger. People of high income live in their “Garden of Eden", completely separated from their low-income neighbors, thinking that they were all evil persons. There are always a certain percentage of good people and bad people in an each social group. Bias, scare and conflicts grow as a result of high level of separation and resulted lacking of contact.
These problems are believed to be associated with a poorly designed built environment and an un healthy (or “dysfunctional”) mode of development, commonly addressed as Urban Sprawl which has properties as: low density, rigidly separated land use, inaccessible huge blocks and an absence of vibrant city center (Reid et al., 2003) . A serious of multi-dimensional social, economic, environment and health problems are believed to be a result of Urban Sprawl as follows: higher driving distance, high car ownership, more polluted air, higher car accident death rate , less physical activity, high obesity rate, and high hypertension (Reid et al., 2003).
The present urban environment, especially in the U.S. tends to be designed for cars rather than human. The low density and high level of separation of different age group, different income group and different use also the separation from natural environment actually make the city a big prison where residents are trapped inside and have nowhere to escape to. When you live in a city where without a car you cannot anywhere (which is the case for lots of U.S. cities), when you are not old enough or too old to drive a car, then you are totally trapped. You can connect your friends through phones, facebook and other electric based methods, but you can never see them in person. The only thing you can do is to wait until your parents come back and willing to take you somewhere. This may be one of the reasons why people get bad impression on teenage drivers. After years of “jail" they are so desperate to rush their way somewhere else. If you are too old, you are thrown into special elder citizen residential facilities far away from normal social life. From then on you are never going to see the younger. People of high income live in their “Garden of Eden", completely separated from their low-income neighbors, thinking that they were all evil persons. There are always a certain percentage of good people and bad people in an each social group. Bias, scare and conflicts grow as a result of high level of separation and resulted lacking of contact.
These problems are believed to be associated with a poorly designed built environment and an un healthy (or “dysfunctional”) mode of development, commonly addressed as Urban Sprawl which has properties as: low density, rigidly separated land use, inaccessible huge blocks and an absence of vibrant city center (Reid et al., 2003) . A serious of multi-dimensional social, economic, environment and health problems are believed to be a result of Urban Sprawl as follows: higher driving distance, high car ownership, more polluted air, higher car accident death rate , less physical activity, high obesity rate, and high hypertension (Reid et al., 2003).
2. Methodology
One serious problem of urban sprawl is its degrading effects on public health. Causal relationships were found between walking time, BMI, and hypertension (Reid et al., 2003). The focus of this project is to discuss the influence of urban sprawl, one of the serious problems caused by poorly-designed modern built environment, on public health. The analysis was done on the counties within Pennsylvania.
2.1 Urban Sprawl Index Metrics
Urban sprawl is commonly evaluated with sprawl index, which include the following four aspects: “Residential Density”, “Land Use Mix Factor”, “Centralization Level” and “Street Network Accessibility” (Reid & Rolf Pendall, 2006).
2.1.1 Residential Density
This factor is measured with 4 variables: population density of county, low density (lower than 1500 persons per square mile) population of the total population, high density (higher than 12500 persons per square mile) population of the total population and urban area population.
All of these variables are indicators of the dwelling density of the total population. Population density of a county is calculated by summing up the population of each county from block group. Low density population percentage and high density population percentage are descriptions of population density distribution pattern. Urban Area is, from the definition of census bureau, is places with “Places of 2,500 or more persons incorporated as cities, villages, boroughs (except in Alaska and New York), and towns” (http://www.census.gov/population/censusdata/urdef.txt). According to census, densely populated is a necessary but not sufficient for a place to be referred to as an urban area. So this variable reflects more than just density, but also some other features of urbanization. The urban area density is also measured by summing up population from block group level.
