Map-Based Modeling of
People’s Knowledge, Perceptions and Willingness
to Participate in Green Infrastructure Alternatives
to Traditional Storm-Water Management:
A Case Study in Syracuse
Wanjun Peng1, A. Karen Baptiste2 and Lindsay Speer3
1. Master Candidate
State University of New York College of Environmental Science and Forestry
2. Assistant Professor , Colgate University 3. The Partnership for Onondaga Creek
November 17th, 2010
A voice for the Midland Community and the environment
advocating for better, nonpolluting solutions
for Onondaga Creek – Since 2000
Partnership for Onondaga Creek
2009 - lspeer@mrss.com
Environmental
Justice
Green Infrastructure
means…
• Injustice of Midland plant
will not be repeated
• Onondaga Creek & Harbor
Brook will be cleaner
• Community investment
and beautification instead of
further disruption
Midland Sewage Plant
Partnership for Onondaga Creek
2010 - lspeer@mrss.com
Green
Infrastructure
•Solution to capacity problems
with underground storage –
reduce the rain!
•Proposed by Onondaga
Nation
•POC: “If you can stop the
Armory Sewage Plant, you can
stop the Phase III Pipeline too!”
•County Executive Mahoney has
promised that the pipeline will
not be built.
Partnership for Onondaga Creek
2010 - lspeer@mrss.com
Engaging
Community
Dr. April Baptiste
Map of land use / GI possibilities in the
077 sewershed, including University
Neighborhood. Part of presentation by
POC to DEC & Onondaga County, 2008. 
Partnership for Onondaga Creek
2010 - lspeer@mrss.com
• During 2008 and 2009 the POC, Syracuse University and SUNY-ESF
cooperated to obtain over 200 surveys among five neighborhoods of the
Midland and Clinton Sewershed to assess public attitudes toward Green
Infrastructure (GI) Implementation.
Objective
• Explore to what extent individual demographics
influence environmental attitudes toward GI
• Explore at what level neighborhood contexts are
statistically related to environmental attitudes toward GI
• Use maps to visualize environmental attitudes toward GI
implementation across neighborhoods
* Three building blocks of environmental attitudes:
Environmental knowledge
Environmental perception
Environmental behavioral intention.
Methods---At individual level
• Summarize responses using
descriptive statistics in Minitab
software
• Look for factors that explain
responses using regression models
in SAS software
Methods---At individual level
• Summarize responses using
Descriptive Statistics in
Minitab software
• Look for influence of
demographics, knowledge,
perception of benefits and
cost on responses using
Regression Models in SAS
software
Methods---At neighborhood level
• Divide survey responses
by neighborhood
• Apply regression
models in SAS software
to explore the influence
of neighborhood
context
• Map predicted
responses (from
models) across the city
using ArcMap Software
RESULTS at the Individual Level:
The effect of demographic background and perception of benefits and cost
on behavioral intention to implement GI if provided for free
(216 Individual Surveys)
Overall GI (All), rain barrels (RB), trees (T), rain garden (RG),
porous driveway/sidewalks (P) and curbside extensions (CE).
Participation likelihood for free GI
All
Residency
Ownership
Age
Education
Ethnicity
Gender
Knowledge
+
The influence of benefit
The influence of cost
+
-
RB
T
RG
P
*
CE
*
*
+
+
*
+
+
+
+
+
+
+
-
+
RESULTS at the Individual Level:
The effect of demographic background and perception of benefits and cost
on behavioral intention to implement GI if savings on water bills ensued
(216 Individual Surveys)
Overall GI (All), rain barrels (RB), trees (T), rain garden (RG),
porous driveway/sidewalks (P) and curbside extensions (CE).
Participation likelihood for savings
All
RB
T
RG
P
CE
Residency
Ownership
Age
Education
Ethnicity
Gender
Knowledge
+
+
+
+
+
The influence of benefit
The influence of cost
+
+
+
+
+
-
RESULTS at the Neighborhood Level:
The effect of neighborhood context on respondents’ behavioral
intention to implement GI if provided for free
(17 block groups with 136 Individual Surveys)
Overall GI (All), rain barrels (RB), trees (T), rain garden (RG),
porous driveway/sidewalks (P) and curbside extensions (CE).
Participation likelihood for free GI
All
RB
T
RG
P
CE
+
+
Neighborhood Canopy Cover (%)
Neighborhood Imp. Surface (%)
The percent of houses
occupied by owner
+
The percent of African Americans
The percent of females
Average age of neighbors
The percent below poverty level
+
+
+
+
RESULTS at the Neighborhood Level:
The effect of neighborhood context on respondents’ behavioral
intention to implement GI if they were to realize savings on their water
bill
(17 block groups with 136 Individual Surveys)
Overall GI (All), rain barrels (RB), trees (T), rain garden (RG),
porous driveway/sidewalks (P) and curbside extensions (CE).
Participation likelihood for savings
All
RB
T
RG
P
CE
Canopy percentage
Impervious percentage
The percent of houses
occupied by owner
The percent of African
Americans
The percent of female
Age
The percent of below poverty
-
+
-
-
+
Spatial Prediction
•
Model estimation of participation likelihood in overall GI implementation, if
free, in Syracuse, NY (10-25: low to high)
Spatial Prediction
•
Model estimation of participation likelihood in overall GI implementation, if
savings on water bills were to occur, in Syracuse, NY (10-25: low to high)
Conclusion
• At the individual level, the likelihood of participation is
significantly related to knowledge and perceived benefit of
GI.
• At the neighborhood scale, the variation in willingness to
implement any GI, with the exception of curbside
extension, can be explained primarily by a higher existing
tree canopy cover, a lower percent of African American
population, and an older population.
• The findings can provide valuable information to
government officials working to find cheaper, publically
acceptable, and environmentally friendly ways to control
storm water management.
Acknowledge
I would like to thank Myrna Hall and Stephen
V. Stehman from SUNY ESF for their guidance and
expertise during my research in spite of their tight
schedule.