2.1.2 Land Use Mix Factor
This metric describes the accessibility of everyday life facilities of working, shopping, education, recreational activities, etc. Mixed land use is one of the important principles of “Smart Growth”. Mixed functions in a neighborhood can generate more communications and increases the level of safety and quality of life (http://www.smartgrowth.org/principles/mix_land.php) and is also believed to influence human behavior and is related to people’s health status. The variables in this metric discussed in the project are grocery shopping and public school accessibility with service area of one mile radius. Grocery shop data is obtained from ReferenceUSA and School information is from CCD. The process goes as: geocoding the facility, using buffer analysis to calculate the service area of the facility, summing up the population falls inside the service area of the facility and finally comparing this population with the total population.
2.1.3 Centralization Level
Centralization is the opposite status of Urban Sprawl, the more centralized, the less sprawl an area is. The variables studied in this metric are: (1) population density variation across block group and (2) density gradient. Population density variation is calculated by standard deviation divided by average population density of block group. Density gradient is visualized with Kernal Density Smooth Raster map generated from block group with population value on their centroids.
Street Network Accessibility One of the direct evident of a sprawled built environment is inaccessible street pattern with very small number of intersections and huge blocks. This leads to a car dominant travel pattern and results in a total block out of people without an access to cars as children and senior citizens. Variables studied in this metric are: (1) average block length and (2) small block (with a length of less than 500 feet) percentage.
2.2 Health Status
The Health Status Data is obtained from the “County Health Ranking & Road Map”. The candidate health outcome variables might be influenced by Urban Sprawl was chosen as follows:
• Premature Death
• Physical Health Status("Poor Health Percent" and "Physical Unhealthy Days")
• Mental Health Status ("Mentally Unhealthy Days")
• Unhealthy Behavior ("Smoker Rate", "Excessive Drinking Rate")
• Obesity Rate
• Diabetes Rate
• "Physical inactiveness"
• Motor Vehicle Death
The data source of that are a combination of “National Center for Health Statistics”, “BRFSS” of CDC and “Dartmouth Atlas of Health Care”. The county level health data was directly aggregated to the map.
3. Results
3.1 Urban Sprawl Metric
One serious problem of urban sprawl is its degrading effects on public health. Causal relationships were found between walking time, BMI, and hypertension (Reid et al., 2003). The focus of this project is to discuss the influence of urban sprawl, one of the serious problems caused by poorly-designed modern built environment, on public health. The analysis was done on the counties within Pennsylvania.
2.1 Urban Sprawl Index Metrics
Urban sprawl is commonly evaluated with sprawl index, which include the following four aspects: “Residential Density”, “Land Use Mix Factor”, “Centralization Level” and “Street Network Accessibility” (Reid & Rolf Pendall, 2006).
2.1.1 Residential Density
This factor is measured with 4 variables: population density of county, low density (lower than 1500 persons per square mile) population of the total population, high density (higher than 12500 persons per square mile) population of the total population and urban area population.
All of these variables are indicators of the dwelling density of the total population. Population density of a county is calculated by summing up the population of each county from block group. Low density population percentage and high density population percentage are descriptions of population density distribution pattern. Urban Area is, from the definition of census bureau, is places with “Places of 2,500 or more persons incorporated as cities, villages, boroughs (except in Alaska and New York), and towns” (http://www.census.gov/population/censusdata/urdef.txt). According to census, densely populated is a necessary but not sufficient for a place to be referred to as an urban area. So this variable reflects more than just density, but also some other features of urbanization. The urban area density is also measured by summing up population from block group level.
2.1.2 Land Use Mix Factor
This metric describes the accessibility of everyday life facilities of working, shopping, education, recreational activities, etc. Mixed land use is one of the important principles of “Smart Growth”. Mixed functions in a neighborhood can generate more communications and increases the level of safety and quality of life (http://www.smartgrowth.org/principles/mix_land.php) and is also believed to influence human behavior and is related to people’s health status. The variables in this metric discussed in the project are grocery shopping and public school accessibility with service area of one mile radius. Grocery shop data is obtained from ReferenceUSA and School information is from CCD. The process goes as: geocoding the facility, using buffer analysis to calculate the service area of the facility, summing up the population falls inside the service area of the facility and finally comparing this population with the total population.
2.1.3 Centralization Level
Centralization is the opposite status of Urban Sprawl, the more centralized, the less sprawl an area is. The variables studied in this metric are: (1) population density variation across block group and (2) density gradient. Population density variation is calculated by standard deviation divided by average population density of block group. Density gradient is visualized with Kernal Density Smooth Raster map generated from block group with population value on their centroids.
Street Network Accessibility One of the direct evident of a sprawled built environment is inaccessible street pattern with very small number of intersections and huge blocks. This leads to a car dominant travel pattern and results in a total block out of people without an access to cars as children and senior citizens. Variables studied in this metric are: (1) average block length and (2) small block (with a length of less than 500 feet) percentage.
2.2 Health Status
The Health Status Data is obtained from the “County Health Ranking & Road Map”. The candidate health outcome variables might be influenced by Urban Sprawl was chosen as follows:
• Premature Death
• Physical Health Status("Poor Health Percent" and "Physical Unhealthy Days")
• Mental Health Status ("Mentally Unhealthy Days")
• Unhealthy Behavior ("Smoker Rate", "Excessive Drinking Rate")
• Obesity Rate
• Diabetes Rate
• "Physical inactiveness"
• Motor Vehicle Death
The data source of that are a combination of “National Center for Health Statistics”, “BRFSS” of CDC and “Dartmouth Atlas of Health Care”. The county level health data was directly aggregated to the map.
3. Results
3.1 Urban Sprawl Metric
3.2 Health Status
3.3 General Findings
For the four metrics of input parameters, the patterns of the color distribution within each set are pretty similar, which indicate the interconnection between variables within each metric set. For the output health metrics, the Premature Death Rate, Car Crash Death Percent, Diabetes Percentage, Physical Inactive Percentage, Physical Unhealthy Day and Poor Health Percentage possess similar patterns to some extend and show some relationship with the urban sprawl measurement as block size and grocery store accessibility.
3.4 Interesting and Unexpected Relations
In this part of the analysis, I picked out several groups of input Urban Sprawl variables and the output health variables and intend to verify some relationship by visualize them within the same map using gradual color and gradual symbol.
3.4.1 Average Block Length and Diabetes Rate
In this part of the analysis, I picked out several groups of input Urban Sprawl variables and the output health variables and intend to verify some relationship by visualize them within the same map using gradual color and gradual symbol.
3.4.1 Average Block Length and Diabetes Rate
From the map above, we can see the average block length mainly follows the pattern of diabetes rate, especially in the northern and middle of the south of the State. There is one exception right in the center of Pennsylvania, which is the location of the “Centre County”.
3.4.2 Large Block Percentage and Car Crash Death Rate
3.4.2 Large Block Percentage and Car Crash Death Rate
The high percentage of large block (with a length greater than 500 feet) in a County is related to its car crash death rate. Again, the “Centre County” is an exception.
3.4.3 Poor Access to Grocery Store and Car Crash Death Rate
3.4.3 Poor Access to Grocery Store and Car Crash Death Rate
It is natural to think that with poor access to grocery store, one tends to be more prone to weight problem, but actually, from the analysis shows that grocery store accessibility and the obesity rate doesn’t have a significant relationship. But surprisingly, the poor grocery store accessibility highly accords with car crash death rate.
4. Conclusion
From this analysis of Urban Sprawl metric and public health, several pattern are revealed:
• health variables of Premature Death Rate, Car Crash Death Percentage, Diabetes Percentage, and Physical health status measurement of Physical Inactive Percentage, Physical Unhealthy Day and Poor Health Percentage share similar pattern to some extent
• The Street Network Accessibility variables of average block length and large block percentage have some relationship with Diabetes and Car Crash Death Rate
• Poor accessibility to grocery store is significantly related to Car Crash Death Rate
5. Limitations
• Some confounding factor as socioeconomic feature, generation and age construction of the studied area are not controlled.
• Patterns are mainly judged by visualization, further quantities should be done to further prove those findings.
6. Data Source
• Geographic and Population:
o PA County Boundary and PA state road and PA local road shapefile
(intended to be used in network analysis but haven't succeeded) from PASDA
(http://www.pasda.psu.edu/uci/SearchResults.aspx?shortcutKeyword=base&search&sessionID=260115920201312820440)
o Block Group shapefile and Block shapefile from Census Bureau: http://www.census.gov/geo/maps-data/data/tiger-line.html
o Democratic information and some user tools for retriving data from census bureau:http://www2.census.gov/acs2011_5yr/summaryfile/
o Urban Area shapefile from census: ftp://ftp2.census.gov/geo/tiger/TIGER2013/UAC/
o PA County Boundary and PA state road and PA local road shapefile (intended to be used in network analysis but haven't succeeded) from PASDA (http://www.pasda.psu.edu/uci/SearchResults.aspx?shortcutKeyword=base&searchType=shortcut&sessionID=260115920201312820440)
• GroceryStore informationo from referenceUSA: http://www.referenceusa.com/UsBusiness/Result/9a6f3c2da04c4d1badd411a9c0b026c9
• Health Status Data on County Level:
http://www.countyhealthrankings.org/app/pennsylvania/2013/downloads
• School Information from CCD: http://nces.ed.gov/ccd/elsi/tableGenerator.aspx
BibliographyReid et al., E. (2003). Relationship Between Urban Sprawl and Physical Activity, Obesity, and Morbidity. American Journal of Health Promotion, 45-47.
Reid, E., & Rolf Pendall. (2006). Measuring Sprawl and Its Transportation Impacts . Transportation Research Record: Journal of the Transportation Research Board, 175 - 183.
4. Conclusion
From this analysis of Urban Sprawl metric and public health, several pattern are revealed:
• health variables of Premature Death Rate, Car Crash Death Percentage, Diabetes Percentage, and Physical health status measurement of Physical Inactive Percentage, Physical Unhealthy Day and Poor Health Percentage share similar pattern to some extent
• The Street Network Accessibility variables of average block length and large block percentage have some relationship with Diabetes and Car Crash Death Rate
• Poor accessibility to grocery store is significantly related to Car Crash Death Rate
5. Limitations
• Some confounding factor as socioeconomic feature, generation and age construction of the studied area are not controlled.
• Patterns are mainly judged by visualization, further quantities should be done to further prove those findings.
6. Data Source
• Geographic and Population:
o PA County Boundary and PA state road and PA local road shapefile
(intended to be used in network analysis but haven't succeeded) from PASDA
(http://www.pasda.psu.edu/uci/SearchResults.aspx?shortcutKeyword=base&search&sessionID=260115920201312820440)
o Block Group shapefile and Block shapefile from Census Bureau: http://www.census.gov/geo/maps-data/data/tiger-line.html
o Democratic information and some user tools for retriving data from census bureau:http://www2.census.gov/acs2011_5yr/summaryfile/
o Urban Area shapefile from census: ftp://ftp2.census.gov/geo/tiger/TIGER2013/UAC/
o PA County Boundary and PA state road and PA local road shapefile (intended to be used in network analysis but haven't succeeded) from PASDA (http://www.pasda.psu.edu/uci/SearchResults.aspx?shortcutKeyword=base&searchType=shortcut&sessionID=260115920201312820440)
• GroceryStore informationo from referenceUSA: http://www.referenceusa.com/UsBusiness/Result/9a6f3c2da04c4d1badd411a9c0b026c9
• Health Status Data on County Level:
http://www.countyhealthrankings.org/app/pennsylvania/2013/downloads
• School Information from CCD: http://nces.ed.gov/ccd/elsi/tableGenerator.aspx
BibliographyReid et al., E. (2003). Relationship Between Urban Sprawl and Physical Activity, Obesity, and Morbidity. American Journal of Health Promotion, 45-47.
Reid, E., & Rolf Pendall. (2006). Measuring Sprawl and Its Transportation Impacts . Transportation Research Record: Journal of the Transportation Research Board, 175 - 183.
RSS Feed
